From dave.banham@tde.alstom.com Mon Oct 2 02:45:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Mon, 02 Oct 2000 02:45:00 -0000 Subject: Q: Where is the symbol window ? Message-ID: Fabrice, The F4 key only works if you don't have num-lock selected! (I think this bug has already been reported.) I take your point about using the file list view in the symbol window, but I very rarely do this because I am not using SN as an IDE tool - I only us it to find program symbols and their relationships. However, on the odd occasion that I have used the file list view I have found that filtering it with a file name pattern that includes the extension doesn't work. e.g. '*fred.h' I have even tried to escape the dot with a slash, but have had no success. So I end up using '*fred*' instead. Regards Dave Banham > If you are in the 'navigation' symbol window you can get the > 'Symbols' window > back by pressing F4 or from the menu bar as Windows->New > Window->Symbols. But F4 doesn't seem to work on the WinNT version. > I personally find the 'Symbols' window to be of little use > because I use the > retriever to look up specific symbol and to show its position > in the Editor > window, which I have added as a second pane. Except the retriever doesn't allow you to search for a specific file. -- Fabrice Gautier fabrice_gautier@sdesigns.com From dkirby@orchestream.com Mon Oct 2 07:17:00 2000 From: dkirby@orchestream.com (Kirby, Dave) Date: Mon, 02 Oct 2000 07:17:00 -0000 Subject: Some Python (and other) questions Message-ID: I have recently upgraded to 4.5.2 for its Python support, and have some questions. Firstly, the python support seems to be only partial - nothing shows up in the class view and the Xref and Include tabs are greyed out. However the information for the classes, methods etc are in the database since they can be seen in the Retriever view, so what gives? Is this a bug in the python parser? Secondly, has anyone tried creating python wrappers for the SN APIs? This would make it easy to write scripts for analysing the information in the DB (as suggested by a previous poster) and for creating parsers written in python. I may have a go myself when I have time, but dont dont want to duplicate the work if someone else has already done it. It should be easy using SWIG, and this would also enable the APIs to be called from TCL, Perl, Guile etc. Lastly, why doesnt the mailing list digest set the 'Reply To' field? It is a minor point, but it would make the list much easier to use. Dave K From lohr@laas.fr Mon Oct 2 10:09:00 2000 From: lohr@laas.fr (Christophe Lohr) Date: Mon, 02 Oct 2000 10:09:00 -0000 Subject: how may I find unused methodes ? Message-ID: <200010021708.e92H8qp20573@cruzeiro.laas.fr> Hi ! How may I find unused methodes in my C++ project with source navigator ? (just like "lint" do for C code) How can I save call tree of methodes (Xref window) in a textual file ? How can I save list of all methodes (Symbols window) in a textual file ? Thanks for your help. Christophe. From spolk@redhat.com Mon Oct 2 10:41:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 02 Oct 2000 10:41:00 -0000 Subject: Some Python (and other) questions References: Message-ID: <4.2.0.58.20001002104207.00d796e0@pop.cygnus.com> At 03:14 PM 10/2/00 +0100, Kirby, Dave wrote: >I have recently upgraded to 4.5.2 for its Python support, and have some >questions. > >Firstly, the python support seems to be only partial - nothing shows up in >the class view and the Xref and Include tabs are greyed out. However the >information for the classes, methods etc are in the database since they can >be seen in the Retriever view, so what gives? Is this a bug in the python >parser? You need to ask the original author. >Secondly, has anyone tried creating python wrappers for the SN APIs? This >would make it easy to write scripts for analysing the information in the DB >(as suggested by a previous poster) and for creating parsers written in >python. I may have a go myself when I have time, but dont dont want to >duplicate the work if someone else has already done it. It should be easy >using SWIG, and this would also enable the APIs to be called from TCL, Perl, >Guile etc. We welcome the contributiions. >Lastly, why doesnt the mailing list digest set the 'Reply To' field? It is >a minor point, but it would make the list much easier to use. None of the sources.redhat.com lists do this that I am aware. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From thomas.heller@ion-tof.com Mon Oct 2 12:22:00 2000 From: thomas.heller@ion-tof.com (Thomas Heller) Date: Mon, 02 Oct 2000 12:22:00 -0000 Subject: Some Python (and other) questions References: <4.2.0.58.20001002104207.00d796e0@pop.cygnus.com> Message-ID: <004d01c02c9d$90ad0680$4500a8c0@thomas> > >I have recently upgraded to 4.5.2 for its Python support, and have some > >questions. > > > >Firstly, the python support seems to be only partial - nothing shows up in > >the class view and the Xref and Include tabs are greyed out. However the > >information for the classes, methods etc are in the database since they can > >be seen in the Retriever view, so what gives? Is this a bug in the python > >parser? > > You need to ask the original author. Ok, so here we go: The python parser is modeled after pyclrbr.py, with the improvement that it also finds global variables (apart from classes, methods and functions). No xref info at all (simply because I have no idea how to do this). No import statements are parsed, so no include info is generated. > > >Secondly, has anyone tried creating python wrappers for the SN APIs? This > >would make it easy to write scripts for analysing the information in the DB > >(as suggested by a previous poster) and for creating parsers written in > >python. I may have a go myself when I have time, but dont dont want to > >duplicate the work if someone else has already done it. It should be easy > >using SWIG, and this would also enable the APIs to be called from TCL, Perl, > >Guile etc. Thoma From dave.banham@tde.alstom.com Tue Oct 3 00:49:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Tue, 03 Oct 2000 00:49:00 -0000 Subject: how may I find unused methodes ? Message-ID: > How may I find unused methodes in my C++ project with source navigator ? >(just like "lint" do for C code) SN is a code comprehension (and IDE) tool and not a static analysis tool. Moreover, there are lots of improvements to be made to SN's code comprehension facilities before we even start to consider static analysis. I am also under the impression that Lint will report on unused classes, objects, methods, etc. in C++, just as it reports on unused items in C (although I have only used SN and Lint on my embedded C projects). Have a look at PC-Lint by Gimpel: www.gimpel.com. Regards Dave Banham From dave.banham@tde.alstom.com Tue Oct 3 00:56:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Tue, 03 Oct 2000 00:56:00 -0000 Subject: Some Python (and other) questions Message-ID: >>Lastly, why doesnt the mailing list digest set the 'Reply To' field? It is >>a minor point, but it would make the list much easier to use. >None of the sources.redhat.com lists do this that I am aware. The reason for this (I guess), and the setting of the 'reply to' field, in the non-digest version, to the originating author rather than the SN mailing list e-mail address, is to avoid automated 'out of the office', etc. replies junking up the list and everyone's e-mail. If you want to reply to the list you must explicitly set the 'send-to' field to that of the list otherwise you will just be replyingto the original author - off-line. As a member of several other mailing lists which do not do this, I can attest to the amount of automated junk that I receive. So please Red Hat do not change this policy! Regards Dave Banham From mdejong@cygnus.com Tue Oct 3 09:44:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 03 Oct 2000 09:44:00 -0000 Subject: Source navigator roadmap, features list, etc.? References: Message-ID: On Thu, 28 Sep 2000, Ben Elliston wrote: > I agree completely. I love SN, but it does have a lot of rough edges > in the UI, and it's buggy. I, too, have a lot of suggestions and bug > reports and would like to know what the future of SN is going to be > > IMHO, it should be a relatively high priority to establish a GNATS bug > reporting database for S-N (as other projects on sources.redhat.com do). > This will encourage folks to submit bug reports and for us to track them. > > Ben We already have one, but we do not want to "roll it out" until 5.0 is ready. Mo DeJong Red Hat Inc From mdejong@cygnus.com Tue Oct 3 16:02:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 03 Oct 2000 16:02:00 -0000 Subject: Source navigator roadmap, features list, etc.? References: <5D2A198567EED311BD0F009027A25C87402387@SRVMAIL-ANN> Message-ID: On Wed, 27 Sep 2000, William Gacquer wrote: > Yahoo! The community rules! > > More seriously, if I can give you my wishlist... > > 1/ automatic recognition of new files in the project directories Not sure about that one. Perhaps a feature to rescan for new files would be a good idea, but it should not be automatic. > 2/ a better interface with CVS ( why not an integration of tkCVS in SN?) This one is a well known problem that we need to deal with. > 3/ an expand/collapse button in the project tree view (expansion is the > default but my project is so big that I would prefer a collapse by default) > 4/ Drag&Drop with KDE/Gnome (I known, that's Tk's stuff) What exactly are you itching to drop into SN? Drag and Drop is something that seems really cool but I find to be less than useful unless it is very consistent across applications. > 5/ a UML-istic class view (even if it's not editable) > 6/ clicking on xref-ed objects/functions opens the file where the > object/function is used (by default, it opens the file a function > implementation) Yup, I looked into this before but I could not fix it without a redesign. > 7/ a clever "replace-in-files" > 8/ an interface with GNAT (with tkGNAT?) > 9/ a autoconf/automake project creation as in kdevelop This one is hard to do in a general way, but we could do something like writing a configure.in for the user in the same way that we write a Makefile. > 10/ a possibility to hide/show all SN-controlled windows This is already implemented, see the Windows->Iconize Project menu item. > 11/ a search-and-replace dialog that can jump directly to the next occurence > by clicking on replace > 12/ working real-regular-expressions everywhere The trick will be how the user can tell the system to use a regexp vs a glob for a given entry box. Mo DeJong Red Hat Inc From leonp@plris.com Wed Oct 4 00:53:00 2000 From: leonp@plris.com (leonp@plris.com) Date: Wed, 04 Oct 2000 00:53:00 -0000 Subject: Source navigator roadmap, features list, etc.? Message-ID: <4.3.2.7.0.20001004105753.00b142d8@plris.com> At 16:02 03/10/2000 -0700, Mo DeJong wrote: >On Wed, 27 Sep 2000, William Gacquer wrote: > > 9/ a autoconf/automake project creation as in kdevelop > >This one is hard to do in a general way, but we could do >something like writing a configure.in for the user in the >same way that we write a Makefile. As far as I was able to understand the SN was designed for use also (mainly?) in Embedded Programming, where you need to have a little bit simpler and closer approach to a lot of build process parameters. Absence of this simple approach in kdevelop (which is very good IDE!) was the main reason that I left my attempts to sufficiently and comfortably integrate the embedded projects into it, while one of the main difficulties was this autoconf/automake stuff. I shall be very disappointed if this issue will be introduced also in SN, which is my only IDE today. Thanks to Cygnus/RH for the good SW. Leon Pollak leonp@plris.com From tshead@k-3d.com Wed Oct 4 02:42:00 2000 From: tshead@k-3d.com (Timothy M. Shead) Date: Wed, 04 Oct 2000 02:42:00 -0000 Subject: Source navigator roadmap, features list, etc.? References: <4.3.2.7.0.20001004105753.00b142d8@plris.com> Message-ID: <39DAFB81.9010107@k-3d.com> leonp@plris.com wrote: > At 16:02 03/10/2000 -0700, Mo DeJong wrote: > >> On Wed, 27 Sep 2000, William Gacquer wrote: >> > 9/ a autoconf/automake project creation as in kdevelop >> >> This one is hard to do in a general way, but we could do >> something like writing a configure.in for the user in the >> same way that we write a Makefile. > > As far as I was able to understand the SN was designed for use also > (mainly?) in Embedded Programming, where you need to have a little bit > simpler and closer approach to a lot of build process parameters. > Absence of this simple approach in kdevelop (which is very good IDE!) > was the main reason that I left my attempts to sufficiently and > comfortably integrate the embedded projects into it, while one of the > main difficulties was this autoconf/automake stuff. I shall be very > disappointed if this issue will be introduced also in SN, which is my > only IDE today. Remember that you can always use external commands to build any way you like - I prefer to use my own makefiles, so my "Build Command" is set to "make", "Directory" to my top-level project directory, and "Build Targets" to "". This shouldn't make anyone nervous :) Timothy M. Shead From spolk@redhat.com Wed Oct 4 10:25:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 04 Oct 2000 10:25:00 -0000 Subject: Source navigator roadmap, features list, etc.? References: <5D2A198567EED311BD0F009027A25C87402387@SRVMAIL-ANN> Message-ID: <4.2.0.58.20001004102624.00dad380@pop.cygnus.com> At 04:02 PM 10/3/00 -0700, Mo DeJong wrote: >On Wed, 27 Sep 2000, William Gacquer wrote: > > > Yahoo! The community rules! > > > > More seriously, if I can give you my wishlist... > > > > 1/ automatic recognition of new files in the project directories > >Not sure about that one. Perhaps a feature to rescan for new files >would be a good idea, but it should not be automatic. I think that we should have a checkbox on the directory selection dialog that says "Live directory". If this is checked, the directory will be scanned every time the project is refreshed. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From berek@usa.net Wed Oct 4 13:27:00 2000 From: berek@usa.net (Berek) Date: Wed, 04 Oct 2000 13:27:00 -0000 Subject: SN Printing Under WIN/NT References: <5D2A198567EED311BD0F009027A25C87402387@SRVMAIL-ANN> <4.2.0.58.20001004102624.00dad380@pop.cygnus.com> Message-ID: <005501c02e41$70241a20$7113a8c0@concord.com> Lot's of problems printing hierarchy and cross reference charts under NT. I have SN installed on 1/2 dozen PC's. Some use Postscript drivers, others PCL, some use both. All have same results printing on 1/2 dozen networked printers. Basically, printing is garbled: letters, symbols, and lines overlaying each other. Have to resort to doing NT alt-print screen and pasting into MS Word, then printing from Word. In Class view, C++ class members that are "private" have symbol (transparent diamond) for "public". Private members should have no symbol at all. Both public and private have the public symbol. I'm concerned about comments made by SN Development (Syd Polk?) about priority given to Perl parser. I know of more than 1/2 dozen companies in my area that have been forced to scale back use of SN because of many problems with C++ parser: 3Com, Lucent, Cienna, Concord Communications, Broadband Access Systems, Cisco, Fidelity... Unless C++ parsing bugs are fixed, SN cannot be relied upon as a development tool. We desparately need accurate cross references. Those of us that us SN are forced to fall back on grep's. Please review your priorities. Tx Steve Dow ----- Original Message ----- From: "Syd Polk" To: Sent: Wednesday, October 04, 2000 13:27 Subject: RE: Source navigator roadmap, features list, etc.? > At 04:02 PM 10/3/00 -0700, Mo DeJong wrote: > >On Wed, 27 Sep 2000, William Gacquer wrote: > > > > > Yahoo! The community rules! > > > > > > More seriously, if I can give you my wishlist... > > > > > > 1/ automatic recognition of new files in the project directories > > > >Not sure about that one. Perhaps a feature to rescan for new files > >would be a good idea, but it should not be automatic. > > I think that we should have a checkbox on the directory selection dialog > that says "Live directory". If this is checked, the directory will be > scanned every time the project is refreshed. > > > Syd Polk spolk@redhat.com > Engineering Manager +1 415 777 9810 x 241 > Red Hat, Inc. > > > > From Fabrice_Gautier@sdesigns.com Wed Oct 4 14:25:00 2000 From: Fabrice_Gautier@sdesigns.com (Fabrice Gautier) Date: Wed, 04 Oct 2000 14:25:00 -0000 Subject: Q: Where is the symbol window ? Message-ID: <8AE4B526B977D411841F00A0CC334020052C83@cuz-exchange.sdesigns.net> > > > Fabrice, > The F4 key only works if you don't have num-lock selected! (I > think this bug has > already been reported.) Thanks ... But I do use the keypad. So ALT-W N S will have to do... > I have used the file list view I have found that filtering it > with a file name > pattern that includes the extension doesn't work. e.g. > '*fred.h' I have even > tried to escape the dot with a slash, but have had no > success. So I end up using > '*fred*' instead. Yes, I noticed that one too, but I would use *fred .h* in your case. Regards -- Fabrice Gautier fabrice_gautier@sdesigns.com From mdejong@cygnus.com Wed Oct 4 15:27:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 04 Oct 2000 15:27:00 -0000 Subject: SN Printing Under WIN/NT References: <005501c02e41$70241a20$7113a8c0@concord.com> Message-ID: > I'm concerned about comments made by SN Development (Syd Polk?) about > priority given to Perl parser. Syd was talking about a possible contribution from someone in the net community. We are not going to write a perl parser unless someone contracts with us to write it. > I know of more than 1/2 dozen companies in my > area that have been forced to scale back use of SN because of many problems > with C++ parser: 3Com, Lucent, Cienna, Concord Communications, Broadband > Access Systems, Cisco, Fidelity... Unless C++ parsing bugs are fixed, SN > cannot be relied upon as a development tool. Humm, if all of these companies are having problems why have they not contacted us about a support/development contract to improve the C++ parser support? > We desparately need accurate > cross references. Those of us that us SN are forced to fall back on grep's. > Please review your priorities. Our current priority is Source-Navigator 5.0. If your company needs some particular feature, you have the following options: 1. Implement it yourself 2. Contract with us to implement it 3. Hire someone else to do it cheers Mo DeJong Red Hat Inc From paul@toolscenter.org Wed Oct 4 16:50:00 2000 From: paul@toolscenter.org (Paul Selormey) Date: Wed, 04 Oct 2000 16:50:00 -0000 Subject: SN Printing Under WIN/NT References: Message-ID: <002101c02e5e$2b5750c0$60a65a8f@paulselo> Hello Mo DeJong, > > [...] > Humm, if all of these companies are having problems why have they > not contacted us about a support/development contract to improve > the C++ parser support? Please how much will it cost to contract you. I think many here will help to contribute that amount which will give us a good parser for C/C++ in parser and Java where possible. > Our current priority is Source-Navigator 5.0. If your company needs > some particular feature, you have the following options: We have heard much about this, but it is not building on the current existing codes? What is new here? since no one seems to know what is going on except the RedHat staff. > 1. Implement it yourself > 2. Contract with us to implement it > 3. Hire someone else to do it Now, tell me SN used to be a commercial product. Was the principle the same too? If you buy it and it does not work the way you expect, then implement it yourself, pay us more to implement it and asked somebody to do it for you :-) It sounds interesting though. Best regards, Paul. From mdejong@cygnus.com Wed Oct 4 17:04:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 04 Oct 2000 17:04:00 -0000 Subject: SN Printing Under WIN/NT References: <002101c02e5e$2b5750c0$60a65a8f@paulselo> Message-ID: On Thu, 5 Oct 2000, Paul Selormey wrote: > Hello Mo DeJong, > > > > [...] > > Humm, if all of these companies are having problems why have they > > not contacted us about a support/development contract to improve > > the C++ parser support? > > Please how much will it cost to contract you. I think many here will help to > contribute that amount which will give us a good parser for C/C++ in parser > and Java where possible. The cost naturally depends on what you want fixed and how long it takes to fix the problem. When you give away free software, you need to charge for support and services to make ends meet. > > Our current priority is Source-Navigator 5.0. If your company needs > > some particular feature, you have the following options: > > We have heard much about this, but it is not building on the current > existing codes? Source-Navigator 5.0 includes a large Itcl upgrade (1.5 -> 3.0) but it is not a "start from scratch" rewrite like mozilla or anything. > What is new here? since no one seems to know what is going on except the > RedHat > staff. We have purposely tried to add no new features for 5.0. We are upgrading the core technologies we depend on (Tcl/Tk/Itcl), but we are not adding a bunch of new user visible features. I am not sure what you mean by "no one seems to know what is going on", we have explained this on the list a number of times. > > 1. Implement it yourself > > 2. Contract with us to implement it > > 3. Hire someone else to do it > > Now, tell me SN used to be a commercial product. Was the principle the same > too? > If you buy it and it does not work the way you expect, then implement it > yourself, pay > us more to implement it and asked somebody to do it for you :-) With a closed source product, you do not have the option to fix it yourself or hire someone else to do it. You are at the mercy of the vendor. This is one of the critical diffs of an open vs closed code. We are not saying that you have to fix the code yourself, but you do have the freedom to do so. Would you buy a car that could only be serviced at the dealer? That said, we think people will recognize the logic of contracting with us to fix problems or add new features, since we already know the code base. cheers Mo DeJong Red Hat Inc From dave.banham@tde.alstom.com Thu Oct 5 00:52:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Thu, 05 Oct 2000 00:52:00 -0000 Subject: Source navigator roadmap, features list, etc.? Message-ID: >I think that we should have a checkbox on the directory selection dialog >that says "Live directory". If this is checked, the directory will be >scanned every time the project is refreshed. You'd have to implement explicit exclusions too. Regards Dave Banham From wgacquer@ubisoft.fr Thu Oct 5 01:16:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Thu, 05 Oct 2000 01:16:00 -0000 Subject: Source navigator roadmap, features list, etc.? Message-ID: <5D2A198567EED311BD0F009027A25C874334C5@SRVMAIL-ANN> or some kind of "extended real file view": dir | file | SN | VC | ------------------------------------------| a | toto.c | yes | yes | a/b | foo.h | no | yes | z/zz | albert.txt | no | no | The SN and VC (version control) columns could be filled with checkboxes. Of course, this one could be more complete than the one above. Maybe we could discuss the features that should be included in that kind of view, if you accept it, of course. William -----Original Message----- From: dave.banham@tde.alstom.com [ mailto:dave.banham@tde.alstom.com ] Sent: jeudi 5 octobre 2000 09:10 To: sourcenav@sources.redhat.com Subject: RE: Source navigator roadmap, features list, etc.? >I think that we should have a checkbox on the directory selection dialog >that says "Live directory". If this is checked, the directory will be >scanned every time the project is refreshed. You'd have to implement explicit exclusions too. Regards Dave Banham From MChalla@USURF.Com Thu Oct 5 06:23:00 2000 From: MChalla@USURF.Com (Matt Challacombe) Date: Thu, 05 Oct 2000 06:23:00 -0000 Subject: build and fortran Message-ID: <00100507185901.01679@MatCha> Hi, I'm trying to get SNavigator to work with fortran90 (I have *.F90s and they get ccp'ed to *.f90s then compiled) and I'm having a few problems I'm hoping the list can help me with. I've grabbed the SN4.5.2 tarball, and when I ./configure I get the following: Configuring for a i686-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. How to avoid this? I see an option for suffixes etc with --help, so this is promising. I also tried the binaries and Ian's advice: "Two ways to add a rule for fortran, one is using sn_add_simple_ide_rule in .../etc/sn_prop.cfg sn_add_simple_ide_rule Fortran \ -suffix-list {.o.f90} \ -action "$<" \ -tool fortran \ -description "Compile a Fortran file." \ -file-type Fortran replace "-tool fortran" with whatever fortran compiler you are using. " I did this, with the approporiate suffixes compiler etc, but SNavigator does not recognice my *.F90 files. Is their something more like a reboot? Thanks, Matt From bje@redhat.com Thu Oct 5 06:28:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 05 Oct 2000 06:28:00 -0000 Subject: build and fortran References: <00100507185901.01679@MatCha> Message-ID: Hi, I've grabbed the SN4.5.2 tarball, and when I ./configure I get the following: Configuring for a i686-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. You need to build in a separate tree to the source tree: $ mkdir build && cd build $ ../sn-src/configure [options] "Two ways to add a rule for fortran, one is using sn_add_simple_ide_rule in .../etc/sn_prop.cfg sn_add_simple_ide_rule Fortran \ -suffix-list {.o.f90} \ -action "$<" \ -tool fortran \ -description "Compile a Fortran file." \ -file-type Fortran replace "-tool fortran" with whatever fortran compiler you are using. " This is for building, not parsing. Look in S-N's project preferences for the extensions that it recognises as Fortran files (and not that the Fortran parser doesn't handle F90 yet). I did this, with the approporiate suffixes compiler etc, but SNavigator does not recognice my *.F90 files. Is their something more like a reboot? I should hope not! Hope this helps, Ben From dvenkita@ecel.uwa.edu.au Thu Oct 5 09:34:00 2000 From: dvenkita@ecel.uwa.edu.au (Daniel V.) Date: Thu, 05 Oct 2000 09:34:00 -0000 Subject: vim integration Message-ID: <200010051628.AAA15195@rand.lab.ecel.uwa.edu.au> Hi. I'm just beginning to use sourcenav, and was wondering if anyone has written some tcl code (rc.tcl) to properly integrate with vim. More specifically, I would like sourcenav to jump to the line referenced (eg when I find a function declaration) instead of just opening the file, like the Emacs integration works now. I got the impression from the docs that implementing some tcl code to run when an edit view is opened, that invokes vim with the + line argument would probably do it, but I have no experience with tcl. If someone has already implemented this, I would be very grateful for the code. The other small question I had was, is there a way to change the cursor colour in the integrated editor? (Because I use a black background, which means I can't see the cursor for quick changes) Thanks, Daniel. From spolk@redhat.com Thu Oct 5 12:06:00 2000 From: spolk@redhat.com (Syd Polk) Date: Thu, 05 Oct 2000 12:06:00 -0000 Subject: SN Printing Under WIN/NT References: <5D2A198567EED311BD0F009027A25C87402387@SRVMAIL-ANN> <4.2.0.58.20001004102624.00dad380@pop.cygnus.com> <005501c02e41$70241a20$7113a8c0@concord.com> Message-ID: <4.2.0.58.20001005120625.00db3b10@pop.cygnus.com> At 04:26 PM 10/4/00 -0400, Berek wrote: >Lot's of problems printing hierarchy and cross reference charts under NT. I >have SN installed on 1/2 dozen PC's. Some use Postscript drivers, others >PCL, some use both. All have same results printing on 1/2 dozen networked >printers. Basically, printing is garbled: letters, symbols, and lines >overlaying each other. Have to resort to doing NT alt-print screen and >pasting into MS Word, then printing from Word. The real problem is that there are no font preferences that just apply to printing, and the fonts for screen and printing don't match very well. We need to do some work at some point cleaning this up, as well as an overall preferences rewrite. >In Class view, C++ class members that are "private" have symbol (transparent >diamond) for "public". Private members should have no symbol at all. Both >public and private have the public symbol. > >I'm concerned about comments made by SN Development (Syd Polk?) about >priority given to Perl parser. I know of more than 1/2 dozen companies in my >area that have been forced to scale back use of SN because of many problems >with C++ parser: 3Com, Lucent, Cienna, Concord Communications, Broadband >Access Systems, Cisco, Fidelity... Unless C++ parsing bugs are fixed, SN >cannot be relied upon as a development tool. We desparately need accurate >cross references. Those of us that us SN are forced to fall back on grep's. >Please review your priorities. C++ parser is a high priority; we use it internally. Thanks for your feedback; it really is appreciated. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From berek@usa.net Thu Oct 5 13:07:00 2000 From: berek@usa.net (Berek) Date: Thu, 05 Oct 2000 13:07:00 -0000 Subject: SN Printing Under WIN/NT References: Message-ID: <006901c02f07$db6acce0$7113a8c0@concord.com> See imbedded comments below. ----- Original Message ----- From: "Mo DeJong" To: "Berek" Sent: Thursday, October 05, 2000 14:19 Subject: Re: SN Printing Under WIN/NT > > > Humm, if all of these companies are having problems why have they > > > not contacted us about a support/development contract to improve > > > the C++ parser support? > > > > Because they bought SN for evaluation: to see if it could improve the > > development and sustaining engineering process. SN works well for standard > > C, but it is simply not usable for C++. Since the folks that evaluated SN > > (at these companies) are experienced s/w engineers, they know that SN is not > > a release quality product (Beta at best) for C++ code analysis, and they > > have a good feeling for the effort and time it will take before these > > problems are resolved. When they bought the product, they assumed they > > wouldn't have to shell out additional money to Red Hat (Cygnus at the time) > > to fix bugs that shouldn't have existed (in release quality software) in the > > first place. > > > > I've been a s/w engineer for thiry-four years and I have never seen a > > release quality product with so many serious and blatant defects. > > You must be joking. I have seen plenty of "released" software products > that a much much worse than SN. While I would agree with you that > the parsers need some serious reworking, saying SN in the worst > product in 34 years is really a stretch. At least it does not > core dump :) Sorry...I stand by my statement. We're not talking about frivilous or insignificant bugs, we're talking about basic functionality: SN does not work as advertized for C++. Nor can it print xref charts under Windows NT for, apparently, both C and C++. And...it does core dump, at least dbimp croaks on an address violation when parsing large projects. Yes, I did report this problem. No, it's not a parsing error, it's a boundary/threshhold problem. The large project I refer to contains over 560,000 lines of code. It's comprised of much smaller projects, each of which parses correctly when done separately (dbimp doesn't croak). When I create a new project consisting of two or more of these smaller projects, dbimp "goes south". > > > It raises > > concern about the level and degree of testing applied to this product. 99% > > of these (C++ and printing) problems should have been caught and fixed > > before SN ever hit the streets. I apologize for the strong language, but > > when I relate to Red Hat engineering the problems we've encountered with > > their s/w and the response we get is "pay us to fix 'em" or "do it > > yourself", it sends my blood pressure through the roof. If any of the > > companies I've worked for had that sort of attitude, they'd have been out of > > business long ago. My $0.02 worth. > > I hear what you are saying. This issue does need to get solved, > and we are looking for help to get the right solution in place. > The trick here is that our business is customer driven. We are > going to have a hard time justifying 3 to 4 months of work on > the C++ parser if we do not have a customer that is paying for it. You mean to tell me that there aren't enough customers in the United States using C++ to justify fixing very basic, fundamental flaws in the C++ parser? (Not to mention fixing bugs with printing xref and hierarchy charts.) > > If you like, I can give you the names and phone numbers of people that have > > evaluated SN at Concord Communications, Broadband, and 3Com. I can also tell > > you exactly what they're going to say: "we can't use SN for C++ development > > until these bugs are fixed". > > > > Cheers > > Steve Dow > > So we are stuck at a catch-22. We need to figure out how to > more forward, how about asking these folks "if the C++ > support in SN was fixed, would you consider purchasing > support and using the product?" Have you done a market survey yet? I find it hard to believe that managememt at Red Hat thinks the PERL market is larger than C++. My feeling is that any survey is going to show (in decreasing order of use): COBOL, std C, C++, all others. > If the answer if "yes", and we can get that in a nice > pie chart to present to management, then it would > be a lot easier to get resources allocated to doing it. I just happen to have such a chart in my office. I'd be delighted to forward it to you [smile]. > cheers > Mo DeJong > Red Hat Inc > From berek@usa.net Thu Oct 5 13:13:00 2000 From: berek@usa.net (Berek) Date: Thu, 05 Oct 2000 13:13:00 -0000 Subject: SN Printing Under WIN/NT References: <5D2A198567EED311BD0F009027A25C87402387@SRVMAIL-ANN> <4.2.0.58.20001004102624.00dad380@pop.cygnus.com> <4.2.0.58.20001005120625.00db3b10@pop.cygnus.com> Message-ID: <007001c02f08$b755c7f0$7113a8c0@concord.com> Thank you [smile]. I really love SN. If you guys manage to clean up the C++ parser so that it's as reliable as a compiler (the closer to 100%, the better off we'll be...we absolutely have to have accurate xrefs), this product will be a developer's "dream come true". ----- Original Message ----- From: "Syd Polk" To: "Berek" ; Sent: Thursday, October 05, 2000 15:08 Subject: Re: SN Printing Under WIN/NT > At 04:26 PM 10/4/00 -0400, Berek wrote: > >Lot's of problems printing hierarchy and cross reference charts under NT. I > >have SN installed on 1/2 dozen PC's. Some use Postscript drivers, others > >PCL, some use both. All have same results printing on 1/2 dozen networked > >printers. Basically, printing is garbled: letters, symbols, and lines > >overlaying each other. Have to resort to doing NT alt-print screen and > >pasting into MS Word, then printing from Word. > > The real problem is that there are no font preferences that just apply to > printing, and the fonts for screen and printing don't match very well. We > need to do some work at some point cleaning this up, as well as an overall > preferences rewrite. > > > >In Class view, C++ class members that are "private" have symbol (transparent > >diamond) for "public". Private members should have no symbol at all. Both > >public and private have the public symbol. > > > >I'm concerned about comments made by SN Development (Syd Polk?) about > >priority given to Perl parser. I know of more than 1/2 dozen companies in my > >area that have been forced to scale back use of SN because of many problems > >with C++ parser: 3Com, Lucent, Cienna, Concord Communications, Broadband > >Access Systems, Cisco, Fidelity... Unless C++ parsing bugs are fixed, SN > >cannot be relied upon as a development tool. We desparately need accurate > >cross references. Those of us that us SN are forced to fall back on grep's. > >Please review your priorities. > > C++ parser is a high priority; we use it internally. > > Thanks for your feedback; it really is appreciated. > > Syd Polk spolk@redhat.com > Engineering Manager +1 415 777 9810 x 241 > Red Hat, Inc. > > > > > From mdejong@cygnus.com Thu Oct 5 13:28:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Thu, 05 Oct 2000 13:28:00 -0000 Subject: SN Printing Under WIN/NT References: <006901c02f07$db6acce0$7113a8c0@concord.com> Message-ID: On Thu, 5 Oct 2000, Berek wrote: > > You must be joking. I have seen plenty of "released" software products > > that a much much worse than SN. While I would agree with you that > > the parsers need some serious reworking, saying SN in the worst > > product in 34 years is really a stretch. At least it does not > > core dump :) > > Sorry...I stand by my statement. We're not talking about frivilous or > insignificant bugs, we're talking about basic functionality: SN does not > work as advertized for C++. Nor can it print xref charts under Windows NT > for, apparently, both C and C++. > > And...it does core dump, at least dbimp croaks on an address violation when > parsing large projects. Perhaps you missed the smiley face on that sentence. > Yes, I did report this problem. No, it's not a > parsing error, it's a boundary/threshhold problem. The large project I refer > to contains over 560,000 lines of code. It's comprised of much smaller > projects, each of which parses correctly when done separately (dbimp doesn't > croak). When I create a new project consisting of two or more of these > smaller projects, dbimp "goes south". Without a way to reproduce the error, it is unlikely we are going to be able to fix it. > > I hear what you are saying. This issue does need to get solved, > > and we are looking for help to get the right solution in place. > > The trick here is that our business is customer driven. We are > > going to have a hard time justifying 3 to 4 months of work on > > the C++ parser if we do not have a customer that is paying for it. > > You mean to tell me that there aren't enough customers in the United States > using C++ to justify fixing very basic, fundamental flaws in the C++ parser? > (Not to mention fixing bugs with printing xref and hierarchy charts.) Perhaps you misunderstand the meaning of "customer" in this context. We currently have actual customers that are paying us to work a number of projects that include Source-Navigator. This set of folks is much smaller than the sum of all the C++ users in the U.S. > Have you done a market survey yet? I find it hard to believe that managememt > at Red Hat thinks the PERL market is larger than C++. My feeling is that any > survey is going to show (in decreasing order of use): COBOL, std C, C++, all > others. I am not sure why you are so worked up about Perl. A Perl parser was talked about in the context of things that people on the net might want to donate. If you are so interested in C++ support, why don't you champion the cause of writing a new C++ parser for Source-Navigator? It sounds as though you are uninterested in allocating any capital resources to a new parser, so how about organizing others to help you achieve your goal? cheers Mo DeJong Red Hat Inc From berek@usa.net Thu Oct 5 13:49:00 2000 From: berek@usa.net (Berek) Date: Thu, 05 Oct 2000 13:49:00 -0000 Subject: SN Printing Under WIN/NT References: Message-ID: <008501c02f0d$af522350$7113a8c0@concord.com> See imbedded comments below [smile]. ----- Original Message ----- From: "Mo DeJong" To: Sent: Thursday, October 05, 2000 16:28 Subject: Re: SN Printing Under WIN/NT > On Thu, 5 Oct 2000, Berek wrote: > > > > You must be joking. I have seen plenty of "released" software products > > > that a much much worse than SN. While I would agree with you that > > > the parsers need some serious reworking, saying SN in the worst > > > product in 34 years is really a stretch. At least it does not > > > core dump :) > > > > Sorry...I stand by my statement. We're not talking about frivilous or > > insignificant bugs, we're talking about basic functionality: SN does not > > work as advertized for C++. Nor can it print xref charts under Windows NT > > for, apparently, both C and C++. > > > > And...it does core dump, at least dbimp croaks on an address violation when > > parsing large projects. > > Perhaps you missed the smiley face on that sentence. > > > Yes, I did report this problem. No, it's not a > > parsing error, it's a boundary/threshhold problem. The large project I refer > > to contains over 560,000 lines of code. It's comprised of much smaller > > projects, each of which parses correctly when done separately (dbimp doesn't > > croak). When I create a new project consisting of two or more of these > > smaller projects, dbimp "goes south". > > Without a way to reproduce the error, it is unlikely we are going > to be able to fix it. Right. I wish I could send you my code base, but Concord Comm. would have some serious objections [smile]. > > > I hear what you are saying. This issue does need to get solved, > > > and we are looking for help to get the right solution in place. > > > The trick here is that our business is customer driven. We are > > > going to have a hard time justifying 3 to 4 months of work on > > > the C++ parser if we do not have a customer that is paying for it. > > > > You mean to tell me that there aren't enough customers in the United States > > using C++ to justify fixing very basic, fundamental flaws in the C++ parser? > > (Not to mention fixing bugs with printing xref and hierarchy charts.) > > Perhaps you misunderstand the meaning of "customer" in this context. > We currently have actual customers that are paying us to work > a number of projects that include Source-Navigator. This set > of folks is much smaller than the sum of all the C++ users > in the U.S. I understand. But, if the C++ parser were cleaned up, you'd have a lot more C++ customers and a lot more revenue than you do now. > > > Have you done a market survey yet? I find it hard to believe that managememt > > at Red Hat thinks the PERL market is larger than C++. My feeling is that any > > survey is going to show (in decreasing order of use): COBOL, std C, C++, all > > others. > > I am not sure why you are so worked up about Perl. A Perl parser > was talked about in the context of things that people on the net > might want to donate. If you are so interested in C++ support, > why don't you champion the cause of writing a new C++ parser > for Source-Navigator? It sounds as though you are uninterested > in allocating any capital resources to a new parser, so how > about organizing others to help you achieve your goal? There you go again with the "if you don't like the situation, fix it yourself" attitude [smile]. Let me put it a little more bluntly: I work 60 plus hours/week at my job at Concord. The weekends (when I have both Sat and Sun off) I devote to chores around the house/yard and spending a few minutes here and there with my wife and kids (once in a great while I even manage to sneak in a hug or two). I spend my "spare time" coming up to speed in such things as Java, web page development, and papers on protocols and networking (work-related). I have to do this to stay current and remain competitive. Now, that having been said, how much time do you think I can devote to fixing SN bugs [smile]? This is not my job, it's your job. BTW, my four Golden Retrievers and two cats don't seem to know me anymore. I can't remember the last time I was able to take the dogs for a walk. I have a proposition for you. You dedicate some of your free time to my wife, kids, dogs and cats, and I'll use the time saved to fix SN bugs. Deal [grin]? Later -Steve Dow > cheers > Mo DeJong > Red Hat Inc > From bje@redhat.com Thu Oct 5 14:06:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 05 Oct 2000 14:06:00 -0000 Subject: SN Printing Under WIN/NT References: <006901c02f07$db6acce0$7113a8c0@concord.com> Message-ID: And...it does core dump, at least dbimp croaks on an address violation when parsing large projects. Yes, I did report this problem. No, it's not a parsing error, it's a boundary/threshhold problem. The large project I refer to contains over 560,000 lines of code. It's comprised of much smaller projects, each of which parses correctly when done separately (dbimp doesn't croak). When I create a new project consisting of two or more of these smaller projects, dbimp "goes south". It probably wouldn't take much to track this bug down. It's possible to run the parser manually, capture the output and then feed it into dbimp by hand (under a debugger). I assume you're not willing to share your 560,000 lines of code? Ben From erayo@cs.bilkent.edu.tr Sun Oct 8 09:37:00 2000 From: erayo@cs.bilkent.edu.tr (Eray 'exa' Ozkural) Date: Sun, 08 Oct 2000 09:37:00 -0000 Subject: Sourcenav for debian Message-ID: <20001008194348.A400@borg.cs.bilkent.edu.tr> Hi, I'm building a debian package for sourcenav. I've managed to configure and compile sourcenav for debian, making small changes where necessary but the "make install" seems to get stuck in a loop: The source is configured and built in a parallel directory. Then, cd $(builddir)/snavigator; \ $(MAKE) install prefix=$(installdir) DESTDIR=$(debdir)/tmp \ mandir=$(installdir)/share/man \ infodir=$(installdir)/share/info \ SN_SDK_EXEC_PREFIX=$(installdir) where srcdir=$(shell pwd) debdir=$(srcdir)/debian builddir=$(debdir)/build installdir=$(debdir)/tmp/usr The install-recursive target in snavigator/Makefile.in seems to cause the infinite loop. What do you think? Thanks, __ Eray From irox@redhat.com Mon Oct 9 13:38:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Mon, 09 Oct 2000 13:38:00 -0000 Subject: C++ parser problem (RFC) Message-ID: <200010092035.NAA25638@manx.soma.redhat.com> Hi, Can all the people who are having problems please try to send in test cases (easy, minimal, code snippets) with details instructions. If you can't get a really good reproducible test case try to post an outline, somebody might be have similar problems and beable to produce a good test case. If you want your bugs fixed and (for whatever reason) can't fix it yourself, remember the better a bug report you give us the easier it is for us to fix. Thanks, Ian. P.S.: We do want to fix all the bugs in SN regardless of who's paying us. P.S.S.: Please keep posts on topic. From pallas@CS.Stanford.EDU Mon Oct 9 14:09:00 2000 From: pallas@CS.Stanford.EDU (Joseph Pallas) Date: Mon, 09 Oct 2000 14:09:00 -0000 Subject: A small fix Message-ID: I was never able to produce a small test-case for my crash, but I was able to analyze it enough to convince myself that I understood the problem and to hazard a fix. I've been running with the following patch for some time and I haven't seen any problems. Most important, it cured my SEGV crashes in dbimp. All it does is specify that copies need to be made of macro names when they're inserted in the search table, instead of using the ephemeral string that was passed in. I'm convinced it's harmless and fixes at least one bug. joe Index: snavigator/parsers/cpp/cpplib/macro.c =================================================================== RCS file: /home/pallas/cvsroot/source_navigator/snavigator/parsers/cpp/cpplib/macro.c,v retrieving revision 1.1.1.1 diff -c -r1.1.1.1 macro.c *** macro.c 2000/08/14 20:58:28 1.1.1.1 --- macro.c 2000/09/01 17:33:41 *************** *** 90,96 **** entry.key_len = sString.leng; entry.data = 0; entry.data_len = 0; ! entry.flag = 0; if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) { --- 90,96 ---- entry.key_len = sString.leng; entry.data = 0; entry.data_len = 0; ! entry.flag = SEARCH_DUP_KEY; if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) { From irox@redhat.com Mon Oct 9 14:11:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Mon, 09 Oct 2000 14:11:00 -0000 Subject: build and fortran Message-ID: <200010092108.OAA25688@manx.soma.redhat.com> On Thu, 5 Oct 2000, Matt Challacombe wrote: > I also tried the binaries and Ian's advice: > > > "Two ways to add a rule for fortran, one is using > sn_add_simple_ide_rule in .../etc/sn_prop.cfg > > sn_add_simple_ide_rule Fortran \ > -suffix-list {.o.f90} \ > -action "$<" \ > -tool fortran \ > -description "Compile a Fortran file." \ > -file-type Fortran > > replace "-tool fortran" with whatever fortran compiler you are using. " > > I did this, with the approporiate suffixes compiler etc, but > SNavigator does not recognice my *.F90 files. Is their something more > like a reboot? What platform are you using? Note that you say your files have a .F90 extension and the build rule is looking for .f90 files. This might cause the problem you are seeing. Simply change the -suffix-list arguement to {.o.F90} so that it will pick up your .F90 files. Also look for: sn_add_parser fortran \ -suffix {*.f *.for *.FOR} \ -brow_cmd $sn_path(bindir)/fbrowser \ -high_cmd $sn_path(bindir)/fbrowser \ -brow_switch "" \ -high_switch "-h" \ -case 1 You can add .F90 to the -suffix arguement list so they are add to the SN project, but as BJE says, the fortran parsers don't support F90. Assume the fortran parser chokes on the F90 files you can add them under "others" (files without parsers). Look for: sn_add_parser others -suffix $other_ext and add *.F90 to the $other_ext variable set just above that line. (This is all inside the .../etc/sn_prop.cfg file). Ian. From spolk@redhat.com Mon Oct 9 14:42:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 09 Oct 2000 14:42:00 -0000 Subject: A small fix References: Message-ID: <4.2.0.58.20001009144355.00d74ae0@pop.cygnus.com> If you write a small ChangeLog entry, I will generate a patch. At 02:08 PM 10/9/00 -0700, Joseph Pallas wrote: >I was never able to produce a small test-case for my crash, but I was able >to analyze it enough to convince myself that I understood the problem and >to hazard a fix. I've been running with the following patch for some time >and I haven't seen any problems. Most important, it cured my SEGV crashes >in dbimp. All it does is specify that copies need to be made of macro >names when they're inserted in the search table, instead of using the >ephemeral string that was passed in. I'm convinced it's harmless and >fixes at least one bug. > >joe > >Index: snavigator/parsers/cpp/cpplib/macro.c >=================================================================== >RCS file: >/home/pallas/cvsroot/source_navigator/snavigator/parsers/cpp/cpplib/macro.c,v >retrieving revision 1.1.1.1 >diff -c -r1.1.1.1 macro.c >*** macro.c 2000/08/14 20:58:28 1.1.1.1 >--- macro.c 2000/09/01 17:33:41 >*************** >*** 90,96 **** > entry.key_len = sString.leng; > entry.data = 0; > entry.data_len = 0; >! entry.flag = 0; > > if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) > { >--- 90,96 ---- > entry.key_len = sString.leng; > entry.data = 0; > entry.data_len = 0; >! entry.flag = SEARCH_DUP_KEY; > > if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) > { Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mdejong@cygnus.com Mon Oct 9 15:53:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 09 Oct 2000 15:53:00 -0000 Subject: Sourcenav for debian References: <20001008194348.A400@borg.cs.bilkent.edu.tr> Message-ID: On Sun, 8 Oct 2000, Eray 'exa' Ozkural wrote: > Hi, > > I'm building a debian package for sourcenav. Great! We really appreciate you help getting SN packaged up for debian. You are going to run into the old Tcl/Tk version problems, but I am sure you already know about that. SN is going to be easier to get working with Tcl/Tk 8.3.2 (when compared to insight), but it is work that still needs to be done and tested on all the platforms we support. > I've managed to > configure and compile sourcenav for debian, making small changes > where necessary but the "make install" seems to get stuck in a > loop: > > The source is configured and built in a parallel directory. > Then, > > cd $(builddir)/snavigator; \ > $(MAKE) install prefix=$(installdir) DESTDIR=$(debdir)/tmp \ > mandir=$(installdir)/share/man \ > infodir=$(installdir)/share/info \ > SN_SDK_EXEC_PREFIX=$(installdir) Could this be a problem related to passing prefix=$(installdir) to the make install goal? I doubt we have tested that. We know it works when you run ./configure with a --prefix argument, but that is the only way it is currently "supported". cheers Mo DeJong Red Hat Inc From vanpouck@imec.be Tue Oct 10 06:54:00 2000 From: vanpouck@imec.be (vanpouck) Date: Tue, 10 Oct 2000 06:54:00 -0000 Subject: build settings dissapear Message-ID: <39E31FE5.BDD3D89F@imec.be> Hello, I' m trying to change the linker of my C++ files from g++ to ld. I change the linker prgram in the build settings, but when I start a build these settings have dissapeared. Is this a known bug ? are there any solutions ? I have revission 4.5.2 Greetings, Bart Please put me in CC when anyone replys. From mharris@opensourceadvocate.org Tue Oct 10 11:06:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 11:06:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) References: Message-ID: On Mon, 9 Oct 2000, Mo DeJong wrote: >> I'm building a debian package for sourcenav. > >Great! We really appreciate you help getting SN >packaged up for debian. You are going to run into >the old Tcl/Tk version problems, but I am sure you >already know about that. SN is going to be easier >to get working with Tcl/Tk 8.3.2 >(when compared to insight), but it is work >that still needs to be done and tested on all >the platforms we support. I'm working on SN packages for Red Hat as well. Any suggestions/advice you may have would be very much appreciated. ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Microsoft Windows(tm). A thirty-two bit extension and graphical shell to a sixteen bit patch to an eight bit operating system originally coded for a four bit microprocessor which was written by a two-bit company that can't stand one bit of competition. From mdejong@cygnus.com Tue Oct 10 11:25:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 10 Oct 2000 11:25:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) References: Message-ID: On Tue, 10 Oct 2000, Mike A. Harris wrote: > On Mon, 9 Oct 2000, Mo DeJong wrote: > > >> I'm building a debian package for sourcenav. > > > >Great! We really appreciate you help getting SN > >packaged up for debian. You are going to run into > >the old Tcl/Tk version problems, but I am sure you > >already know about that. SN is going to be easier > >to get working with Tcl/Tk 8.3.2 > >(when compared to insight), but it is work > >that still needs to be done and tested on all > >the platforms we support. > > I'm working on SN packages for Red Hat as well. Any > suggestions/advice you may have would be very much appreciated. My advice would be, don't bother. We are planning on making RPMs for SN once 5.0 is out. This is going to be a long process because we also need to update our Tcl and Tk installs. When finished, SN will use Tcl/Tk 8.3.2, but that is still some months off. Mo DeJong Red Hat Inc From mharris@opensourceadvocate.org Tue Oct 10 11:40:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 11:40:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) References: Message-ID: On Tue, 10 Oct 2000, Mo DeJong wrote: >> >Great! We really appreciate you help getting SN >> >packaged up for debian. You are going to run into >> >the old Tcl/Tk version problems, but I am sure you >> >already know about that. SN is going to be easier >> >to get working with Tcl/Tk 8.3.2 >> >(when compared to insight), but it is work >> >that still needs to be done and tested on all >> >the platforms we support. >> >> I'm working on SN packages for Red Hat as well. Any >> suggestions/advice you may have would be very much appreciated. > >My advice would be, don't bother. We are planning on making >RPMs for SN once 5.0 is out. This is going to be a long >process because we also need to update our Tcl and Tk >installs. When finished, SN will use Tcl/Tk 8.3.2, >but that is still some months off. Hmm.. I'm going to take a stab at it anyway just to see what problems arise as I hate mixing RPM's and tgz installs on a Red Hat system. If it is too messy, I might just drop the idea after, but I want to have a shot at it anyways. I'll post my results here. Thanks for the warning. ;o) ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Need general help or technical support with Red Hat Linux 6.2? Join the user support mailing list by sending a message to "zoot-list-request@redhat.com" with the word "subscribe" on the subject line. From spolk@redhat.com Tue Oct 10 11:59:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 10 Oct 2000 11:59:00 -0000 Subject: A small fix References: Message-ID: <4.2.0.58.20001010115914.00dd2730@pop.cygnus.com> Patch has been posted at ftp://sources.redhat.com/pub/sourcenav/releases/patches/0008-pallas-001009-p atch. At 02:08 PM 10/9/00 -0700, Joseph Pallas wrote: >I was never able to produce a small test-case for my crash, but I was able >to analyze it enough to convince myself that I understood the problem and >to hazard a fix. I've been running with the following patch for some time >and I haven't seen any problems. Most important, it cured my SEGV crashes >in dbimp. All it does is specify that copies need to be made of macro >names when they're inserted in the search table, instead of using the >ephemeral string that was passed in. I'm convinced it's harmless and >fixes at least one bug. > >joe > >Index: snavigator/parsers/cpp/cpplib/macro.c >=================================================================== >RCS file: >/home/pallas/cvsroot/source_navigator/snavigator/parsers/cpp/cpplib/macro.c,v >retrieving revision 1.1.1.1 >diff -c -r1.1.1.1 macro.c >*** macro.c 2000/08/14 20:58:28 1.1.1.1 >--- macro.c 2000/09/01 17:33:41 >*************** >*** 90,96 **** > entry.key_len = sString.leng; > entry.data = 0; > entry.data_len = 0; >! entry.flag = 0; > > if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) > { >--- 90,96 ---- > entry.key_len = sString.leng; > entry.data = 0; > entry.data_len = 0; >! entry.flag = SEARCH_DUP_KEY; > > if(( pentry = (*pbtrMacro->search)( &pbtrMacro, entry ))) > { Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From wgacquer@ubisoft.fr Tue Oct 10 12:04:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Tue, 10 Oct 2000 12:04:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) Message-ID: <5D2A198567EED311BD0F009027A25C87460EBC@SRVMAIL-ANN> I would suggest following Mo's advice. If you don't want SN to interfer with the other RPMS, just get the sources and use the "--use-prefix" option in the configure step. For instance, I installed SN on tens of linux boxes using the "configure --use-prefix=/usr/local/snavigator" and it gave me less headache than playing with conflicts from the sn modified tcl/tk or grep. In fact, on most of the computers here, /usr/local/snavigator is a NFS symbolic link. The only thing that one has on his PATH is a shell script that launches SN. I'll wait for SN 5.0 RPMS. When is is scheduled? William -----Original Message----- From: Mike A. Harris [ mailto:mharris@opensourceadvocate.org ] Sent: mardi 10 octobre 2000 20:40 To: Mo DeJong Cc: sourcenav@sourceware.cygnus.com Subject: Re: SN for Red Hat (was Re: Sourcenav for debian) On Tue, 10 Oct 2000, Mo DeJong wrote: >> >Great! We really appreciate you help getting SN >> >packaged up for debian. You are going to run into >> >the old Tcl/Tk version problems, but I am sure you >> >already know about that. SN is going to be easier >> >to get working with Tcl/Tk 8.3.2 >> >(when compared to insight), but it is work >> >that still needs to be done and tested on all >> >the platforms we support. >> >> I'm working on SN packages for Red Hat as well. Any >> suggestions/advice you may have would be very much appreciated. > >My advice would be, don't bother. We are planning on making >RPMs for SN once 5.0 is out. This is going to be a long >process because we also need to update our Tcl and Tk >installs. When finished, SN will use Tcl/Tk 8.3.2, >but that is still some months off. Hmm.. I'm going to take a stab at it anyway just to see what problems arise as I hate mixing RPM's and tgz installs on a Red Hat system. If it is too messy, I might just drop the idea after, but I want to have a shot at it anyways. I'll post my results here. Thanks for the warning. ;o) ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Need general help or technical support with Red Hat Linux 6.2? Join the user support mailing list by sending a message to "zoot-list-request@redhat.com" with the word "subscribe" on the subject line. From mharris@opensourceadvocate.org Tue Oct 10 12:43:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 12:43:00 -0000 Subject: Can't configure in source dir.. Message-ID: I get the following error after untarring the SN source, cd'ing into the dir and trying to configure: 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure Configuring for a i586-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. No matter what directory I put the sources in I get this. I've got 3 copies of the untarred sources right now and it still does not configure no matter what. Looking at the configure script I find: # Don't allow builds at Cygnus to be done with srcdir==builddir. We # can assume srcdir==. in this case, because of code above. if test "$srcdir" = .; then echo '***' "Cannot configure in source tree." >&2 echo '***' "Make a parallel directory and configure there." >&2 exit 1 fi Short of removing this line from the script, what is the recommended method to build this thing. ;o) I just followed the instructions in the top level README file and it does not work. 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure --prefix=/opt/source-navigator --srcdir=. Configuring for a i586-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. $ ./configure --prefix=/opt/source-navigator \ --srcdir=. --secretdecoderring=~/.decoderrings/sourcenav.ring configure: Unrecognized option: "--secretdecoderring=/home/mharris/.decoderrings/sourcenav.ring"; use --help for usage. Is there some problem with compiling this as non-root? I really don't want to have to be root to build packages.. Any ideas? ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- [Favorite quotes of Linus Torvalds - Sept 6, 2000] I'm a bastard. I have absolutely no clue why people can ever think otherwise. Yet they do. People think I'm a nice guy, and the fact is that I'm a scheming, conniving bastard who doesn't care for any hurt feelings or lost hours of work if it just results in what I consider to be a better system. And I'm not just saying that. I'm really not a very nice person. I can say "I don't care" with a straight face, and really mean it. -- Linus Torvalds on linux-kernel mailing list From irox@redhat.com Tue Oct 10 13:12:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Tue, 10 Oct 2000 13:12:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: <39E3781A.4B1D5F41@redhat.com> Have you tried building out side of the the source tree? E.g. $ mkdir /home/user/sn/src $ mkdir /home/user/sn/build $ cd /home/user/sn/src $ tar -xvzf ...../.tar.gz $ cd /home/user/sn/build $ ../src/configure --prefix=/home/user/sn/install $ make all-snavigator $ make install-snavigator This should help. Ian. "Mike A. Harris" wrote: > > I get the following error after untarring the SN source, cd'ing > into the dir and trying to configure: > > 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure > Configuring for a i586-pc-linux-gnu host. > *** Cannot configure in source tree. > *** Make a parallel directory and configure there. > > No matter what directory I put the sources in I get this. I've > got 3 copies of the untarred sources right now and it still does > not configure no matter what. Looking at the configure script I > find: > > # Don't allow builds at Cygnus to be done with > srcdir==builddir. We > # can assume srcdir==. in this case, because of code above. > if test "$srcdir" = .; then > echo '***' "Cannot configure in source tree." >&2 > echo '***' "Make a parallel directory and configure > there." >&2 > exit 1 > fi > > Short of removing this line from the script, what is the > recommended method to build this thing. ;o) I just followed the > instructions in the top level README file and it does not work. > > 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure > --prefix=/opt/source-navigator --srcdir=. > Configuring for a i586-pc-linux-gnu host. > *** Cannot configure in source tree. > *** Make a parallel directory and configure there. > > $ ./configure --prefix=/opt/source-navigator \ > --srcdir=. --secretdecoderring=~/.decoderrings/sourcenav.ring > configure: Unrecognized > option: "--secretdecoderring=/home/mharris/.decoderrings/sourcenav.ring"; use > --help for usage. > > Is there some problem with compiling this as non-root? I really > don't want to have to be root to build packages.. Any ideas? > > ---------------------------------------------------------------------- > Mike A. Harris - Linux advocate - Open source advocate > Computer Consultant - Capslock Consulting > Copyright 2000 all rights reserved > ---------------------------------------------------------------------- > > [Favorite quotes of Linus Torvalds - Sept 6, 2000] > I'm a bastard. I have absolutely no clue why people can ever think > otherwise. Yet they do. People think I'm a nice guy, and the fact is that > I'm a scheming, conniving bastard who doesn't care for any hurt feelings > or lost hours of work if it just results in what I consider to be a better > system. And I'm not just saying that. I'm really not a very nice person. > I can say "I don't care" with a straight face, and really mean it. > -- Linus Torvalds on linux-kernel mailing list From bje@redhat.com Tue Oct 10 13:15:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 10 Oct 2000 13:15:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: Hi Mike, I get the following error after untarring the SN source, cd'ing into the dir and trying to configure: 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure Configuring for a i586-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. No matter what directory I put the sources in I get this. I've got 3 copies of the untarred sources right now and it still does not configure no matter what. Looking at the configure script I find: It means exactly what the error says: you can't run configure in the source tree. Instead, do this: $ mkdir build && cd build $ ../path/to/src/configure [--switches] Sorry about the confusion. Syd -- this is a good candidate for an FAQ. Cheers, Ben From mdejong@cygnus.com Tue Oct 10 13:16:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 10 Oct 2000 13:16:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: On Tue, 10 Oct 2000, Mike A. Harris wrote: > I get the following error after untarring the SN source, cd'ing > into the dir and trying to configure: > > 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure > Configuring for a i586-pc-linux-gnu host. > *** Cannot configure in source tree. > *** Make a parallel directory and configure there. Just build it in a build dir: mkdir build ; cd build ../SN452/configure --prefix=... make make install That should do it. Mo DeJong Red Hat Inc From bje@redhat.com Tue Oct 10 13:16:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 10 Oct 2000 13:16:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) References: Message-ID: I'm working on SN packages for Red Hat as well. Any suggestions/advice you may have would be very much appreciated. One suggestion I have is to pay close attention to annotating documentation files in the RPM spec. Too many third-party RPMs are generated that don't differentiate documentation files from executables. Running `rpm -qd ' is extremely useful. Thanks for doing this work! Ben From wgacquer@ubisoft.fr Tue Oct 10 13:26:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Tue, 10 Oct 2000 13:26:00 -0000 Subject: Can't configure in source dir.. Message-ID: <5D2A198567EED311BD0F009027A25C87460EEF@SRVMAIL-ANN> cut and paste this ... ---------------------sn_install.sh----------------------- #! /bin/bash echo "Welcome to SourceNavigator" tar xvfz SN452.tgz cd SN452-source mkdir linuxdist cd linuxdist ../configure --prefix=/usr/local/snavigator make make install chmod -R a+r /usr/local/snavigator cd ../.. rm -rf SN452-source echo "SN was installed in /usr/local/snavigator" exit 0 ----------------------sn.sh------------------------------ #! /bin/bash export SN=/usr/local/snavigator export PATH=$SN/bin:$PATH snavigator -----Original Message----- From: Mike A. Harris [ mailto:mharris@opensourceadvocate.org ] Sent: mardi 10 octobre 2000 21:43 To: sourcenav@sourceware.cygnus.com Subject: Can't configure in source dir.. I get the following error after untarring the SN source, cd'ing into the dir and trying to configure: 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure Configuring for a i586-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. No matter what directory I put the sources in I get this. I've got 3 copies of the untarred sources right now and it still does not configure no matter what. Looking at the configure script I find: # Don't allow builds at Cygnus to be done with srcdir==builddir. We # can assume srcdir==. in this case, because of code above. if test "$srcdir" = .; then echo '***' "Cannot configure in source tree." >&2 echo '***' "Make a parallel directory and configure there." >&2 exit 1 fi Short of removing this line from the script, what is the recommended method to build this thing. ;o) I just followed the instructions in the top level README file and it does not work. 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure --prefix=/opt/source-navigator --srcdir=. Configuring for a i586-pc-linux-gnu host. *** Cannot configure in source tree. *** Make a parallel directory and configure there. $ ./configure --prefix=/opt/source-navigator \ --srcdir=. --secretdecoderring=~/.decoderrings/sourcenav.ring configure: Unrecognized option: "--secretdecoderring=/home/mharris/.decoderrings/sourcenav.ring"; use --help for usage. Is there some problem with compiling this as non-root? I really don't want to have to be root to build packages.. Any ideas? ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- [Favorite quotes of Linus Torvalds - Sept 6, 2000] I'm a bastard. I have absolutely no clue why people can ever think otherwise. Yet they do. People think I'm a nice guy, and the fact is that I'm a scheming, conniving bastard who doesn't care for any hurt feelings or lost hours of work if it just results in what I consider to be a better system. And I'm not just saying that. I'm really not a very nice person. I can say "I don't care" with a straight face, and really mean it. -- Linus Torvalds on linux-kernel mailing list From mharris@opensourceadvocate.org Tue Oct 10 13:37:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 13:37:00 -0000 Subject: Can't configure in source dir.. References: <39E3781A.4B1D5F41@redhat.com> Message-ID: On Tue, 10 Oct 2000, Ian Roxborough wrote: >Have you tried building out side of the the source tree? I did try once, yes and it failed with a different error so I figured I did it wrong and reverted to the initial defacto standard way as documented in the README. >E.g. > >$ mkdir /home/user/sn/src >$ mkdir /home/user/sn/build >$ cd /home/user/sn/src >$ tar -xvzf ...../.tar.gz >$ cd /home/user/sn/build >$ ../src/configure --prefix=/home/user/sn/install >$ make all-snavigator >$ make install-snavigator > >This should help. Ouch, my brain hurts. ;o) This is the first time I've ever seen a package require such an oddball build method. ;o) I'll try it though. It might be better to patch the build process than to try to come up with a convoluted RPM build method to work around it. Any idea _why_ it does this? I am almost positive 451 did not require this. Looking at the configure script it appears that it is done for some internal cygnus reason.. Is there a: ./configure --Im-not-cygnus-so-dont-do-that option at all? ;o) Thanks for the suggestion! I'll give it a whirl. TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Red Hat FAQ tip: Having trouble upgrading RPM 3.0.x to RPM 4.0.x? Upgrade first to version 3.0.5, and then to 4.0.x. All packages are available on Red Hat's ftp sites: ftp://ftp.redhat.com ftp://rawhide.redhat.com From mharris@opensourceadvocate.org Tue Oct 10 13:41:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 13:41:00 -0000 Subject: Can't configure in source dir.. References: <5D2A198567EED311BD0F009027A25C87460EEF@SRVMAIL-ANN> Message-ID: On Tue, 10 Oct 2000, William Gacquer wrote: >---------------------sn_install.sh----------------------- >#! /bin/bash >echo "Welcome to SourceNavigator" >tar xvfz SN452.tgz >cd SN452-source >mkdir linuxdist >cd linuxdist >../configure --prefix=/usr/local/snavigator >make >make install >chmod -R a+r /usr/local/snavigator >cd ../.. >rm -rf SN452-source >echo "SN was installed in /usr/local/snavigator" >exit 0 > >----------------------sn.sh------------------------------ >#! /bin/bash >export SN=/usr/local/snavigator >export PATH=$SN/bin:$PATH >snavigator AHHH! Now _that_ makes it much more clear. I think I can easily work with that in RPM. Thanks for the detailed instructions. It seems to be working now.. Perhaps the upper level README should be patched to give more detailed build instructions. If nobody feels like it, I'll volunteer to make the documentation clearer if someone is willing to integrate my effort. Thanks. ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Red Hat FAQ tip: Having trouble upgrading RPM 3.0.x to RPM 4.0.x? Upgrade first to version 3.0.5, and then to 4.0.x. All packages are available on Red Hat's ftp sites: ftp://ftp.redhat.com ftp://rawhide.redhat.com From spolk@redhat.com Tue Oct 10 13:46:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 10 Oct 2000 13:46:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: <4.2.0.58.20001010134750.00dd9320@pop.cygnus.com> At 01:16 PM 10/10/00 -0700, Mo DeJong wrote: >On Tue, 10 Oct 2000, Mike A. Harris wrote: > > > I get the following error after untarring the SN source, cd'ing > > into the dir and trying to configure: > > > > 10 mharris@asdf:~/rpmbuild/SOURCES/SN452-source$ ./configure > > Configuring for a i586-pc-linux-gnu host. > > *** Cannot configure in source tree. > > *** Make a parallel directory and configure there. > >Just build it in a build dir: > >mkdir build ; cd build >../SN452/configure --prefix=... >make >make install > >That should do it. > >Mo DeJong >Red Hat Inc One major wrinkle is the source RPM's are basically set up to build the app in the source tree. It is kind of hard to work around that. We will have to solve this problem when we do our RPMs. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Tue Oct 10 13:51:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 10 Oct 2000 13:51:00 -0000 Subject: Can't configure in source dir.. References: <39E3781A.4B1D5F41@redhat.com> Message-ID: <4.2.0.58.20001010135207.00dd5570@pop.cygnus.com> >Ouch, my brain hurts. ;o) This is the first time I've ever seen >a package require such an oddball build method. ;o) I'll try it >though. It might be better to patch the build process than to >try to come up with a convoluted RPM build method to work around >it. > >Any idea _why_ it does this? I am almost positive 451 did not >require this. Looking at the configure script it appears that it >is done for some internal cygnus reason.. Is there a: > >./configure --Im-not-cygnus-so-dont-do-that Building for multiple platforms for the same source tree is the reason why we build in a separate directory. It looks like I used the wrong version of autoconf to generate the configure file; I will investigate a patch soon. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mharris@opensourceadvocate.org Tue Oct 10 13:52:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 13:52:00 -0000 Subject: SN for Red Hat (was Re: Sourcenav for debian) References: Message-ID: On Wed, 11 Oct 2000, Ben Elliston wrote: > I'm working on SN packages for Red Hat as well. Any > suggestions/advice you may have would be very much appreciated. > >One suggestion I have is to pay close attention to annotating documentation >files in the RPM spec. Too many third-party RPMs are generated that don't >differentiate documentation files from executables. Running `rpm -qd ' >is extremely useful. Yep, not to worry, I try to make my RPMs adhere to as many standards - official, and defacto - as possible, and also make life easy for the end user. Also, I try to maintain compatibility between Red Hat 6.x and the new 7.x although I haven't had a chance to fully test my environment out yet. I also do everything as non-root to make things even more flexible. If I get a working spec, I'll submit it back for possible inclusion to you guys, and you may also feel free to patch it as necessary. >Thanks for doing this work! No problem. The benefit comes from everything learned about RPM during the process. ;o) I've learned a lot about packaging from other things I'm packaging so I thought tackling something a bit more insane might help increase the knowledge of RPM internals a bit. ;o) Take care! ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- "Facts do not cease to exist because they are ignored." - Aldous Huxley From bje@redhat.com Tue Oct 10 13:55:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 10 Oct 2000 13:55:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: I did try once, yes and it failed with a different error so I figured I did it wrong and reverted to the initial defacto standard way as documented in the README. It says this in the README? Where? Ouch, my brain hurts. ;o) This is the first time I've ever seen a package require such an oddball build method. ;o) I'll try it though. It might be better to patch the build process than to try to come up with a convoluted RPM build method to work around it. Any idea _why_ it does this? I am almost positive 451 did not require this. Looking at the configure script it appears that it is done for some internal cygnus reason.. Is there a: It's a Cygnus build convention -- and it's extremely useful in a development environment, where it's undesirable to litter your source tree with object files. Yes, it's a bit unusual, but it's better (IMHO) this way. Ben From mharris@opensourceadvocate.org Tue Oct 10 14:04:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 14:04:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: On Wed, 11 Oct 2000, Ben Elliston wrote: >Date: Wed, 11 Oct 2000 07:55:15 +1100 (EST) >From: Ben Elliston >To: Mike A. Harris >Cc: Ian Roxborough , sourcenav@sourceware.cygnus.com >Content-Type: TEXT/PLAIN; charset=US-ASCII >Subject: Re: Can't configure in source dir.. > > I did try once, yes and it failed with a different error so I figured > I did it wrong and reverted to the initial defacto standard way as > documented in the README. > >It says this in the README? Where? I might not have been clear enough there. What I meant is I tried what the README says, and it does not work. Here is from the SN452 top level readme: ---- It is now possible to automatically configure and build a variety of tools with one command. To build all of the tools contained herein, run the ``configure'' script here, e.g.: ./configure make ---- The instructions you all gave me though seem to work, so things are definitely looking better now. ;o) You guys give fast answers! I went to get a coffee after writing my question and when I came back I had 3 answers! ;o) I like that. ;o) > Ouch, my brain hurts. ;o) This is the first time I've ever seen a > package require such an oddball build method. ;o) I'll try it > though. It might be better to patch the build process than to try to > come up with a convoluted RPM build method to work around it. > > Any idea _why_ it does this? I am almost positive 451 did not require > this. Looking at the configure script it appears that it is done for > some internal cygnus reason.. Is there a: > >It's a Cygnus build convention -- and it's extremely useful in a development >environment, where it's undesirable to litter your source tree with object >files. Yes, it's a bit unusual, but it's better (IMHO) this way. Right, I can see the usefulness now, especially if cross building, however it should be documented more clearly in the README or some other appropriate file. Myself, I just create "clean" and "distclean" targets, or I create the makefile to spew out .o's, and other output files into an output dir of some kind. I've never seen this particular approach before though, so it threw me for a loop. ;o) TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- If you're looking for Linux books, guides, and other documentation, visit the Linux Documentation Project homepage: http://linuxdoc.org From bje@redhat.com Tue Oct 10 14:39:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 10 Oct 2000 14:39:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: I might not have been clear enough there. What I meant is I tried what the README says, and it does not work. Here is from the SN452 top level readme: ---- It is now possible to automatically configure and build a variety of tools with one command. To build all of the tools contained herein, run the ``configure'' script here, e.g.: ./configure make Yep, this is no longer true. Someone want to update this? The instructions you all gave me though seem to work, so things are definitely looking better now. ;o) You guys give fast answers! I went to get a coffee after writing my question and when I came back I had 3 answers! ;o) I like that. ;o) We've got coverage around the globe. :-) Ben From tromey@cygnus.com Tue Oct 10 14:56:00 2000 From: tromey@cygnus.com (Tom Tromey) Date: Tue, 10 Oct 2000 14:56:00 -0000 Subject: Can't configure in source dir.. References: Message-ID: <87d7h8wdyw.fsf@creche.cygnus.com> >>>>> "Mike" == Mike A Harris writes: Mike> Any idea _why_ it does this? I am almost positive 451 did not Mike> require this. Looking at the configure script it appears that Mike> it is done for some internal cygnus reason.. You can delete that code from the configure script. It should have been removed before making the source available (I think). If you want the full answer, I added that code to configure because in some cases -- namely, multilibbed target libraries -- it is not possible to configure in the source tree. These cases do not affect S-N. Tom From mharris@opensourceadvocate.org Tue Oct 10 15:25:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 15:25:00 -0000 Subject: Can't configure in source dir.. References: <87d7h8wdyw.fsf@creche.cygnus.com> Message-ID: On 10 Oct 2000, Tom Tromey wrote: >Mike> Any idea _why_ it does this? I am almost positive 451 did not >Mike> require this. Looking at the configure script it appears that >Mike> it is done for some internal cygnus reason.. > >You can delete that code from the configure script. It should have >been removed before making the source available (I think). Ok, thanks. I've got it working as is right now, but if I need to, I'll patch it out for the RPMS.. It seems to work now except for a small disk space problem I just had. ;o) >If you want the full answer, I added that code to configure because in >some cases -- namely, multilibbed target libraries -- it is not >possible to configure in the source tree. These cases do not affect >S-N. Ok, thanks. One more question: How much disk space does a complete build require? I had quite a bit of space and it filled the disk.. ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- "If it isn't source, it isn't software." -- NASA From mharris@opensourceadvocate.org Tue Oct 10 23:30:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 10 Oct 2000 23:30:00 -0000 Subject: Build warnings, and install problems. Message-ID: I'm now building from scratch using: mkdir build-redhat cd build-redhat ../configure --prefix=/opt/source-navigator make all-snavigator During the build I get tonnes of: In file included from ../../../db/PORT/../include/db.h:45, from ../../../db/PORT/../recno/rec_utils.c:44: ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' redefined /usr/include/endian.h:47: warning: this is the location of the previous definiti [SNIP] T/../recno ../../../db/PORT/../recno/rec_seq.c In file included from ../../../db/PORT/../include/db.h:45, from ../../../db/PORT/../recno/rec_seq.c:45: ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' redefined /usr/include/endian.h:47: warning: this is the location of the previous definition Then finally: TRACE> make install-snavigator /bin/sh ../mkinstalldirs /usr /usr make[1]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator' Making install in misc make[2]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc' Making install in libutils make[3]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' make[4]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' make[4]: Nothing to be done for `install-exec-am'. /bin/sh ../../../../snavigator/mkinstalldirs /usr/share/sdk/lib mkdir /usr/share/sdk mkdir /usr/share/sdk/lib make[4]: *** [install-sdkLIBRARIES] Error 255 make[4]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' make[3]: *** [install-am] Error 2 make[3]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' make[2]: *** [install-recursive] Error 1 make[2]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator' make: *** [install-snavigator] Error 2 Bad exit status from /home/mharris/rpmbuild/BUILDROOT/rpm-tmp.81690 (%install) Is the "make install-snavigator" target the proper one? I tried just "make install" and it fails too with: make[2]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/po' if test -r ../../../grep/../../grep/mkinstalldirs; then \ ../../../grep/../../grep/mkinstalldirs /usr/share; \ else \ ../../../grep/mkinstalldirs /usr/share; \ fi cp: cannot create regular file `/usr/share/locale/de/LC_MESSAGES/#inst.26406#': Permission denied installing ../../../grep/po/de.gmo as /usr/share/locale/de/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/es/LC_MESSAGES/#inst.26427#': Permission denied installing ../../../grep/po/es.gmo as /usr/share/locale/es/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/fr/LC_MESSAGES/#inst.26448#': Permission denied installing ../../../grep/po/fr.gmo as /usr/share/locale/fr/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/ko/LC_MESSAGES/#inst.26469#': Permission denied installing ../../../grep/po/ko.gmo as /usr/share/locale/ko/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/nl/LC_MESSAGES/#inst.26490#': Permission denied installing ../../../grep/po/nl.gmo as /usr/share/locale/nl/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/no/LC_MESSAGES/#inst.26511#': Permission denied installing ../../../grep/po/no.gmo as /usr/share/locale/no/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/pl/LC_MESSAGES/#inst.26532#': Permission denied installing ../../../grep/po/pl.gmo as /usr/share/locale/pl/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/ru/LC_MESSAGES/#inst.26553#': Permission denied installing ../../../grep/po/ru.gmo as /usr/share/locale/ru/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/sl/LC_MESSAGES/#inst.26574#': Permission denied installing ../../../grep/po/sl.gmo as /usr/share/locale/sl/LC_MESSAGES/grep.mo cp: cannot create regular file `/usr/share/locale/sv/LC_MESSAGES/#inst.26595#': Permission denied installing ../../../grep/po/sv.gmo as /usr/share/locale/sv/LC_MESSAGES/grep.mo if test "grep" = "gettext"; then \ if test -r ../../../grep/../../grep/mkinstalldirs; then \ ../../../grep/../../grep/mkinstalldirs /usr/share/gettext/po; \ else \ ../../../grep/mkinstalldirs /usr/share/gettext/po; \ fi; \ /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c -m 644 ../../../grep/po/Makefile.in.in \ /usr/share/gettext/po/Makefile.in.in; \ else \ : ; \ fi make[2]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/po' Making install in src make[2]: Entering directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/src' /bin/sh ../../../grep/mkinstalldirs /usr/bin /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c grep /usr/bin/grep cp: cannot create regular file `/usr/bin/#inst.26629#': Permission denied /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c egrep /usr/bin/egrep cp: cannot create regular file `/usr/bin/#inst.26651#': Permission denied /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c fgrep /usr/bin/fgrep cp: cannot create regular file `/usr/bin/#inst.26673#': Permission denied make[2]: *** [install-binPROGRAMS] Error 1 make[2]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/src' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep' make: *** [install-grep] Error 2 Bad exit status from /home/mharris/rpmbuild/BUILDROOT/rpm-tmp.74770 (%install) Looks to me like it isn't installing in my BUILDROOT. Looks like it is trying to install on my system instead. If this is true, can someone suggest how to have it install elsewhere? Does SN require any part of the build process to be "root"? Is it possible for an ordinary user to build it, then install it in ~/source-navigator fully functionally? If so how? I am getting it built now, but I don't know if it actually works or not yet.. Thanks again for all the previous help, and TIA! TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Need general help or technical support with Red Hat Linux 6.2? Join the user support mailing list by sending a message to "zoot-list-request@redhat.com" with the word "subscribe" on the subject line. From bje@redhat.com Tue Oct 10 23:35:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 10 Oct 2000 23:35:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: Hi Mike, During the build I get tonnes of: In file included from ../../../db/PORT/../include/db.h:45, from ../../../db/PORT/../recno/rec_utils.c:44: ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' redefined /usr/include/endian.h:47: warning: this is the location of the I submitted a patch recently that eliminates this annoying warning so that you can see real warnings instead. It'll be in the next release. Looks to me like it isn't installing in my BUILDROOT. Looks like it is trying to install on my system instead. If this is true, can someone suggest how to have it install elsewhere? How did you configure the tree? Does SN require any part of the build process to be "root"? Is it possible for an ordinary user to build it, then install it in ~/source-navigator fully functionally? If so how? No, you can be an ordinary user to build and install S-N: /path/to/src/configure --prefix=$HOME/source-navigator .. is all you need. Ben From mharris@opensourceadvocate.org Wed Oct 11 00:05:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 11 Oct 2000 00:05:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: On Wed, 11 Oct 2000, Ben Elliston wrote: >Hi Mike, > > During the build I get tonnes of: > > In file included from ../../../db/PORT/../include/db.h:45, > from ../../../db/PORT/../recno/rec_utils.c:44: > ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' > redefined > /usr/include/endian.h:47: warning: this is the location of the > >I submitted a patch recently that eliminates this annoying warning so that >you can see real warnings instead. It'll be in the next release. Ok good, so it isn't something I foobed. ;o) > Looks to me like it isn't installing in my BUILDROOT. Looks like it > is trying to install on my system instead. If this is true, can > someone suggest how to have it install elsewhere? > >How did you configure the tree? Hmm. I thought I put that in my message: %setup -n SN452-source %build # Clean out the buildroot dir to prevent common errors #[ $RPM_BUILD_ROOT != "/" ] && rm -rf $RPM_BUILD_ROOT mkdir build-redhat cd build-redhat ../configure --prefix=/opt/source-navigator make all-snavigator %install cd build-redhat make install-snavigator The %install section fails with all sorts of permission denied messages and appears it is trying to install into the system dirs instead of the build root. > Does SN require any part of the build process to be "root"? Is it > possible for an ordinary user to build it, then install it in > ~/source-navigator fully functionally? If so how? > >No, you can be an ordinary user to build and install S-N: > > /path/to/src/configure --prefix=$HOME/source-navigator > >.. is all you need. Ok, great. My above configure line should set everything up properly for a package that will install into /opt/source-navigator. The install section follows instructions either Mo, Ian or you sent me.. It doesn't work though. I don't want to compile as root on this system, so I can't do a test install other than as an ordinary user, and it doesn't work for me. I don't see why the install section would require me to have write access to /opt/ since it shouldn't be putting things there during RPM build install time, only at package install time. Any ideas? Thanks, TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- [Quote: Linus Torvalds - Aug 27, 2000 - linux-kernel mailing list] "And I'm right. I'm always right, but in this case I'm just a bit more right than I usually am." -- Linus Torvalds From mdejong@cygnus.com Wed Oct 11 00:09:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 11 Oct 2000 00:09:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: On Wed, 11 Oct 2000, Mike A. Harris wrote: > ../configure --prefix=/opt/source-navigator > messages and appears it is trying to install into the system dirs > instead of the build root. > The install section follows instructions either Mo, Ian or you > sent me.. It doesn't work though. I don't want to compile as > root on this system, Mike, you are trying to install into a directory that is owned by root. You need to su to root before running "make install" or else use a --prefix that your normal user has access to. If you are building an RPM, that needs to be done as root. You will still be able to run the program as a normal user, as the perms are set properly by the install. Mo DeJong Red Hat Inc From bje@redhat.com Wed Oct 11 00:10:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 11 Oct 2000 00:10:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: The install section follows instructions either Mo, Ian or you sent me.. It doesn't work though. I don't want to compile as root on this system, so I can't do a test install other than as an ordinary user, and it doesn't work for me. I don't see why the install section would require me to have write access to /opt/ since it shouldn't be putting things there during RPM build install time, only at package install time. Forgetting RPM for a moment, are you able to build/install from the shell? Ben From mharris@opensourceadvocate.org Wed Oct 11 00:22:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 11 Oct 2000 00:22:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: On Wed, 11 Oct 2000, Mo DeJong wrote: >> ../configure --prefix=/opt/source-navigator > >> messages and appears it is trying to install into the system dirs >> instead of the build root. > >> The install section follows instructions either Mo, Ian or you >> sent me.. It doesn't work though. I don't want to compile as >> root on this system, > >Mike, you are trying to install into a directory that is owned >by root. You need to su to root before running "make install" >or else use a --prefix that your normal user has access >to. If you are building an RPM, that needs to be done as >root. You will still be able to run the program as a normal >user, as the perms are set properly by the install. Sorry, I wasn't clear in my message. I'm doing all of this via RPM as non-root, so the "make install" should be installing to $RPM_BUILD_ROOT/%_prefix, and it should work as non-root as I build all packages as non-root. They won't install with "rpm -i" as non-root, but there is no reason a non-root user shouldn't be able to build RPM's. I've been doing it for months now. The permissions and ownership will be set in the %files section, which I haven't gotten to yet. ;o) If the install target forces owner/group change during install, then I won't be able to use "make install". Perm changes should work ok, but not the o/g. I guess I should have mentioned more clearly how my build process is set up. Any ideas? ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- "Facts do not cease to exist because they are ignored." - Aldous Huxley From mharris@opensourceadvocate.org Wed Oct 11 00:55:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 11 Oct 2000 00:55:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: On Wed, 11 Oct 2000, Ben Elliston wrote: >Date: Wed, 11 Oct 2000 18:10:05 +1100 (EST) >From: Ben Elliston >To: Mike A. Harris >Cc: Source Navigator List >Content-Type: TEXT/PLAIN; charset=US-ASCII >Subject: Re: Build warnings, and install problems. > > The install section follows instructions either Mo, Ian or you sent > me.. It doesn't work though. I don't want to compile as root on this > system, so I can't do a test install other than as an ordinary user, > and it doesn't work for me. I don't see why the install section would > require me to have write access to /opt/ since it shouldn't be putting > things there during RPM build install time, only at package install > time. > >Forgetting RPM for a moment, are you able to build/install from the shell? As non-root yes, but only if I put /home/mharris/source-nav as the prefix of course. I haven't tested if the program actually works or not yet though. ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- "A Firewall is really much like a sophisticated traffic cop; it detects and stops unauthorized or suspicious movement in or out of the network. But security is more than a Firewall; it's a process. You can't just put in a Firewall and think you're secure." From jra@simpod.com Wed Oct 11 06:54:00 2000 From: jra@simpod.com (Jason Andrews) Date: Wed, 11 Oct 2000 06:54:00 -0000 Subject: syntax highlighting Message-ID: <39E3C8A4.1EFAA222@simpod.com> Hi, I have integrated a new parser with sorucenav and I am looking for documentation on syntax highlighting. Any advice on where to find it or which parser to look at as an example is appreciated. Thanks, Jason From spolk@redhat.com Wed Oct 11 09:14:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 11 Oct 2000 09:14:00 -0000 Subject: Build warnings, and install problems. References: Message-ID: <39E492A3.7628CEAA@redhat.com> "Mike A. Harris" wrote: > > I'm now building from scratch using: > > mkdir build-redhat > cd build-redhat > ../configure --prefix=/opt/source-navigator > make all-snavigator > > During the build I get tonnes of: > > In file included from ../../../db/PORT/../include/db.h:45, > from ../../../db/PORT/../recno/rec_utils.c:44: > ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' > redefined > /usr/include/endian.h:47: warning: this is the location of the > previous definiti > > [SNIP] > T/../recno ../../../db/PORT/../recno/rec_seq.c > In file included from ../../../db/PORT/../include/db.h:45, > from ../../../db/PORT/../recno/rec_seq.c:45: > ../../../db/PORT/include/compat.h:108: warning: `BYTE_ORDER' > redefined > /usr/include/endian.h:47: warning: this is the location of the > previous definition These warnings are known and are warnings; they should not have stopped the compilation. > Then finally: > TRACE> make install-snavigator > /bin/sh ../mkinstalldirs /usr /usr > make[1]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator' > Making install in misc > make[2]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc' > Making install in libutils > make[3]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' > make[4]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' > make[4]: Nothing to be done for `install-exec-am'. > /bin/sh ../../../../snavigator/mkinstalldirs /usr/share/sdk/lib > mkdir /usr/share/sdk > mkdir /usr/share/sdk/lib > make[4]: *** [install-sdkLIBRARIES] Error 255 This is all of the output? It looks like you just redirected stdout or stderr, and the other is missing. What happens when you do the following: cd /home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils make install > make[4]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' > make[3]: *** [install-am] Error 2 > make[3]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc/libutils' > make[2]: *** [install-recursive] Error 1 > make[2]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator/misc' > make[1]: *** [install-recursive] Error 1 > make[1]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/snavigator' > make: *** [install-snavigator] Error 2 > Bad exit status from > /home/mharris/rpmbuild/BUILDROOT/rpm-tmp.81690 (%install) > > Is the "make install-snavigator" target the proper one? I tried > just "make install" and it fails too with: Yes, it is. > make[2]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/po' > if test -r ../../../grep/../../grep/mkinstalldirs; then \ > ../../../grep/../../grep/mkinstalldirs /usr/share; \ > else \ > ../../../grep/mkinstalldirs /usr/share; \ > fi > cp: cannot create regular file > `/usr/share/locale/de/LC_MESSAGES/#inst.26406#': You do not have your permission correct. Did you use --prefix when you configured? Are you building as root? > Permission denied > installing ../../../grep/po/de.gmo as > /usr/share/locale/de/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/es/LC_MESSAGES/#inst.26427#': > Permission denied > installing ../../../grep/po/es.gmo as > /usr/share/locale/es/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/fr/LC_MESSAGES/#inst.26448#': > Permission denied > installing ../../../grep/po/fr.gmo as > /usr/share/locale/fr/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/ko/LC_MESSAGES/#inst.26469#': > Permission denied > installing ../../../grep/po/ko.gmo as > /usr/share/locale/ko/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/nl/LC_MESSAGES/#inst.26490#': > Permission denied > installing ../../../grep/po/nl.gmo as > /usr/share/locale/nl/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/no/LC_MESSAGES/#inst.26511#': > Permission denied > installing ../../../grep/po/no.gmo as > /usr/share/locale/no/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/pl/LC_MESSAGES/#inst.26532#': > Permission denied > installing ../../../grep/po/pl.gmo as > /usr/share/locale/pl/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/ru/LC_MESSAGES/#inst.26553#': > Permission denied > installing ../../../grep/po/ru.gmo as > /usr/share/locale/ru/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/sl/LC_MESSAGES/#inst.26574#': > Permission denied > installing ../../../grep/po/sl.gmo as > /usr/share/locale/sl/LC_MESSAGES/grep.mo > cp: cannot create regular file > `/usr/share/locale/sv/LC_MESSAGES/#inst.26595#': > Permission denied > installing ../../../grep/po/sv.gmo as > /usr/share/locale/sv/LC_MESSAGES/grep.mo > if test "grep" = "gettext"; then \ > if test -r ../../../grep/../../grep/mkinstalldirs; then \ > ../../../grep/../../grep/mkinstalldirs > /usr/share/gettext/po; \ > else \ > ../../../grep/mkinstalldirs /usr/share/gettext/po; \ > fi; \ > /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c > -m 644 ../../../grep/po/Makefile.in.in \ > /usr/share/gettext/po/Makefile.in.in; \ > else \ > : ; \ > fi > make[2]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/po' > Making install in src > make[2]: Entering directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/src' > /bin/sh ../../../grep/mkinstalldirs /usr/bin > /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c > grep /usr/bin/grep > cp: cannot create regular file > `/usr/bin/#inst.26629#': Permission denied > /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c > egrep /usr/bin/egrep > cp: cannot create regular file > `/usr/bin/#inst.26651#': Permission denied > /bin/sh /home/mharris/rpmbuild/BUILD/SN452-source/install-sh -c > fgrep /usr/bin/fgrep > cp: cannot create regular file > `/usr/bin/#inst.26673#': Permission denied > make[2]: *** [install-binPROGRAMS] Error 1 > make[2]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep/src' > make[1]: *** [install-recursive] Error 1 > make[1]: Leaving directory > `/home/mharris/rpmbuild/BUILD/SN452-source/build-redhat/grep' > make: *** [install-grep] Error 2 > Bad exit status from > /home/mharris/rpmbuild/BUILDROOT/rpm-tmp.74770 (%install) > > Looks to me like it isn't installing in my BUILDROOT. Looks like > it is trying to install on my system instead. If this is true, > can someone suggest how to have it install elsewhere? > > Does SN require any part of the build process to be "root"? Is > it possible for an ordinary user to build it, then install it in > ~/source-navigator fully functionally? If so how? > > I am getting it built now, but I don't know if it actually works > or not yet.. > > Thanks again for all the previous help, and TIA! > > TTYL > > ---------------------------------------------------------------------- > Mike A. Harris - Linux advocate - Open source advocate > Computer Consultant - Capslock Consulting > Copyright 2000 all rights reserved > ---------------------------------------------------------------------- > > Need general help or technical support with Red Hat Linux 6.2? Join the user > support mailing list by sending a message to "zoot-list-request@redhat.com" > with the word "subscribe" on the subject line. From bje@redhat.com Wed Oct 11 11:23:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 11 Oct 2000 11:23:00 -0000 Subject: syntax highlighting References: <39E3C8A4.1EFAA222@simpod.com> Message-ID: Hi, I have integrated a new parser with sorucenav and I am looking for documentation on syntax highlighting. The syntax highlighters have specific hooks into the `hyper' interpreter. For an example, go to the `hyper' source directory and grep for `ch_highlight' in highlight.c. The idea is that an init function is called once to initialise things and then an entry point is called periodically whenever S-N's editor wants a memory buffer (usually the file currently loaded into the editor) to be scanned for highlighting information. Such scanners are usually written using Flex. Any advice on where to find it or which parser to look at as an example is appreciated. Take a look at hyper/*.l. Yes, it's not as clean as it ought to be. Highlighters should be loaded into the interpreter dynamically. Cheers, Ben From m_manohar@yahoo.com Thu Oct 12 00:26:00 2000 From: m_manohar@yahoo.com (M Manohar) Date: Thu, 12 Oct 2000 00:26:00 -0000 Subject: C++ parser problem (RFC) Message-ID: <20001012072552.12246.qmail@web3506.mail.yahoo.com> here is a code snippet ---------- class dummy { public: int no_ref; int func(void) {no_ref = 5;}; int func1(void) {no_ref = 5;}; }; snippet end ---------- Trying to generate xref for no_ref gives nothing Hope this helps -Manohar --- Ian Roxborough wrote: > > > Hi, > > Can all the people who are having problems please > try to > send in test cases (easy, minimal, code snippets) > with details > instructions. If you can't get a really good > reproducible test > case try to post an outline, somebody might be have > similar > problems and beable to produce a good test case. > > If you want your bugs fixed and (for whatever > reason) can't > fix it yourself, remember the better a bug report > you give us > the easier it is for us to fix. > > Thanks, > Ian. > > P.S.: We do want to fix all the bugs in SN > regardless of who's paying us. > > P.S.S.: Please keep posts on topic. > __________________________________________________ Do You Yahoo!? Get Yahoo! Mail - Free email you can access from anywhere! http://mail.yahoo.com/ From ashok.raj@intel.com Fri Oct 13 08:37:00 2000 From: ashok.raj@intel.com (Raj, Ashok) Date: Fri, 13 Oct 2000 08:37:00 -0000 Subject: problem with source navigator when run on linux kernel sources... Message-ID: <9319DDF797C4D211AC4700A0C96B7C9404AC1AA9@orsmsx42.jf.intel.com> Hello. I tried both the 4.5.1 and 4.5.2 against linux sources, after it starts creating the project, one of the sub processes cbrowser??? seems to just spin for ever. the progress bar shows 6% and never moves. Have you tried the snavigator on the linux tree? on small projects it seems ok. i would really like to use it on the kernel sources.. if you get this to work or know of a workaround, please let me know. btw: i tried running strace on it, and it did not print anything, seems like its spinning somewhere in the code. ashokr From akoifman@mediaone.net Fri Oct 13 13:52:00 2000 From: akoifman@mediaone.net (Alex Koifman) Date: Fri, 13 Oct 2000 13:52:00 -0000 Subject: Search find not highlighted? Message-ID: <000b01c03557$b816c220$500110ac@qb.local> Hi, ???? I just installed Source-Navigator 4.5.2 on Windows NT 4.0 (patch 5).???? Rebooted. ???? When in the Edit window, I do Control-F and type the search string, press Search.???? The Navigator finds the searched string, but it is not highlighted.???? I move the dialog window around, still not highlighted.???? I can see the found string.???? When I Cancel, the found string gets highlighted. ???? Why? ???? When select (double click) a search string and press "Find selection or next pattern" the found string is highlighted. ???? Thanks a lot, ________________________________ Alex Koifman srcnav.gif From bje@air.net.au Fri Oct 13 14:44:00 2000 From: bje@air.net.au (Ben Elliston) Date: Fri, 13 Oct 2000 14:44:00 -0000 Subject: problem with source navigator when run on linux kernel sources... References: <9319DDF797C4D211AC4700A0C96B7C9404AC1AA9@orsmsx42.jf.intel.com> Message-ID: Hi, I tried both the 4.5.1 and 4.5.2 against linux sources, after it starts creating the project, one of the sub processes cbrowser??? seems to just spin for ever. the progress bar shows 6% and never moves. Have you tried the snavigator on the linux tree? on small projects it seems ok. i would really like to use it on the kernel sources.. What version of the kernel source tree are you trying to parse? That should make it possible to reproduce. Ben From irox@redhat.com Fri Oct 13 15:25:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 13 Oct 2000 15:25:00 -0000 Subject: Search find not highlighted? References: <000b01c03557$b816c220$500110ac@qb.local> Message-ID: <39E78BBA.76682E95@redhat.com> Hi, (Please avoid posting in HTML as not every is using an HTML enabled email client.) > When in the Edit window, I do Control-F and type the search string, > press Search. The Navigator finds the searched string, but it is not > highlighted. I move the dialog window around, still not highlighted. > I can see the found string. When I Cancel, the found string gets > highlighted. > > Why? Thinking back, there was a macro of something that was meant to be set(?) when building tk, ALWAYS_SHOW_SELECTION I think, that allows multiple selection, otherwise the selection would be only where the focus was (?). I think ALWAYS_SHOW_SELECTION should have been set at configure time (but I could be wrong). This is the only thing the springs to mind. Ian. > When select (double click) a search string and press "Find selection > or next pattern" the found string is highlighted. > > Thanks a lot, > ________________________________ > Alex Koifman > > Name: srcnav.gif > srcnav.gif Type: GIF Image (image/gif) > Encoding: base64 From akoifman@mediaone.net Fri Oct 13 15:30:00 2000 From: akoifman@mediaone.net (Alex Koifman) Date: Fri, 13 Oct 2000 15:30:00 -0000 Subject: Search find not highlighted? References: <000b01c03557$b816c220$500110ac@qb.local> <39E78BBA.76682E95@redhat.com> Message-ID: <001b01c03565$6a1dc9c0$500110ac@qb.local> Ian, Thanks for your reply. I am new to Source Navigator. Where do I set ALWAYS_SHOW_SELECTION? BTW, this used to work in the older version that we bought, v 99r1. Thanks, Alex. ----- Original Message ----- From: "Ian Roxborough" To: "Alex Koifman" Cc: Sent: Friday, October 13, 2000 6:24 PM Subject: Re: Search find not highlighted? > > Hi, > > (Please avoid posting in HTML as not every is using an HTML > enabled email client.) > > > When in the Edit window, I do Control-F and type the search string, > > press Search. The Navigator finds the searched string, but it is not > > highlighted. I move the dialog window around, still not highlighted. > > I can see the found string. When I Cancel, the found string gets > > highlighted. > > > > Why? > > Thinking back, there was a macro of something that was meant to be > set(?) when building tk, ALWAYS_SHOW_SELECTION I think, that allows > multiple selection, otherwise the selection would be only where the > focus was (?). I think ALWAYS_SHOW_SELECTION should have been set at > configure time (but I could be wrong). > > This is the only thing the springs to mind. > > Ian. > > > When select (double click) a search string and press "Find selection > > or next pattern" the found string is highlighted. > > > > Thanks a lot, > > ________________________________ > > Alex Koifman > > > > Name: srcnav.gif > > srcnav.gif Type: GIF Image (image/gif) > > Encoding: base64 From jra@simpod.com Fri Oct 13 15:31:00 2000 From: jra@simpod.com (Jason Andrews) Date: Fri, 13 Oct 2000 15:31:00 -0000 Subject: syntax highlighting Message-ID: <39E6E4B0.2E4040EF@simpod.com> Can you explain the relationship between what is in the share/etc/sn_prop.cfg file specifically the -high_cmd argument for adding parsers and the info you gave me below? I have entered my new parser in the -brow_cmd and also the -high_cmd I added hooks to the highlight.c stuff you mentioned below but this code is never called for my parser's language. I guess I don't see the connection between my flex based parser and these other flex files for syntax highlighting. Maybe I need something else for the -high_cmd field. Any help is much appreciated. Jason -------------- Hi, I have integrated a new parser with sorucenav and I am looking for documentation on syntax highlighting. The syntax highlighters have specific hooks into the `hyper' interpreter. For an example, go to the `hyper' source directory and grep for `ch_highlight' in highlight.c. The idea is that an init function is called once to initialise things and then an entry point is called periodically whenever S-N's editor wants a memory buffer (usually the file currently loaded into the editor) to be scanned for highlighting information. Such scanners are usually written using Flex. Any advice on where to find it or which parser to look at as an example is appreciated. Take a look at hyper/*.l. Yes, it's not as clean as it ought to be. Highlighters should be loaded into the interpreter dynamically. Cheers, Ben From irox@redhat.com Fri Oct 13 16:00:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 13 Oct 2000 16:00:00 -0000 Subject: Search find not highlighted? Message-ID: <200010132257.PAA02898@manx.soma.redhat.com> On Fri, 13 Oct 2000, Alex Koifman wrote: > I am new to Source Navigator. Where do I set ALWAYS_SHOW_SELECTION? > BTW, this used to work in the older version that we bought, v 99r1. This should set a build time, CFLAGS=..... -DALWAYS_SHOW_SELECTION, I can't remember the SN windows build steps off hand, but adding it to CFLAGS should work. Ian. > ----- Original Message ----- > From: "Ian Roxborough" > To: "Alex Koifman" > Cc: > Sent: Friday, October 13, 2000 6:24 PM > Subject: Re: Search find not highlighted? > > > > > > Hi, > > > > (Please avoid posting in HTML as not every is using an HTML > > enabled email client.) > > > > > When in the Edit window, I do Control-F and type the search string, > > > press Search. The Navigator finds the searched string, but it is not > > > highlighted. I move the dialog window around, still not highlighted. > > > I can see the found string. When I Cancel, the found string gets > > > highlighted. > > > > > > Why? > > > > Thinking back, there was a macro of something that was meant to be > > set(?) when building tk, ALWAYS_SHOW_SELECTION I think, that allows > > multiple selection, otherwise the selection would be only where the > > focus was (?). I think ALWAYS_SHOW_SELECTION should have been set at > > configure time (but I could be wrong). > > > > This is the only thing the springs to mind. > > > > Ian. > > > > > When select (double click) a search string and press "Find selection > > > or next pattern" the found string is highlighted. > > > > > > Thanks a lot, > > > ________________________________ > > > Alex Koifman > > > > > > Name: srcnav.gif > > > srcnav.gif Type: GIF Image (image/gif) > > > Encoding: base64 > > From bje@air.net.au Fri Oct 13 17:27:00 2000 From: bje@air.net.au (Ben Elliston) Date: Fri, 13 Oct 2000 17:27:00 -0000 Subject: syntax highlighting References: <39E6E4B0.2E4040EF@simpod.com> Message-ID: Can you explain the relationship between what is in the share/etc/sn_prop.cfg file specifically the -high_cmd argument for adding parsers and the info you gave me below? I have entered my new parser in the -brow_cmd and also the -high_cmd The -high_cmd is no longer applicable. In fact, it needs to be removed. Once upon a time, syntax highlighting was done by an external program. I added hooks to the highlight.c stuff you mentioned below but this code is never called for my parser's language. I guess I don't see the connection between my flex based parser and these other flex files for syntax highlighting. Maybe I need something else for the -high_cmd field. Can you post a diff for your changes to highlight.c? Ben From berek@usa.net Fri Oct 13 17:37:00 2000 From: berek@usa.net (Berek) Date: Fri, 13 Oct 2000 17:37:00 -0000 Subject: problem with source navigator when run on linux kernel sources... References: Message-ID: I've had the same problem, only on NT 4.0. It's dbimp. Goes into a loop and pegs the cpu. Cannot reproduce at will. Happens one out of eight or nine times. Reparsing same project. 560,000 plus lines. Does not happen on smaller projects or subprojects. Only when combining subprojects into a larger 500,000+ lines of code. May not be memory-related because the machine(s) (PCs) on which this is happening both have 512 meg real and 1 gig virtual. > -----Original Message----- > From: sourcenav-owner@sources.redhat.com > [ mailto:sourcenav-owner@sources.redhat.com]On Behalf Of Ben Elliston > Sent: Friday, October 13, 2000 17:45 > To: Raj, Ashok > Cc: 'sourcenav@sources.redhat.com' > Subject: Re: problem with source navigator when run on linux kernel > sources... > > > Hi, > > I tried both the 4.5.1 and 4.5.2 against linux sources, after it > starts creating the project, one of the sub processes cbrowser??? > seems to just spin for ever. the progress bar shows 6% and never > moves. Have you tried the snavigator on the linux tree? on small > projects it seems ok. i would really like to use it on the kernel > sources.. > > What version of the kernel source tree are you trying to parse? > That should > make it possible to reproduce. > > Ben > > From akoifman@mediaone.net Sat Oct 14 10:32:00 2000 From: akoifman@mediaone.net (Alex Koifman) Date: Sat, 14 Oct 2000 10:32:00 -0000 Subject: Search find not highlighted? References: <200010132257.PAA02898@manx.soma.redhat.com> Message-ID: <002001c03604$b2d8dde0$7cfeda18@ne.mediaone.net> SN for Windows comes as an install. There is no Makefile to set flags. Should I wait for a patch? Will someone add the flag to the Windows build? Why would the flag be set otherwise? Thanks a lot, Alex. ----- Original Message ----- From: "Ian Roxborough" To: "Alex Koifman" Cc: Sent: Friday, October 13, 2000 6:57 PM Subject: Re: Search find not highlighted? > > On Fri, 13 Oct 2000, Alex Koifman wrote: > > I am new to Source Navigator. Where do I set ALWAYS_SHOW_SELECTION? > > BTW, this used to work in the older version that we bought, v 99r1. > > This should set a build time, CFLAGS=..... -DALWAYS_SHOW_SELECTION, > I can't remember the SN windows build steps off hand, but adding it > to CFLAGS should work. > > Ian. > > > ----- Original Message ----- > > From: "Ian Roxborough" > > To: "Alex Koifman" > > Cc: > > Sent: Friday, October 13, 2000 6:24 PM > > Subject: Re: Search find not highlighted? > > > > > > > > > > Hi, > > > > > > (Please avoid posting in HTML as not every is using an HTML > > > enabled email client.) > > > > > > > When in the Edit window, I do Control-F and type the search string, > > > > press Search. The Navigator finds the searched string, but it is not > > > > highlighted. I move the dialog window around, still not highlighted. > > > > I can see the found string. When I Cancel, the found string gets > > > > highlighted. > > > > > > > > Why? > > > > > > Thinking back, there was a macro of something that was meant to be > > > set(?) when building tk, ALWAYS_SHOW_SELECTION I think, that allows > > > multiple selection, otherwise the selection would be only where the > > > focus was (?). I think ALWAYS_SHOW_SELECTION should have been set at > > > configure time (but I could be wrong). > > > > > > This is the only thing the springs to mind. > > > > > > Ian. > > > > > > > When select (double click) a search string and press "Find selection > > > > or next pattern" the found string is highlighted. > > > > > > > > Thanks a lot, > > > > ________________________________ > > > > Alex Koifman > > > > > > > > Name: srcnav.gif > > > > srcnav.gif Type: GIF Image (image/gif) > > > > Encoding: base64 > > > > From jra@simpod.com Sat Oct 14 18:13:00 2000 From: jra@simpod.com (Jason Andrews) Date: Sat, 14 Oct 2000 18:13:00 -0000 Subject: syntax highlighting References: Message-ID: <39E85C35.7D953470@simpod.com> Ben Elliston wrote: > Can you explain the relationship between what is in the > share/etc/sn_prop.cfg file specifically the -high_cmd argument for > adding parsers and the info you gave me below? I have entered my new > parser in the -brow_cmd and also the -high_cmd > > The -high_cmd is no longer applicable. In fact, it needs to be > removed. Once upon a time, syntax highlighting was done by an external > program. > > I added hooks to the highlight.c stuff you mentioned below but this > code is never called for my parser's language. I guess I don't see > the connection between my flex based parser and these other flex files > for syntax highlighting. Maybe I need something else for the -high_cmd > field. > > Can you post a diff for your changes to highlight.c? > > Ben Thanks, but I got it. The key was to edit sninit.tcl and add my language to the list of those with built-in-highlighting. Now the functions in highlight.c are called as you described. From spolk@redhat.com Mon Oct 16 08:55:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 16 Oct 2000 08:55:00 -0000 Subject: Search find not highlighted? References: <200010132257.PAA02898@manx.soma.redhat.com> <002001c03604$b2d8dde0$7cfeda18@ne.mediaone.net> Message-ID: <39EB25AA.C23888AF@redhat.com> Well, it is possible to build the source tarball for Windows, just not easily. The next build will have this flag set. Alex Koifman wrote: > > SN for Windows comes as an install. There is no Makefile to set flags. > > Should I wait for a patch? Will someone add the flag to the Windows build? > Why would the flag be set otherwise? > > Thanks a lot, > > Alex. > ----- Original Message ----- > From: "Ian Roxborough" > To: "Alex Koifman" > Cc: > Sent: Friday, October 13, 2000 6:57 PM > Subject: Re: Search find not highlighted? > > > > > On Fri, 13 Oct 2000, Alex Koifman wrote: > > > I am new to Source Navigator. Where do I set ALWAYS_SHOW_SELECTION? > > > BTW, this used to work in the older version that we bought, v 99r1. > > > > This should set a build time, CFLAGS=..... -DALWAYS_SHOW_SELECTION, > > I can't remember the SN windows build steps off hand, but adding it > > to CFLAGS should work. > > > > Ian. > > > > > ----- Original Message ----- > > > From: "Ian Roxborough" > > > To: "Alex Koifman" > > > Cc: > > > Sent: Friday, October 13, 2000 6:24 PM > > > Subject: Re: Search find not highlighted? > > > > > > > > > > > > > > Hi, > > > > > > > > (Please avoid posting in HTML as not every is using an HTML > > > > enabled email client.) > > > > > > > > > When in the Edit window, I do Control-F and type the search string, > > > > > press Search. The Navigator finds the searched string, but it is > not > > > > > highlighted. I move the dialog window around, still not > highlighted. > > > > > I can see the found string. When I Cancel, the found string gets > > > > > highlighted. > > > > > > > > > > Why? > > > > > > > > Thinking back, there was a macro of something that was meant to be > > > > set(?) when building tk, ALWAYS_SHOW_SELECTION I think, that allows > > > > multiple selection, otherwise the selection would be only where the > > > > focus was (?). I think ALWAYS_SHOW_SELECTION should have been set at > > > > configure time (but I could be wrong). > > > > > > > > This is the only thing the springs to mind. > > > > > > > > Ian. > > > > > > > > > When select (double click) a search string and press "Find selection > > > > > or next pattern" the found string is highlighted. > > > > > > > > > > Thanks a lot, > > > > > ________________________________ > > > > > Alex Koifman > > > > > > > > > > Name: srcnav.gif > > > > > srcnav.gif Type: GIF Image (image/gif) > > > > > Encoding: base64 > > > > > > From mdejong@cygnus.com Mon Oct 16 13:48:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 16 Oct 2000 13:48:00 -0000 Subject: SN 4.5.2 passed the 1000 downloads mark on freshmeat. Message-ID: Wow! SN downloads are really picking up on freshmeat. We just passed the 1000 downloads mark for the 4.5.2 release. http://freshmeat.net/projects/source-navigator Mo DeJong Red Hat Inc From justin@ee.byu.edu Mon Oct 16 19:58:00 2000 From: justin@ee.byu.edu (Justin Tripp) Date: Mon, 16 Oct 2000 19:58:00 -0000 Subject: SourceNavigator. Message-ID: Dear SourceNavigator Team, My name is Justin Tripp and I am a member of the GNU software evaluation team. It came up in discussion as to whether or not SourceNavigator would have a command-line interface. Is it in your plans to have a command-line interface? Thank you for helping us with this information. .justin. ------------------------------------------------------------------------ Justin Leonard Tripp justin@ee.byu.edu Configurable Computing Laboratory Research Assistant CB 461 x8-7206 Electrical and Computer Engineering Department Brigham Young University From mdejong@cygnus.com Mon Oct 16 20:09:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 16 Oct 2000 20:09:00 -0000 Subject: SourceNavigator. References: Message-ID: On Mon, 16 Oct 2000, Justin Tripp wrote: > > Dear SourceNavigator Team, > My name is Justin Tripp and I am a member of the GNU software evaluation > team. It came up in discussion as to whether or not SourceNavigator would > have a command-line interface. Is it in your plans to have a command-line > interface? Thank you for helping us with this information. > > .justin We are planning on doing a command line interface to the parsers and database generation code, but not the GUI itself. It would be quite a bit of work to add a command line interface, but not impossible. Mo DeJong Red Hat Inc From justin@ee.byu.edu Mon Oct 16 20:14:00 2000 From: justin@ee.byu.edu (Justin Tripp) Date: Mon, 16 Oct 2000 20:14:00 -0000 Subject: SourceNavigator. References: Message-ID: Thanks, I think you have done a bang-up job on the gui and it would be difficult to reimplement everything as a command-line. Today, we spent 30 minutes, scratching our heads trying to get CodeWarrior to actually import some .c and .h files into a new project. I had alot easier time working with SourceNavigator doing a similar sort of task. Keep up the good work. .justin. On Mon, 16 Oct 2000, Mo DeJong wrote: ;)On Mon, 16 Oct 2000, Justin Tripp wrote: ;) ;)> ;)> Dear SourceNavigator Team, ;)> My name is Justin Tripp and I am a member of the GNU software evaluation ;)> team. It came up in discussion as to whether or not SourceNavigator would ;)> have a command-line interface. Is it in your plans to have a command-line ;)> interface? Thank you for helping us with this information. ;)> ;)> .justin ;) ;)We are planning on doing a command line interface to the ;)parsers and database generation code, but not the GUI itself. ;) ;)It would be quite a bit of work to add a command line ;)interface, but not impossible. ;) ;)Mo DeJong ;)Red Hat Inc ;) ------------------------------------------------------------------------ Justin Leonard Tripp justin@ee.byu.edu Configurable Computing Laboratory Research Assistant CB 461 x8-7206 Electrical and Computer Engineering Department Brigham Young University From irox@redhat.com Mon Oct 16 21:03:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Mon, 16 Oct 2000 21:03:00 -0000 Subject: SourceNavigator. References: Message-ID: <39EBCF95.B1A4DCE4@redhat.com> Thanks! I command line interface would be a very cool feature. If for the only reason of report generating, but also for other thing like update database info during the night (cron job style). I think most of the more useful feature could be implemented as tcl scripts to output either in text or HTML. (There is some HTML support but it is current broken in 4.5x). I believe that currently run Source-Navigator requires an X server even when ran as a command line app. (?) There is more info about running SN from the command line in the SDK doc (Programmers Reference Guide) with some examples. Ian. Justin Tripp wrote: > > Thanks, I think you have done a bang-up job on the gui and it would be > difficult to reimplement everything as a command-line. Today, we spent 30 > minutes, scratching our heads trying to get CodeWarrior to actually import > some .c and .h files into a new project. I had alot easier time working > with SourceNavigator doing a similar sort of task. Keep up the good work. > > .justin. > > On Mon, 16 Oct 2000, Mo DeJong wrote: > > ;)On Mon, 16 Oct 2000, Justin Tripp wrote: > ;) > ;)> > ;)> Dear SourceNavigator Team, > ;)> My name is Justin Tripp and I am a member of the GNU software evaluation > ;)> team. It came up in discussion as to whether or not SourceNavigator would > ;)> have a command-line interface. Is it in your plans to have a command-line > ;)> interface? Thank you for helping us with this information. > ;)> > ;)> .justin > ;) > ;)We are planning on doing a command line interface to the > ;)parsers and database generation code, but not the GUI itself. > ;) > ;)It would be quite a bit of work to add a command line > ;)interface, but not impossible. > ;) > ;)Mo DeJong > ;)Red Hat Inc > ;) > > ------------------------------------------------------------------------ > Justin Leonard Tripp justin@ee.byu.edu > Configurable Computing Laboratory Research Assistant CB 461 x8-7206 > Electrical and Computer Engineering Department Brigham Young University From bje@redhat.com Mon Oct 16 21:13:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 16 Oct 2000 21:13:00 -0000 Subject: SourceNavigator. References: Message-ID: My name is Justin Tripp and I am a member of the GNU software evaluation team. It came up in discussion as to whether or not SourceNavigator would have a command-line interface. Is it in your plans to have a command-line interface? Thank you for helping us with this information. This is partly in place already. It's possible to generate S-N projects (ie. parsing source code) from the command line now. Getting at the database is also possible, just slightly more difficult. Ben From tsys@uswest.net Tue Oct 17 08:36:00 2000 From: tsys@uswest.net (Lonnie L VanZandt) Date: Tue, 17 Oct 2000 08:36:00 -0000 Subject: xref and referred by? Message-ID: Why is it that many times the Referenced By information is unavailable? More often than not, I want to see who calls a function and not what that function calls. However, I can't seem to get SN to display the Referenced By information. Do you have a FAQ I should have read before bothering you? Thanks, Lonnie. From bruce.stephens@messagingdirect.com Tue Oct 17 08:45:00 2000 From: bruce.stephens@messagingdirect.com (Bruce Stephens) Date: Tue, 17 Oct 2000 08:45:00 -0000 Subject: xref and referred by? References: Message-ID: Lonnie L VanZandt writes: > Why is it that many times the Referenced By information is > unavailable? More often than not, I want to see who calls a function > and not what that function calls. However, I can't seem to get SN to > display the Referenced By information. xref requires that certain databases be completely constructed. The first time you run source navigator on a project, these can take some time to construct (potentially hours, for large projects and/or slow machines). There's a progress bar in the bottom right corner of any SN windows while this is going on. After the first time, SN should be faster to start up. -- Bruce Stephens Bruce.Stephens@MessagingDirect.com MessagingDirect(UK) Ltd From erayo@cs.bilkent.edu.tr Tue Oct 17 09:31:00 2000 From: erayo@cs.bilkent.edu.tr (Eray 'exa' Ozkural) Date: Tue, 17 Oct 2000 09:31:00 -0000 Subject: sourcenav deb dir layout Message-ID: The following is the dir layout of sourcenav package: There are several problems with it. It doesn't seem to comply with FHS or GNU coding standards. I can fix some of it, but ultimately you should decide how it will play nicely with a system. What is your opinion? -- usr |-- bin |-- html | |-- progref | `-- userguide |-- lib |-- sbin `-- share |-- bitmaps |-- demos |-- doc | `-- sourcenav |-- etc |-- gui `-- sdk |-- api | |-- c | | `-- database | | `-- examples | `-- tcl | |-- database | | `-- examples | `-- misc |-- include |-- lib `-- parsers `-- examples |-- assembly `-- elf From erayo@cs.bilkent.edu.tr Tue Oct 17 09:34:00 2000 From: erayo@cs.bilkent.edu.tr (Eray 'exa' Ozkural) Date: Tue, 17 Oct 2000 09:34:00 -0000 Subject: distclean target incomplete Message-ID: I issue 'make distclean' in parallel build directory. Making distclean in ../parsers/cpp/cpplib make[3]: Entering directory `/home/exa/devel/debian/sourcenav-4.5.2/debian/build/snavigator/parsers/cpp/cpplib' make[3]: *** No rule to make target `distclean'. Stop. make[3]: Leaving directory `/home/exa/devel/debian/sourcenav-4.5.2/debian/build/snavigator/parsers/cpp/cpplib' make[2]: *** [distclean-recursive] Error 1 make[2]: Leaving directory `/home/exa/devel/debian/sourcenav-4.5.2/debian/build/snavigator/db' make[1]: *** [distclean-recursive] Error 1 make[1]: Leaving directory `/home/exa/devel/debian/sourcenav-4.5.2/debian/build/snavigator' make: *** [do-distclean] Error 1 clean target seems to be okay. From spolk@redhat.com Tue Oct 17 10:39:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 17 Oct 2000 10:39:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: <4.2.0.58.20001017103931.01a09100@pop.cygnus.com> At 07:36 PM 10/17/00 +0300, Eray 'exa' Ozkural wrote: >The following is the dir layout of sourcenav package: > >There are several problems with it. It doesn't seem to comply with >FHS or GNU coding standards. I can fix some of it, but ultimately >you should decide how it will play nicely with a system. >What is your opinion? First of all, this project was never written to the FHS or GNU coding standards. Our official coding standards are the Tcl coding standards. This is for all new code. The original code was never written to a standard; we are slowly fixing that. Secondly, what changes are you proposing? >-- usr > |-- bin > |-- html > | |-- progref > | `-- userguide > |-- lib > |-- sbin > `-- share > |-- bitmaps > |-- demos > |-- doc > | `-- sourcenav > |-- etc > |-- gui > `-- sdk > |-- api > | |-- c > | | `-- database > | | `-- examples > | `-- tcl > | |-- database > | | `-- examples > | `-- misc > |-- include > |-- lib > `-- parsers > `-- examples > |-- assembly > `-- elf > > Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From wgacquer@ubisoft.fr Tue Oct 17 16:11:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Tue, 17 Oct 2000 16:11:00 -0000 Subject: help with unused functions Message-ID: <5D2A198567EED311BD0F009027A25C874905F9@SRVMAIL-ANN> Can somebody send me a short tcl script or C code that retrieves all unused functions in a C project? I apologize for asking this kind of question but I am in hurry : 2 days to finish our game.... Thanks in advance. William ________________ William Gacquer Rayman 2 PS2 team - Ubi Simulations - Annecy France From mdejong@cygnus.com Tue Oct 17 16:18:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 17 Oct 2000 16:18:00 -0000 Subject: help with unused functions References: <5D2A198567EED311BD0F009027A25C874905F9@SRVMAIL-ANN> Message-ID: On Wed, 18 Oct 2000, William Gacquer wrote: > Can somebody send me a short tcl script or C code that retrieves all unused > functions in a C project? > > I apologize for asking this kind of question but I am in hurry : 2 days to > finish our game.... > > Thanks in advance. > William Take a look at share/sdk/api/tcl/database/examples/unused.tcl in the install dir. It prints unused variables, you will have to adapt it a bit but it should be of some help. There are other examples in that same dir that can help you get started. Mo DeJong Red Hat Inc From wgacquer@ubisoft.fr Tue Oct 17 16:46:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Tue, 17 Oct 2000 16:46:00 -0000 Subject: help with unused functions Message-ID: <5D2A198567EED311BD0F009027A25C874905FB@SRVMAIL-ANN> Thanks for being so fast! should in look in the ".to" or the ".by" databases? I see no other database to look at to retrieve the "access" status of a function. But when I have an isolated function - foo(), for instance - in the code that neither used nor referring to anything, foo() is visible in .fil, .fu but not .to and .by. Does that mean that SN is not able to find the isolated functions? -----Original Message----- From: Mo DeJong [ mailto:mdejong@cygnus.com ] Sent: mercredi 18 octobre 2000 01:19 To: sourcenav@sourceware.cygnus.com Subject: Re: help with unused functions On Wed, 18 Oct 2000, William Gacquer wrote: > Can somebody send me a short tcl script or C code that retrieves all unused > functions in a C project? > > I apologize for asking this kind of question but I am in hurry : 2 days to > finish our game.... > > Thanks in advance. > William Take a look at share/sdk/api/tcl/database/examples/unused.tcl in the install dir. It prints unused variables, you will have to adapt it a bit but it should be of some help. There are other examples in that same dir that can help you get started. Mo DeJong Red Hat Inc From bje@redhat.com Tue Oct 17 16:55:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 17 Oct 2000 16:55:00 -0000 Subject: help with unused functions References: <5D2A198567EED311BD0F009027A25C874905FB@SRVMAIL-ANN> Message-ID: should in look in the ".to" or the ".by" databases? I see no other database to look at to retrieve the "access" status of a function. Correct. But when I have an isolated function - foo(), for instance - in the code that neither used nor referring to anything, foo() is visible in .fil, .fu but not .to and .by. That's right. Does that mean that SN is not able to find the isolated functions? No, it means it has no cross-references. Ben From erayo@cs.bilkent.edu.tr Tue Oct 17 17:01:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Tue, 17 Oct 2000 17:01:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001017103931.01a09100@pop.cygnus.com> Message-ID: <39ECE896.A12F764F@cs.bilkent.edu.tr> Syd Polk wrote: > > First of all, this project was never written to the FHS or GNU coding > standards. Our official coding standards are the Tcl coding standards. This > is for all new code. The original code was never written to a standard; we > are slowly fixing that. > Excuse my ignorance, I'm not aware of tcl coding standards. Does it indicate where you should be putting docs or binary data? > Secondly, what changes are you proposing? > First, all doc files go to their debian doc dir /usr/share/doc/sourcenav. that doesn't concern you much I guess. demos also go to doc dir. share/bitmaps, share/gui -> share/sourcenav/ i don't know where to put sdk, since sdk is not specific to sourcenav? if it is specific to sourcenav, then it also goes under share/sourcenav it looks like read only arch-independent data. Cheers, > >-- usr > > |-- bin > > |-- html > > | |-- progref > > | `-- userguide > > |-- lib > > |-- sbin > > `-- share > > |-- bitmaps > > |-- demos > > |-- doc > > | `-- sourcenav > > |-- etc > > |-- gui > > `-- sdk > > |-- api > > | |-- c > > | | `-- database > > | | `-- examples > > | `-- tcl > > | |-- database > > | | `-- examples > > | `-- misc > > |-- include > > |-- lib > > `-- parsers > > `-- examples > > |-- assembly > > `-- elf > > > > -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From spolk@redhat.com Tue Oct 17 17:58:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 17 Oct 2000 17:58:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001017103931.01a09100@pop.cygnus.com> <39ECE896.A12F764F@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001017175902.01a3b730@pop.cygnus.com> At 03:02 AM 10/18/00 +0300, Eray Ozkural wrote: >Syd Polk wrote: > > > > First of all, this project was never written to the FHS or GNU coding > > standards. Our official coding standards are the Tcl coding standards. This > > is for all new code. The original code was never written to a standard; we > > are slowly fixing that. > > > >Excuse my ignorance, I'm not aware of tcl coding standards. Does it >indicate where you should be putting docs or binary data? Tcl coding standards do not specify the layout of packages. > > Secondly, what changes are you proposing? > > > >First, all doc files go to their debian doc dir /usr/share/doc/sourcenav. >that doesn't concern you much I guess. > >demos also go to doc dir. > >share/bitmaps, share/gui -> share/sourcenav/ > >i don't know where to put sdk, since sdk is not specific to sourcenav? >if it is specific to sourcenav, then it also goes under share/sourcenav >it looks like read only arch-independent data. > >Cheers, Source-Navigator is designed to integrated with the other gnu tools that Red Hat offers, and has a directory structure which maps that. We will keep these changes in mind, however. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mharris@opensourceadvocate.org Tue Oct 17 23:45:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 17 Oct 2000 23:45:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: On Tue, 17 Oct 2000, Eray 'exa' Ozkural wrote: >Date: Tue, 17 Oct 2000 19:36:36 +0300 (EEST) >From: Eray 'exa' Ozkural >To: sourcenav@sources.redhat.com >Content-Type: TEXT/PLAIN; charset=US-ASCII >Subject: sourcenav deb dir layout > >The following is the dir layout of sourcenav package: > >There are several problems with it. It doesn't seem to comply with >FHS or GNU coding standards. I can fix some of it, but ultimately >you should decide how it will play nicely with a system. >What is your opinion? > >-- usr > |-- bin > |-- html > | |-- progref > | `-- userguide ^^^^^^^^^^^^ I don't like those 2 dirs. They bind "progref, userguide" to sourcenav, however 40 packages could have dirs named "userguide". Perhaps: html | ----sourcenav | ----progref ----userguide No ambiguity then. Actually, /usr/share/html would be a better place IMHO, but.... ;o) TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- If you're looking for Linux books, guides, and other documentation, visit the Linux Documentation Project homepage: http://linuxdoc.org From mharris@opensourceadvocate.org Tue Oct 17 23:57:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Tue, 17 Oct 2000 23:57:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001017175902.01a3b730@pop.cygnus.com> Message-ID: On Tue, 17 Oct 2000, Syd Polk wrote: >>i don't know where to put sdk, since sdk is not specific to sourcenav? >>if it is specific to sourcenav, then it also goes under share/sourcenav >>it looks like read only arch-independent data. >> >>Cheers, > >Source-Navigator is designed to integrated with the other gnu tools that >Red Hat offers, and has a directory structure which maps that. We will keep >these changes in mind, however. The funny thing though, is Source Navigator doesn't follow Red Hat conventions of file locations either. ;o) Red Hat 7.0 of course approaches FHS compliance, and both debian and Red Hat are becoming similar in this manner. The reason I'm pointing this out is because I am trying to package Source Navigator for Red Hat right now, and having similar difficulties. ;o) The default install of a source nav build scatters files in odd locations as I found out the hard way.. ;o) I had it build for /usr/local, and after installing had numerous problems as an ordinary user on my system. Root, everything worked fine, but as a user, I had different problems running scripts, etc.. It turns out, source nav installs grep/egrep/fgrep and a slew of other utils into /usr/local/bin when built for /usr/local, and /usr/local was in the path first, so the sourcenav utils got used instead of /bin/grep, etc.. The problem is that the SourceNav grep and friends do not support all grep commandline options. It took me a few days to figure this out.. ;o) What is the reasoning behind bringing a copy of grep and other utils with sourcenav? Is it for systems not having it? If so, can I safely exclude it from the Red Hat packages? To the fellow creating the deb packages, could you post your deb build files, (not the packages, just the deb equiv of .spec files) to me privately. I'll share my .spec with you as well if you like. Perhaps we could take some of the changes that will be necessary in the build sections of our efforts and incorporate them directly into the Sourcenav makefiles, etc. and send patches in? Well, one thing at a time I suppose. ;o) Take care everyone! TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- "If it isn't source, it isn't software." -- NASA From bje@redhat.com Wed Oct 18 00:16:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 18 Oct 2000 00:16:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: What is the reasoning behind bringing a copy of grep and other utils with sourcenav? Is it for systems not having it? If so, can I safely exclude it from the Red Hat packages? The version of grep that is built by S-N includes a number of command line options that were necessary to make it work better with S-N -- for instance, it emitted progress indication to stderr which S-N used to give visual feedback of grep's progress. I believe the current version eliminates the need for a special version of grep, but I could be mistaken. I'll leave that up to one of the current team to speak about. Ben From mdejong@cygnus.com Wed Oct 18 00:27:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 18 Oct 2000 00:27:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: On Wed, 18 Oct 2000, Ben Elliston wrote: > What is the reasoning behind bringing a copy of grep and other utils > with sourcenav? Is it for systems not having it? If so, can I safely > exclude it from the Red Hat packages? > > The version of grep that is built by S-N includes a number of command line > options that were necessary to make it work better with S-N -- for instance, > it emitted progress indication to stderr which S-N used to give visual > feedback of grep's progress. > > I believe the current version eliminates the need for a special version of > grep, but I could be mistaken. I'll leave that up to one of the current > team to speak about. > > Ben Guys, I thought we had been over this before. Yes, 4.5 uses a custom grep that was forked a couple of years ago. The need for this custom grep has been removed in 5.0 because of all the problems it would create if you tried to make a RPM. I still think it is great that you gents are willing to go through the pain of creating deb/rpm package for 4.5, but I still think you should just put everything in a sn root (like /usr/sourcenav) and be done with it. You are going to have nothing but problems trying to get the 4.5 codebase to play nice with the other executables in /usr. Toasted executables: egrep fgrep grep itclsh3.0 itkwish3.0 tclsh wish Heck, just the fact that you dont want to install /usr/README.TXT is enough. Mo DeJong Red Hat Inc From mharris@opensourceadvocate.org Wed Oct 18 00:30:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 18 Oct 2000 00:30:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: On Wed, 18 Oct 2000, Ben Elliston wrote: > What is the reasoning behind bringing a copy of grep and other utils > with sourcenav? Is it for systems not having it? If so, can I safely > exclude it from the Red Hat packages? > >The version of grep that is built by S-N includes a number of command line >options that were necessary to make it work better with S-N -- for instance, >it emitted progress indication to stderr which S-N used to give visual >feedback of grep's progress. > >I believe the current version eliminates the need for a special version of >grep, but I could be mistaken. I'll leave that up to one of the current >team to speak about. When a specific program or util MUST use it's own local copy of a standard system tool, it should rename it to avoid problems. In other words, Source Nav's "grep" should be renamed: sn-grep or something similar. It broke many system scripts I have, and caused several days of "WTH is going on?" questions.. ;o) You never know how a user's path is going to look. Programs like xcdroast used to include cdrecord, mkisofs, etc.. and called them "mycdrecord", "mymkisofs" to avoid conflict. It also put them in its own private dir not in the path. Just some suggestions... Take care all. TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- If you're looking for Linux books, guides, and other documentation, visit the Linux Documentation Project homepage: http://linuxdoc.org From dave.banham@tde.alstom.com Wed Oct 18 01:11:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Wed, 18 Oct 2000 01:11:00 -0000 Subject: xref and referred by? Message-ID: I too have also experienced this frustrating problem and I don't believe it is anything to do with the fact that the xref database is still being (re-)constructed - at least there is no progress bar indication. I quite often end up using the grep tool to find the references - which it does - but in doing so I am left questioning the accuracy of the xref tool. If xref reports no referenced by information then I will try grep, but if it reports some referenced by information am I seeing the complete picture or just part of it? Regards Dave Banham Bruce Stephens on 17/10/2000 16:45:41 To: sourcenav cc: (bcc: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM) Subject: Re: xref and referred by? Lonnie L VanZandt writes: > Why is it that many times the Referenced By information is > unavailable? More often than not, I want to see who calls a function > and not what that function calls. However, I can't seem to get SN to > display the Referenced By information. xref requires that certain databases be completely constructed. The first time you run source navigator on a project, these can take some time to construct (potentially hours, for large projects and/or slow machines). There's a progress bar in the bottom right corner of any SN windows while this is going on. After the first time, SN should be faster to start up. -- Bruce Stephens Bruce.Stephens@MessagingDirect.com MessagingDirect(UK) Ltd From erayo@cs.bilkent.edu.tr Wed Oct 18 08:39:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Wed, 18 Oct 2000 08:39:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: <39EDA593.8ACFA13C@cs.bilkent.edu.tr> "Mike A. Harris" wrote: > > The funny thing though, is Source Navigator doesn't follow Red > Hat conventions of file locations either. ;o) Red Hat 7.0 of > course approaches FHS compliance, and both debian and Red Hat are > becoming similar in this manner. > Yeah, but deb packages strictly follow FHS compliance. It's a bug if they don't. :) > It turns out, source nav installs grep/egrep/fgrep and a slew of > other utils into /usr/local/bin when built for /usr/local, and > /usr/local was in the path first, so the sourcenav utils got used > instead of /bin/grep, etc.. The problem is that the SourceNav > grep and friends do not support all grep commandline options. It > took me a few days to figure this out.. ;o) > burgh, a hacked grep? oh, no. i hadn't noticed that. > To the fellow creating the deb packages, could you post your deb > build files, (not the packages, just the deb equiv of .spec > files) to me privately. I'll share my .spec with you as well if > you like. > All right, I'm approaching a real package. just a couple of more fixes, and I'm sending them to you. > Perhaps we could take some of the changes that will be necessary > in the build sections of our efforts and incorporate them > directly into the Sourcenav makefiles, etc. and send patches in? > Yep, I'm always trying to make a minimal number of changes (so I can maintain them), I'll send you the changes as well. Cheers, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From spolk@redhat.com Wed Oct 18 08:39:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 18 Oct 2000 08:39:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: <39EDC4F1.B86C087C@redhat.com> "Mike A. Harris" wrote: > > On Tue, 17 Oct 2000, Syd Polk wrote: > > >>i don't know where to put sdk, since sdk is not specific to sourcenav? > >>if it is specific to sourcenav, then it also goes under share/sourcenav > >>it looks like read only arch-independent data. > >> > >>Cheers, > > > >Source-Navigator is designed to integrated with the other gnu tools that > >Red Hat offers, and has a directory structure which maps that. We will keep > >these changes in mind, however. > > The funny thing though, is Source Navigator doesn't follow Red > Hat conventions of file locations either. ;o) Red Hat 7.0 of > course approaches FHS compliance, and both debian and Red Hat are > becoming similar in this manner. > I should have said "Cygnus conventions". The gnu tools group has very little to do with Red Hat 7. This will change over time, obviously. > The reason I'm pointing this out is because I am trying to > package Source Navigator for Red Hat right now, and having > similar difficulties. ;o) The default install of a source nav > build scatters files in odd locations as I found out the hard > way.. ;o) I had it build for /usr/local, and after installing > had numerous problems as an ordinary user on my system. Root, > everything worked fine, but as a user, I had different problems > running scripts, etc.. > > It turns out, source nav installs grep/egrep/fgrep and a slew of > other utils into /usr/local/bin when built for /usr/local, and > /usr/local was in the path first, so the sourcenav utils got used > instead of /bin/grep, etc.. The problem is that the SourceNav > grep and friends do not support all grep commandline options. It > took me a few days to figure this out.. ;o) This is why I have been resistant to RPMs at this time. We are currently in the process of removing grep from our build tree. > What is the reasoning behind bringing a copy of grep and other > utils with sourcenav? Is it for systems not having it? If so, > can I safely exclude it from the Red Hat packages? We added some "features" to grep. Without snavigator's grep, the grep window will not work. > To the fellow creating the deb packages, could you post your deb > build files, (not the packages, just the deb equiv of .spec > files) to me privately. I'll share my .spec with you as well if > you like. > > Perhaps we could take some of the changes that will be necessary > in the build sections of our efforts and incorporate them > directly into the Sourcenav makefiles, etc. and send patches in? > > Well, one thing at a time I suppose. ;o) > > Take care everyone! > TTYL > I have repeatedly stated that we will put out RPM versions of Source-Navigator after we finish the 5.0 release, which should be late this year or early next year. I know it is a long time to wait, but hopefully, these issues will have disappeared by then. From erayo@cs.bilkent.edu.tr Wed Oct 18 08:47:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Wed, 18 Oct 2000 08:47:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: <39EDA633.4ED5D45E@cs.bilkent.edu.tr> Mo DeJong wrote: > Guys, I thought we had been over this before. Yes, 4.5 > uses a custom grep that was forked a couple of years > ago. The need for this custom grep has been removed > in 5.0 because of all the problems it would create if > you tried to make a RPM. What 5.0? Isn't the last release 4.5.2? Thanks, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From mharris@opensourceadvocate.org Wed Oct 18 10:21:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 18 Oct 2000 10:21:00 -0000 Subject: sourcenav deb dir layout References: <39EDC4F1.B86C087C@redhat.com> Message-ID: On Wed, 18 Oct 2000, Syd Polk wrote: >> >Source-Navigator is designed to integrated with the other gnu tools that >> >Red Hat offers, and has a directory structure which maps that. We will keep >> >these changes in mind, however. >> >> The funny thing though, is Source Navigator doesn't follow Red >> Hat conventions of file locations either. ;o) Red Hat 7.0 of >> course approaches FHS compliance, and both debian and Red Hat are >> becoming similar in this manner. >> > >I should have said "Cygnus conventions". The gnu tools group >has very little to do with Red Hat 7. This will change over >time, obviously. Yeah, thats ok.. I realize that it takes time to make the changes. Especially with something the size of source nav. This is a big thing, and it will take some time I suspect. >> It turns out, source nav installs grep/egrep/fgrep and a slew of >> other utils into /usr/local/bin when built for /usr/local, and >> /usr/local was in the path first, so the sourcenav utils got used >> instead of /bin/grep, etc.. The problem is that the SourceNav >> grep and friends do not support all grep commandline options. It >> took me a few days to figure this out.. ;o) > >This is why I have been resistant to RPMs at this time. We are >currently in the process of removing grep from our build tree. Oh, its no biggie once it is known. I fully expected numerous problems.. It's all part of the fun of packaging anything at all really. ;o) >> What is the reasoning behind bringing a copy of grep and other >> utils with sourcenav? Is it for systems not having it? If so, >> can I safely exclude it from the Red Hat packages? > >We added some "features" to grep. Without snavigator's grep, the grep window >will not work. Ok, makes sense. >I have repeatedly stated that we will put out RPM versions of >Source-Navigator after we finish the 5.0 release, which should >be late this year or early next year. I know it is a long time >to wait, but hopefully, these issues will have disappeared by >then. Yep, thats good. Lots of people could benefit from the great stuff already there though, so it doesn't hurt for me to try to put out some RPM's right now. It'll take some time though.. I realize the scope, and am proceeding with caution anyways. I might not get it completed, but I'll certainly have fun trying! ;o) Thanks again, TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- Red Hat Linux: http://www.redhat.com Download for free: ftp://ftp.redhat.com/pub/redhat/redhat-6.2/ From spolk@redhat.com Wed Oct 18 11:16:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 18 Oct 2000 11:16:00 -0000 Subject: sourcenav deb dir layout References: <39EDA633.4ED5D45E@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001018111836.01a2f220@pop.cygnus.com> At 04:31 PM 10/18/00 +0300, Eray Ozkural wrote: >Mo DeJong wrote: > > Guys, I thought we had been over this before. Yes, 4.5 > > uses a custom grep that was forked a couple of years > > ago. The need for this custom grep has been removed > > in 5.0 because of all the problems it would create if > > you tried to make a RPM. > >What 5.0? Isn't the last release 4.5.2? > >Thanks, We are working right now on the next release, which is the 5.0 release. It is an upgrade to the incr Tcl code now in the product. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Wed Oct 18 11:16:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 18 Oct 2000 11:16:00 -0000 Subject: sourcenav deb dir layout References: <39EDA593.8ACFA13C@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001018111730.01a23a50@pop.cygnus.com> At 04:28 PM 10/18/00 +0300, Eray Ozkural wrote: >"Mike A. Harris" wrote: > > > > The funny thing though, is Source Navigator doesn't follow Red > > Hat conventions of file locations either. ;o) Red Hat 7.0 of > > course approaches FHS compliance, and both debian and Red Hat are > > becoming similar in this manner. > > > >Yeah, but deb packages strictly follow FHS compliance. It's a bug >if they don't. :) What is FHS, anyway? Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From erayo@cs.bilkent.edu.tr Wed Oct 18 12:14:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Wed, 18 Oct 2000 12:14:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001018111730.01a23a50@pop.cygnus.com> Message-ID: <39EDEC0B.6360A287@cs.bilkent.edu.tr> Syd Polk wrote: > > >Yeah, but deb packages strictly follow FHS compliance. It's a bug > >if they don't. :) > > What is FHS, anyway? > Ahem, this is something debian developers have been wrestling over for many months. :) >From /usr/share/doc/debian-policy/fhs/fhs.txt.gz Filesystem Hierarchy Standard -- Version 2.1 Filesystem Hierarchy Standard Group edited by Daniel Quinlan ABSTRACT This standard consists of a set of requirements and guidelines for file and directory placement under UNIX-like operating systems. The guidelines are intended to support interoperability of applications, system administration tools, development tools, and scripts as well as greater uniformity of documentation for these systems. April 12, 2000 Please see debian policy documents to see how it is used in debian. Just go to www.debian.org ;) Cheers, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From spolk@redhat.com Wed Oct 18 12:37:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 18 Oct 2000 12:37:00 -0000 Subject: sourcenav deb dir layout References: Message-ID: <4.2.0.58.20001018123857.01a20220@pop.cygnus.com> At 12:27 AM 10/18/00 -0700, Mo DeJong wrote: >On Wed, 18 Oct 2000, Ben Elliston wrote: > > > What is the reasoning behind bringing a copy of grep and other utils > > with sourcenav? Is it for systems not having it? If so, can I safely > > exclude it from the Red Hat packages? > > > > The version of grep that is built by S-N includes a number of command line > > options that were necessary to make it work better with S-N -- for > instance, > > it emitted progress indication to stderr which S-N used to give visual > > feedback of grep's progress. > > > > I believe the current version eliminates the need for a special version of > > grep, but I could be mistaken. I'll leave that up to one of the current > > team to speak about. > > > > Ben > >Guys, I thought we had been over this before. Yes, 4.5 >uses a custom grep that was forked a couple of years >ago. The need for this custom grep has been removed >in 5.0 because of all the problems it would create if >you tried to make a RPM. > >I still think it is great that you gents are willing to >go through the pain of creating deb/rpm package for >4.5, but I still think you should just put everything >in a sn root (like /usr/sourcenav) and be done >with it. You are going to have nothing but problems >trying to get the 4.5 codebase to play nice with >the other executables in /usr. > >Toasted executables: > >egrep >fgrep >grep >itclsh3.0 >itkwish3.0 >tclsh >wish Actually, the only ones of these that "make install" clobbers is grep and wish. We need our version of grep, and the wish is because "make install" assumes you are building a version that will be installed via "INSTALL". Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mharris@opensourceadvocate.org Wed Oct 18 13:50:00 2000 From: mharris@opensourceadvocate.org (Mike A. Harris) Date: Wed, 18 Oct 2000 13:50:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001018111730.01a23a50@pop.cygnus.com> Message-ID: On Wed, 18 Oct 2000, Syd Polk wrote: >> > The funny thing though, is Source Navigator doesn't follow Red >> > Hat conventions of file locations either. ;o) Red Hat 7.0 of >> > course approaches FHS compliance, and both debian and Red Hat are >> > becoming similar in this manner. >> > >> >>Yeah, but deb packages strictly follow FHS compliance. It's a bug >>if they don't. :) > >What is FHS, anyway? The "File Heirarchy Standard". It started out as the "Linux Filesystem Standard (FSSTND)" years ago, which had the goal of defining the meanings of the system directories such as /usr, /bin, /home, /etc, and so on. They renamed it to the FHS with hopes it would be adopted on non-Linux systems as well. Enhancements have been made to the document to solve different problems that have come up, and most distributions, including Red Hat are trying to follow the standard so that there is a common ground of where to place files, config files, libraries, data, etc.. Most distributions are FHS compliant to some degree or another, and Red Hat 7.0 seems to be quite so as I understand. The more compliant all dists are, the more interoperability there is. You can read the FHS document at: http://www.pathname.com/fhs It is one of the most important Linux standards documents there are right now IMHO, so it helps out a lot as a programmer to get as familiar with it as possible, since everyone is going to ultimately be using it. Other standards are the Linux Standard Base: http://www.linuxbase.org And also: http://www.freestandards.org Hope this helps! Take care, TTYL ---------------------------------------------------------------------- Mike A. Harris - Linux advocate - Open source advocate Computer Consultant - Capslock Consulting Copyright 2000 all rights reserved ---------------------------------------------------------------------- #[Mike A. Harris bash tip #3 - how to disable core dumps] # Put the following at the bottom of your ~/.bash_profile ulimit -c 0 From bje@redhat.com Wed Oct 18 14:32:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 18 Oct 2000 14:32:00 -0000 Subject: sourcenav deb dir layout References: <4.2.0.58.20001018123857.01a20220@pop.cygnus.com> Message-ID: Actually, the only ones of these that "make install" clobbers is grep and wish. We need our version of grep, and the wish is because "make install" assumes you are building a version that will be installed via "INSTALL". Is the intention that the INSTALL script will vanish in favour of just `make install'? Ben From DBriggs@dotcast.com Wed Oct 18 16:25:00 2000 From: DBriggs@dotcast.com (Deborah Briggs) Date: Wed, 18 Oct 2000 16:25:00 -0000 Subject: running INSTALL for Linux Message-ID: <52C41B218DE28244B071A1B96DD474F63F502C@DC-SRVR1.dotcast.com> Hi - when I try to run the INSTALL program in the Linux distribution, it complains that it can't find a $DISPLAY variable set. I'm running the installation from a terminal program that I've opened on a Windows host, what should DISPLAY be set to? TIA - Deb From echristo@cygnus.com Wed Oct 18 16:28:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Wed, 18 Oct 2000 16:28:00 -0000 Subject: running INSTALL for Linux References: <52C41B218DE28244B071A1B96DD474F63F502C@DC-SRVR1.dotcast.com> Message-ID: <39EE347C.435F623A@cygnus.com> Deborah Briggs wrote: > > Hi - when I try to run the INSTALL program in the Linux distribution, it > complains that it can't find a $DISPLAY variable set. I'm running the > installation from a terminal program that I've opened on a Windows host, > what should DISPLAY be set to? TIA - > Unless you have an X server on your windows box you won't be able to remote display it. Is it possible for you to use another box? At any rate, you'd set display to the box that you are on, e.g. (under csh) setenv DISPLAY twigboy.cygnus.com:0 (under sh) export DISPLAY=twigboy.cygnus.com:0 -eric From dave.banham@tde.alstom.com Thu Oct 19 03:01:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Thu, 19 Oct 2000 03:01:00 -0000 Subject: xref and referred by? Message-ID: My project is 100% (embedded) C code and I still have problems with xref. I hope that when the C++ parser is fixed that the C parsing capabilities will be improved too. BTW, I notice that sometimes when I right click on a variable in the editor window and select Find Implementation of 'xyz' that nothing happens, but when I select Find Declaration of 'xyz' SN puts up a window asking me choose one of two files to view. The first file is always a .h file and when selected it shows a line with the variable declared extern. Whilst the second file is the .c and shows the line where the variable is implemented (or is that declared?) If I then click on the variable's declared type (which is a typedef'd type), Find Implementation does nothing and Find Declaration switches (correctly) to the file containing the typedef statement for the type. If I now repeat this for the type name (in the typedef statement), Find Implementation does nothing and Find Declaration highlights the type name. Using Xref on the type name, shows that xref doesn't know where the type is used either. Oddly enough, Find Implementation and Find Declaration work fine for function names, as does Xref. Is the cross-reference information broken for variables and types or was it never intended to cover variables and types? (If so why not?) Regards Dave Banham "Berek" on 18/10/2000 16:00:37 Please respond to "Berek" To: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM@GA cc: Subject: Re: xref and referred by? The xref seems to be pretty good with std C; however, it's not reliable with C++, and no, you can't rely on it even if it's reporting some references. Until they eliminate the C++ parsing problems, you'll have to keep verifying xrefs with grep. ----- Original Message ----- From: To: "sourcenav" Sent: Wednesday, October 18, 2000 04:02 Subject: Re: xref and referred by? > > > I too have also experienced this frustrating problem and I don't believe it is > anything to do with the fact that the xref database is still being > (re-)constructed - at least there is no progress bar indication. > > I quite often end up using the grep tool to find the references - which it does > - but in doing so I am left questioning the accuracy of the xref tool. If xref > reports no referenced by information then I will try grep, but if it reports > some referenced by information am I seeing the complete picture or just part of > it? > > Regards > Dave Banham > > > > > > Bruce Stephens on 17/10/2000 16:45:41 > > To: sourcenav > cc: (bcc: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM) > Subject: Re: xref and referred by? > > > > Lonnie L VanZandt writes: > > > Why is it that many times the Referenced By information is > > unavailable? More often than not, I want to see who calls a function > > and not what that function calls. However, I can't seem to get SN to > > display the Referenced By information. > > xref requires that certain databases be completely constructed. The > first time you run source navigator on a project, these can take some > time to construct (potentially hours, for large projects and/or slow > machines). There's a progress bar in the bottom right corner of any > SN windows while this is going on. > > After the first time, SN should be faster to start up. > > -- > Bruce Stephens Bruce.Stephens@MessagingDirect.com > MessagingDirect(UK) Ltd > > > > > > > From berek@usa.net Thu Oct 19 07:58:00 2000 From: berek@usa.net (Berek) Date: Thu, 19 Oct 2000 07:58:00 -0000 Subject: xref and referred by? References: Message-ID: <001f01c039dc$fe98d0b0$216410ac@concord.com> I've experienced the same problem frequently (declaration vs implementation). ----- Original Message ----- From: To: Sent: Thursday, October 19, 2000 04:19 Subject: Re: xref and referred by? > > > My project is 100% (embedded) C code and I still have problems with xref. I hope > that when the C++ parser is fixed that the C parsing capabilities will be > improved too. > > BTW, I notice that sometimes when I right click on a variable in the editor > window and select Find Implementation of 'xyz' that nothing happens, but when I > select Find Declaration of 'xyz' SN puts up a window asking me choose one of two > files to view. The first file is always a .h file and when selected it shows a > line with the variable declared extern. Whilst the second file is the .c and > shows the line where the variable is implemented (or is that declared?) If I > then click on the variable's declared type (which is a typedef'd type), Find > Implementation does nothing and Find Declaration switches (correctly) to the > file containing the typedef statement for the type. If I now repeat this for the > type name (in the typedef statement), Find Implementation does nothing and Find > Declaration highlights the type name. Using Xref on the type name, shows that > xref doesn't know where the type is used either. > > Oddly enough, Find Implementation and Find Declaration work fine for function > names, as does Xref. Is the cross-reference information broken for variables and > types or was it never intended to cover variables and types? (If so why not?) > > Regards > Dave Banham > > > > > > "Berek" on 18/10/2000 16:00:37 > > Please respond to "Berek" > > To: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM@GA > cc: > Subject: Re: xref and referred by? > > > > The xref seems to be pretty good with std C; however, it's not reliable with > C++, and no, you can't rely on it even if it's reporting some references. > Until they eliminate the C++ parsing problems, you'll have to keep verifying > xrefs with grep. > > ----- Original Message ----- > From: > To: "sourcenav" > Sent: Wednesday, October 18, 2000 04:02 > Subject: Re: xref and referred by? > > > > > > > > I too have also experienced this frustrating problem and I don't believe > it is > > anything to do with the fact that the xref database is still being > > (re-)constructed - at least there is no progress bar indication. > > > > I quite often end up using the grep tool to find the references - which it > does > > - but in doing so I am left questioning the accuracy of the xref tool. If > xref > > reports no referenced by information then I will try grep, but if it > reports > > some referenced by information am I seeing the complete picture or just > part of > > it? > > > > Regards > > Dave Banham > > > > > > > > > > > > Bruce Stephens on 17/10/2000 16:45:41 > > > > To: sourcenav > > cc: (bcc: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM) > > Subject: Re: xref and referred by? > > > > > > > > Lonnie L VanZandt writes: > > > > > Why is it that many times the Referenced By information is > > > unavailable? More often than not, I want to see who calls a function > > > and not what that function calls. However, I can't seem to get SN to > > > display the Referenced By information. > > > > xref requires that certain databases be completely constructed. The > > first time you run source navigator on a project, these can take some > > time to construct (potentially hours, for large projects and/or slow > > machines). There's a progress bar in the bottom right corner of any > > SN windows while this is going on. > > > > After the first time, SN should be faster to start up. > > > > -- > > Bruce Stephens Bruce.Stephens@MessagingDirect.com > > MessagingDirect(UK) Ltd > > > > > > > > > > > > > > > > > > > > > > > From DBriggs@dotcast.com Thu Oct 19 08:52:00 2000 From: DBriggs@dotcast.com (Deborah Briggs) Date: Thu, 19 Oct 2000 08:52:00 -0000 Subject: running INSTALL for Linux Message-ID: <52C41B218DE28244B071A1B96DD474F63F502E@DC-SRVR1.dotcast.com> Thanks; now I've got the DISPLAY variable set correctly. and the installation appears normal on my X window. But with the Windows X server that I'm using (from MicroImages), when I invoke SourceNavigator and select "New Project", I'm getting an error that the vendor says is due to the following: A: This error occurs when the UNIX box that MI/X is running against attempts to pass an image that is at a greater bit depth than the display of the local machine is set to. For example, if you are running MI/X on a Windows machine set to display 256 colors (8-bit) and the program you are running tries to pass a 16 bit image, you'll get this error. To correct this, set the parameters of the program you are running to match the resolution of the local machine, or change the resolution of the local machine to match the bit depth of the images that the UNIX box is trying to pass. Since the windows setting is at it's highest level (32 bit TrueColor), I can't raise it. I see that source navigator supports a -colormap parameter, can this be used to fix the problem? Alternatively, if you know of a Windows X server that works with your app, please let me know. Thanks- Deb -----Original Message----- From: Eric Christopher [ mailto:echristo@cygnus.com ] Sent: Wednesday, October 18, 2000 4:39 PM To: Deborah Briggs Cc: 'sourcenav@sources.redhat.com' Subject: Re: running INSTALL for Linux Deborah Briggs wrote: > > Hi - when I try to run the INSTALL program in the Linux distribution, it > complains that it can't find a $DISPLAY variable set. I'm running the > installation from a terminal program that I've opened on a Windows host, > what should DISPLAY be set to? TIA - > Unless you have an X server on your windows box you won't be able to remote display it. Is it possible for you to use another box? At any rate, you'd set display to the box that you are on, e.g. (under csh) setenv DISPLAY twigboy.cygnus.com:0 (under sh) export DISPLAY=twigboy.cygnus.com:0 -eric From echristo@cygnus.com Thu Oct 19 09:57:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Thu, 19 Oct 2000 09:57:00 -0000 Subject: running INSTALL for Linux References: <52C41B218DE28244B071A1B96DD474F63F502E@DC-SRVR1.dotcast.com> Message-ID: <39EF2AA1.BB02B6C0@cygnus.com> > Since the windows setting is at it's highest level (32 bit TrueColor), I > can't raise it. I see that source navigator supports a -colormap parameter, > can this be used to fix the problem? > MI/X only supports 256 colors. It is a little weird because a few years ago, SN only supported 256 colors too. This was the last time I tried this little trick. Anyhow, I'm not sure about the colormap parameter - someone that works a little more frequently on SN will need to pipe up about it. Have you thought about compiling it to just run on windows? Or just running it on a unix system of some sort? > Alternatively, if you know of a Windows X server that works with your app, > please let me know. > I believe eXceed (from Hummingbird) is a little more reliable, however, it is also quite a bit more expensive if that may be an issue for you. -eric From tshead@k-3d.com Thu Oct 19 09:57:00 2000 From: tshead@k-3d.com (Timothy M. Shead) Date: Thu, 19 Oct 2000 09:57:00 -0000 Subject: running INSTALL for Linux References: <52C41B218DE28244B071A1B96DD474F63F502E@DC-SRVR1.dotcast.com> Message-ID: <39EF27E5.8020904@k-3d.com> Deborah Briggs wrote: > Alternatively, if you know of a Windows X server that works with your app, > please let me know. I've used SN successfully with XWin32, from www.xwin32.com. It seems to be much more actively developed than MI/X. Regards, Timothy M. Shead tshead@k-3d.com From mdejong@cygnus.com Thu Oct 19 22:08:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Thu, 19 Oct 2000 22:08:00 -0000 Subject: xref and referred by? References: Message-ID: On Thu, 19 Oct 2000 dave.banham@tde.alstom.com wrote: > My project is 100% (embedded) C code and I still have problems with xref. I hope > that when the C++ parser is fixed that the C parsing capabilities will be > improved too. > > BTW, I notice that sometimes when I right click on a variable in the editor > window and select Find Implementation of 'xyz' that nothing happens, but when I > select Find Declaration of 'xyz' SN puts up a window asking me choose one of two > files to view. The first file is always a .h file and when selected it shows a > line with the variable declared extern. Whilst the second file is the .c and > shows the line where the variable is implemented (or is that declared?) If I > then click on the variable's declared type (which is a typedef'd type), Find > Implementation does nothing and Find Declaration switches (correctly) to the > file containing the typedef statement for the type. If I now repeat this for the > type name (in the typedef statement), Find Implementation does nothing and Find > Declaration highlights the type name. Using Xref on the type name, shows that > xref doesn't know where the type is used either. > > Oddly enough, Find Implementation and Find Declaration work fine for function > names, as does Xref. Is the cross-reference information broken for variables and > types or was it never intended to cover variables and types? (If so why not?) > > Regards > Dave Banham If you could create a small test case that reproduces this error, that would really help. Without a way to reproduce the error, we are not going to be able to fix it. Oh, and please don't include things like, "download mozilla and then do this ...". That does not help, we need a test case with one or two functions that shows the error. Mo DeJong Red Hat Inc From zerbst@tu-harburg.de Fri Oct 20 00:04:00 2000 From: zerbst@tu-harburg.de (Carsten Zerbst) Date: Fri, 20 Oct 2000 00:04:00 -0000 Subject: Documentation ? Message-ID: <39EFE206.929B8438@tu-harburg.de> Hello everybody, I'd like to have a look at the sn. Is the documentation available in a printable format too ? Bye, Carsten -- Dipl. Ing. Carsten Zerbst | Express your needs for the | 2. European Tcl Conference 2001 ! zerbst@tu-harburg.de | http://www.tu-harburg.de/~skfcz | http://www.tu-harburg.de/skf/tcltk From emthornber@iee.org Fri Oct 20 00:25:00 2000 From: emthornber@iee.org (Mark Thornber) Date: Fri, 20 Oct 2000 00:25:00 -0000 Subject: running INSTALL for Linux References: <52C41B218DE28244B071A1B96DD474F63F502E@DC-SRVR1.dotcast.com> Message-ID: <39EFF0AC.BD6D4C8E@iee.org> Deborah, > Alternatively, if you know of a Windows X server that works with your app, > please let me know. Have you tried VNC ( http://www.uk.research.att.com/vnc/ ) ? It's free and works very well. --Mark Thornber From mdejong@cygnus.com Fri Oct 20 00:48:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Fri, 20 Oct 2000 00:48:00 -0000 Subject: running INSTALL for Linux References: <39EFF0AC.BD6D4C8E@iee.org> Message-ID: On Fri, 20 Oct 2000, Mark Thornber wrote: > Deborah, > > > Alternatively, if you know of a Windows X server that works with your app, > > please let me know. > > Have you tried VNC ( http://www.uk.research.att.com/vnc/ ) ? > It's free and works very well. > > --Mark Thornber Say, it would be kind of neat if you used VNC to display a remote Source-Navigator session from a Linux box that also had samba on it. You could then load up a cross compiling gcc and be able to compile a windows application on the Linux box using the IDE, and then run it off the network mounted SMB filesystem. That would provide a really easy way to turn a bunch of Windows boxes into developer workstations without having to buy a bunch of copies of VC++. I think I am going to have to try that here in the office. Mo DeJong Red Hat Inc From mthomas@makesys.com Fri Oct 20 08:35:00 2000 From: mthomas@makesys.com (Michael Thomas) Date: Fri, 20 Oct 2000 08:35:00 -0000 Subject: How to use Source Navigator with Clearcase version 4? Message-ID: <39F06653.2BC113FD@makesys.com> Hi, None of the basic clearcase commands work using Source Navigator on Windows, with ClearCase version 4 on Windows NT. I don't know what properties to set in "...Cygnus\SN452\share\etc\sn_prop.cfg" or what to change to make it work. Please advise. Regards, Mike Thomas From mdejong@cygnus.com Fri Oct 20 11:23:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Fri, 20 Oct 2000 11:23:00 -0000 Subject: How to use Source Navigator with Clearcase version 4? References: <39F06653.2BC113FD@makesys.com> Message-ID: On Fri, 20 Oct 2000, Michael Thomas wrote: > Hi, > > None of the basic clearcase commands work using Source Navigator on > Windows, with ClearCase version 4 on Windows NT. > I don't know what properties to set in > "...Cygnus\SN452\share\etc\sn_prop.cfg" or what to change to make it > work. Sounds like you are going to need to get it working and send in a patch. We don't have a copy of Clearcase here in the office to test with, so we can't really help you. Mo DeJong Red Hat Inc From bstell@netscape.com Fri Oct 20 19:58:00 2000 From: bstell@netscape.com (Brian Stell) Date: Fri, 20 Oct 2000 19:58:00 -0000 Subject: insight fails after I install source-nav Message-ID: <39F10674.49630BA2@netscape.com> I have been using insight for a few weeks now (2000.10.04 tar) on Linux 86 and thought I'd try Source-Navigator. After I installed Source-Nav insight would run and complained: /usr/local/bin/gdb ./mozilla-bin Can't find a usable main.tcl in the following directories: /usr/local/share/gdbtcl /usr/local/gdbtcl /usr/local/gdb/gdbtk/library /usr/local/gdb1.0/gdbtk/library /usr/local/share/gdbtcl/main.tcl: unknown namespace in import pattern "debug::*" unknown namespace in import pattern "debug::*" while executing "namespace import debug::*" (file "/usr/local/share/gdbtcl/main.tcl" line 51) invoked from within "source /usr/local/share/gdbtcl/main.tcl" ("uplevel" body line 1) invoked from within "uplevel \#0 [list source $file]" This probably means that gdb wasn't installed properly. while executing "error $msg" (procedure "tcl_findLibrary" line 84) invoked from within "tcl_findLibrary gdb 1.0 {} main.tcl GDBTK_LIBRARY GDBTK_LIBRARY gdbtk/library gdbtcl {} $debug_startup" (procedure "gdbtk_find_main" line 11) invoked from within "gdbtk_find_main" [accipiter]$ : From mhthomas@yahoo.com Sat Oct 21 12:19:00 2000 From: mhthomas@yahoo.com (Michael Thomas) Date: Sat, 21 Oct 2000 12:19:00 -0000 Subject: Build SN452 on Windows NT4 Message-ID: <20001021191906.27861.qmail@web702.mail.yahoo.com> Hi, Please help: "If you must rebuild Source-Navigator on Windows, please send mail to sourcenav@sourceware.cygnus.com, and we will attempt to post rational build instructions." I'm trying to build SN452 on Windows NT4 so I can patch the integration with Clearcase 4.x. I have all the PC's, OS's, tools, debuggers, etc., except the build environment for Windows NT. I can stop by your offices in Durham if you can show me. Or, can you post something? I live at: Michael Thomas 3402 Ruffin St #101 Raleigh, NC 27607 (919) 5588291 Please advise. Regards, Mike Thomas __________________________________________________ Do You Yahoo!? Yahoo! Messenger - Talk while you surf! It's FREE. http://im.yahoo.com/ From mdejong@cygnus.com Sat Oct 21 13:45:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sat, 21 Oct 2000 13:45:00 -0000 Subject: Build SN452 on Windows NT4 References: <20001021191906.27861.qmail@web702.mail.yahoo.com> Message-ID: > Hi, > > Please help: > "If you must rebuild Source-Navigator on Windows, > please send mail to sourcenav@sourceware.cygnus.com, > and we will attempt to post rational build > instructions." > > I'm trying to build SN452 on Windows NT4 so I can > patch the integration with Clearcase 4.x. I have all > the PC's, OS's, tools, debuggers, etc., except the > build environment for Windows NT. > > I can stop by your offices in Durham if you can show > me. Or, can you post something? See: http://sources.redhat.com/ml/sourcenav/2000-q3/msg00375.html Mo DeJong Red Hat Inc From bje@cygnus.com Sat Oct 21 17:13:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Sat, 21 Oct 2000 17:13:00 -0000 Subject: Build SN452 on Windows NT4 References: <20001021191906.27861.qmail@web702.mail.yahoo.com> Message-ID: I'm trying to build SN452 on Windows NT4 so I can patch the integration with Clearcase 4.x. I have all the PC's, OS's, tools, debuggers, etc., except the build environment for Windows NT. It's not necessary to rebuild to fix integration with version control systems. That is all done from etc/sn_prop.cfg.in. Ben From mhthomas@yahoo.com Sun Oct 22 06:53:00 2000 From: mhthomas@yahoo.com (Michael Thomas) Date: Sun, 22 Oct 2000 06:53:00 -0000 Subject: Build SN452 on Windows NT4 Message-ID: <20001022135347.20817.qmail@web701.mail.yahoo.com> Hi, I tried that on the Windows NT configuration file. Apparently, I might not know what to do. Attached is my attempt for ClearCase version 4.x, w/dynamic views, and no CCase project files. There are still many error messages, warnings, problems I could not resolve just from the configuration file. I'd like the opportunity to clean up those issues, before asking our other developers to use the tool. I don't understand why "checkout-individual-to-stdout" seems to be used for file differences? At least that is the line that runs when I select Difference in SN. Am I missing something? Regards, Mike Thomas --- Ben Elliston wrote: > I'm trying to build SN452 on Windows NT4 so I can > patch the > integration with Clearcase 4.x. I have all the > PC's, OS's, tools, > debuggers, etc., except the build environment for > Windows NT. > > It's not necessary to rebuild to fix integration > with version control > systems. That is all done from etc/sn_prop.cfg.in. > > Ben > __________________________________________________ Do You Yahoo!? Yahoo! Messenger - Talk while you surf! It's FREE. http://im.yahoo.com/ ############################################################################### # # sn_prop.cfg # # Copyright (C) 1997 Cygnus Solutions, Inc. # # Description: # This is a Tcl source file that is read by Source-Navigator at startup. By # modifying this file, it is possible to reconfigure Source-Navigator. # # The most common reason for modifying this file is to integrate revision # control software packages or new language parsers into Source-Navigator. # Refer to the Programmer's Reference Guide for more detailed information. # ############################################################################### ############################################################################### # # Version control system configuration # # To integrate a new third-party version control system into Source-Navigator, # use the Tcl command `sn_add_version_control_system'. Refer to the # Programmer's Reference Guide for further details. # ############################################################################### # GNU Revision Control System (RCS) sn_add_version_control_system rcs -default yes \ -checkin "ci -u" \ -checkin-comment-via stdin \ -checkin-exclusive "ci -l" \ -checkout "co -f" \ -checkout-exclusive "co -f -l" \ -checkout-individual "co -f -r" \ -checkout-individual-to-stdout "co -p" \ -checkout-with-lock yes \ -delete-revision "rcs -o" \ -discard "unco" \ -history "rlog" \ -history-pattern { {"^-----" "^====="} } \ -history-individual "rlog -r" \ -history-individual-pattern { {"^-----" "^====="} } \ -history-replacements { {"[ \t]+" " "} } \ -ignore-dirs RCS \ -lock "rcs -l" \ -lock-individual "rcs -l" \ -revision-number-pattern "^revision (\[0-9.\]+)" \ -symbolic-tags-pattern { {"^symbolic names" "^keyword"} } \ -symbolic-tags-replacements { {"\t" ""} } \ -unlock "rcs -u" \ -unlock-individual "rcs -u" # Concurrent Versions System (CVS) # # This configuration provides lazy locking as described in the CVS # documentation. Files are checked out with read/write permissions. CVS is # multiuser version control system, so locking is not necessary and is # therefore not implemented. sn_add_version_control_system cvs \ -checkin "cvs commit -F " \ -checkin-comment-via file \ -checkout "cvs update -r" \ -checkout-individual "cvs update -r" \ -checkout-individual-to-stdout "cvs update -p -r" \ -checkout-with-lock no \ -delete-revision "cvs admin -o" \ -discard {} \ -history "cvs log" \ -history-pattern { {"^-----" "^====="} } \ -history-individual "cvs log -r" \ -history-individual-pattern { {"^-----" "^====="} } \ -history-replacements { {"[ \t]+" " "} } \ -ignore-dirs CVS \ -revision-number-pattern "^revision (\[0-9.\]+)" \ -symbolic-tags-pattern { {"^symbolic names" "^keyword"} } \ -symbolic-tags-replacements { {"\t" ""} } # AT&T Source Code Control System (SCCS) # # The configuration uses the "sccs" command front-end to the SCCS utilities. # On some systems, failing to do some may cause conflicts with other # similiarly named utilities in the path (e.g. "get"). Using the front-end # also insulates Source-Navigator from implementation details and potential # changes to a vendor's implementation of SCCS. sn_add_version_control_system sccs \ -checkin "sccs delget" \ -checkin-comment-via stdin \ -checkout "sccs get -r" \ -checkout-exclusive "sccs edit" \ -checkout-individual "sccs get -r" \ -checkout-individual-to-stdout "sccs get -p -r" \ -checkout-with-lock yes \ -delete-revision "sccs rmdel -r" \ -discard "sccs unget" \ -history "sccs prs" \ -history-pattern { {"^D" end} } \ -history-individual "sccs prs -r" \ -history-individual-pattern { {"^D" end} } \ -ignore-dirs SCCS \ -lock "sccs get -g -e" \ -lock-individual "sccs get -g -e -r" \ -revision-number-pattern "^D (\[0-9\.\]+)" \ -unlock-individual "sccs unget -n" \ -unlock-individual "sccs unget -n -r" # Rational's ClearCase configuration management system (version 3.x) # # This configuration uses the command line driven ClearCase diff tool to # produce UNIX-compatible diff output on stdout. Then we use Source- # Navigator's built-in diff tool to view the results. sn_add_version_control_system ccase \ -checkin "cleartool checkin -cq" \ -checkin-comment-via stdin \ -checkout "cleartool checkout -nc" \ -checkout-exclusive "cleartool checkout -nc -reserved" \ -checkout-individual "cleartool -nc checkout" \ -checkout-individual-to-stdout "cleartool checkout -o -" \ -checkout-with-lock no \ -delete-revision "cleartool rmver -force" \ -diff-command "cleartool diff -diff_format" \ -diff-ignore-case "" \ -diff-ignore-whitespace "-opt -blank_ignore" \ -discard "cleartool uncheckout" \ -history "cleartool lshistory" \ -history-individual "cleartool lshistory" \ -ignore-dirs archives \ -lock "cleartool reserve -nc" \ -lock-individual "cleartool reserve -nc" \ -revision-number-pattern {create version "(.*)"} \ -symbolic-tags-pattern {create version .*\((.*)\)} \ -title ClearCase \ -unlock "cleartool unreserve -nc" \ -unlock-individual "cleartool unreserve -nc" # Rational's ClearCase configuration management system (version 4.x) # # This configuration uses the command line driven ClearCase diff tool to # produce UNIX-compatible diff output on stdout. Then we use Source- # Navigator's built-in diff tool to view the results. # # ClearCase version 4.x is unsupported. But, these settings support most # SN452 version control features on ClearCase 4.x on Window NT4. Ignore the # warning messages, then you can prove it for yourself. Copy this file # over the original C:\Program Files\Cygnus\SN452\share\etc\sn_prop.cfg. # Set Project Properties|Version Control|Version control system to # ClearCaseHack. The following feature work (ignore most warning messages): # Tools|Revision Control|Check Out # Tools|Revision Control|Check In # Tools|Revision Control|Discard Changes # Tools|Revision Control|Lock # Tools|Revision Control|Unlock # Tools|Revision Control|Compare Revisions ** Note 1 # Tools|Revision Control|Revision Control Editor ** Note 2 # # ** Note 1 - Always select the top revision on right side, wrapping {}'s # ** Note 2 - The Revision Control Editor, Delete Version does not work. # We disabled rmver using lshistory command, producing fewer warnings. # # One helpful trick is to always select a revision, forcing wrapped {}'s. # Also, remove the "-pre" from checkout-individual-to-stdout if you only # want to diff two versions, because its setup for three panels. sn_add_version_control_system ccase2 \ -checkin "cleartool checkin -cq" \ -checkin-comment-via stdin \ -checkout "cleartool checkout -res -nwa -nc " \ -checkout-exclusive "cleartool checkout -res -nwa -nc " \ -checkout-individual "cleartool checkout -res -nc " \ -checkout-individual-to-stdout "cleartool diff -g -pre " \ -checkout-with-lock no \ -delete-revision "cleartool lshistory " \ -diff-command "cleartool diff -diff_format" \ -diff-ignore-case "" \ -diff-ignore-whitespace "-opt -blank_ignore" \ -discard "cleartool uncheckout" \ -history "cleartool lshistory" \ -history-individual "cleartool lshistory" \ -ignore-dirs archives \ -lock "cleartool reserve -nc " \ -lock-individual "cleartool reserve -nc " \ -revision-number-pattern {create version "(.*)"} \ -symbolic-tags-pattern {create version .*\((.*)\)} \ -title ClearCaseHack \ -unlock "cleartool unreserve -nc " \ -unlock-individual "cleartool unreserve -nc " ############################################################################### # # Parser configuration # # To add a parser, insert additional procedure calls to `sn_add_parser' # here. Refer to the Programmer's Reference Guide for further details. # Possible arguments to the function: # -suffix :suffixes for the supported language files # -brow_cmd :parser program # -high_cmd :highlighting program # -editor :external editor # -brow_switch :parser switches # -high_switch :switches used by highlighting program # -case (1, 0) :turn case sensitive on/of # -macros (-m, ""):parser understand macros # ############################################################################### # FORTRAN sn_add_parser fortran \ -suffix {*.f *.for *.FOR} \ -brow_cmd $sn_path(bindir)/fbrowser \ -high_cmd $sn_path(bindir)/fbrowser \ -brow_switch "" \ -high_switch "-h" \ -case 1 # ELF #sn_add_parser elf \ -suffix {*.elf *.am} \ -brow_cmd $sn_path(bindir)/ebrowser \ -high_cmd $sn_path(bindir)/ebrowser \ -high_switch "-h" # ELF (using ApplixWare macro editor) #sn_add_parser elf \ -suffix {*.elf *.am} \ -brow_cmd $sn_path(bindir)/ebrowser \ -high_cmd $sn_path(bindir)/ebrowser \ -brow_switch "$asterx_home_dir/applix -macro sn_imporp -pass" \ -high_switch "-h" # COBOL sn_add_parser cobol \ -suffix {*.cbl *.cob} \ -brow_cmd $sn_path(bindir)/obrowser \ -high_cmd $sn_path(bindir)/obrowser \ -high_switch "-h" \ -case 0 # C++ if {$tcl_platform(platform) == "windows"} { #windows file dialog box doesn't understand with '[]' grouped extensions set cpp_ext {*.h *.H *.c *.C *.cxx *.hxx *.cpp *.hpp *.cc *.hh *.CXX *.HXX *.CPP *.HPP *.CC *.HH} } else { set cpp_ext {*.[hHcC] *.[ch]xx *.[ch]pp *.cc *.hh *.[CH]XX *.[CH]PP *.CC *.HH} } sn_add_parser c++ \ -suffix $cpp_ext \ -brow_cmd $sn_path(bindir)/cbrowser \ -high_cmd $sn_path(bindir)/cbrowser \ -high_switch "-h" \ -macros "-m" # C #sn_add_parser c \ -suffix {*.ec *.ep *.sc *.cpc} \ -brow_cmd $sn_path(bindir)/cbrowser \ -high_cmd $sn_path(bindir)/cbrowser \ -brow_switch "-C" \ -high_switch "-C -h" # Java sn_add_parser java \ -suffix {*.java} \ -brow_cmd $sn_path(bindir)/jbrowser \ -high_cmd $sn_path(bindir)/jbrowser \ -high_switch "-h" # Tcl/Tk and [incr Tcl] sn_add_parser tcl \ -suffix {*.tcl *.itcl *.itk *.tk}\ -brow_cmd $sn_path(bindir)/tbrowser \ -high_cmd $sn_path(bindir)/tbrowser \ -high_switch "-h" # PowerPC 601 assembly language sn_add_parser asm \ -suffix {*.asm *.s *.S}\ -brow_cmd $sn_path(bindir)/abrowser \ -high_cmd $sn_path(bindir)/abrowser # CCITT High Level Language (CHILL) #sn_add_parser chill \ -suffix {*.ch} \ -brow_cmd $sn_path(bindir)/chbrowser \ -high_cmd $sn_path(bindir)/chbrowser # Python sn_add_parser python \ -suffix {*.py *.pyw} \ -brow_cmd $sn_path(bindir)/pybrowser \ -high_cmd $sn_path(bindir)/pybrowser # Files without a parser if {$tcl_platform(platform) == "windows"} { #windows file dialog box doesn't understand with '[]' grouped extensions set other_ext {*.l *.y *Makefile *makefile *Iakefile} } else { set other_ext {*.[ly] *[IMm]akefile} } sn_add_parser others -suffix $other_ext # Addisional build rules # A Build rule for Yacc files. sn_add_simple_ide_rule YACC \ -suffix-list {.y.c} \ -action "$<" \ -tool bison \ -description "Process a yacc file." \ -file-type Yacc # A Build rule for Lex files sn_add_simple_ide_rule LEX \ -suffix-list {.l.c} \ -action "$<" \ -tool flex \ -description "Process a lex file." \ -file-type Lex From erayo@cs.bilkent.edu.tr Sun Oct 22 07:53:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Sun, 22 Oct 2000 07:53:00 -0000 Subject: itcl compatibility problem Message-ID: <39F2F801.310677FA@cs.bilkent.edu.tr> I've built sourcenav debian packages with the tcl/tk 8.3 as I did for insight, but it seems that because of (at least) itcl compatibility it's going to fail in runtime. Meanwhile, I'm going to have to make huge debs with tcl/tk in 'em. In debian, we're using itcl/itk 3.1 so is the next version going to be using that? Another thing, I haven't found any development snapshots on the ftp site. Is that a temporary situation? Being able to look at some development snapshots would be very helpful at this point, because I might try to port the code myself, or help do it. Thanks! -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From segg@infonet.ca Sun Oct 22 11:22:00 2000 From: segg@infonet.ca (Gilles J. Seguin) Date: Sun, 22 Oct 2000 11:22:00 -0000 Subject: Parser, interested developpers References: <39F2F801.310677FA@cs.bilkent.edu.tr> Message-ID: <39F3301B.7C47672C@infonet.ca> Eray Ozkural wrote: > > I've built sourcenav debian packages with the tcl/tk 8.3 > as I did for insight, but it seems that because of (at > least) itcl compatibility it's going to fail in runtime. > > Meanwhile, I'm going to have to make huge debs with tcl/tk > in 'em. > > In debian, we're using itcl/itk 3.1 so is the next version > going to be using that? > Another thing, I haven't found any development snapshots on the > ftp site. Is that a temporary situation? Being able to look > at some development snapshots would be very helpful at this > point, because I might try to port the code myself, or help > do it. Development version, - Where is the development CVS ? Target/commitment was end October. - My undersanding is that the target date for stable release with the tcl/tk 8.3 upgrade is March 2001. Why are deloppers out of this upgrade ? Can we be of some help ? Or why can we ?? - What are the commodity/options that are in revision/enhancement. Parsers need improvement. Parsers need new languages. I will like to know who is having interest on parser. - Documents/tools for parsing. - Can we agree on one parser, - Works that need to be done. - Peoples who can be tester. Gilles J. Seguin segg@infonet.ca From bje@cygnus.com Sun Oct 22 14:27:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Sun, 22 Oct 2000 14:27:00 -0000 Subject: Build SN452 on Windows NT4 References: <20001022135347.20817.qmail@web701.mail.yahoo.com> Message-ID: I don't understand why "checkout-individual-to-stdout" seems to be used for file differences? At least that is the line that runs when I select Difference in SN. Am I missing something? The principle is that the diff tool is sufficiently general that it simply shows the differences between two files. You've got one already -- the one you're working on. To get the other, the sought after revision is checked out to stdout and captured to a temporary file. Ben From mdejong@cygnus.com Sun Oct 22 16:17:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sun, 22 Oct 2000 16:17:00 -0000 Subject: itcl compatibility problem References: <39F2F801.310677FA@cs.bilkent.edu.tr> Message-ID: On Sun, 22 Oct 2000, Eray Ozkural wrote: > I've built sourcenav debian packages with the tcl/tk 8.3 > as I did for insight, but it seems that because of (at > least) itcl compatibility it's going to fail in runtime. > > Meanwhile, I'm going to have to make huge debs with tcl/tk > in 'em. > > In debian, we're using itcl/itk 3.1 so is the next version > going to be using that? SN 5.0 is going to be using itcl 3.0 at first. We will upgrade to itcl 3.2 later. There should not be any problems here, as itcl 3.0 and 3.1 are just about the same. > Another thing, I haven't found any development snapshots on the > ftp site. Is that a temporary situation? Being able to look > at some development snapshots would be very helpful at this > point, because I might try to port the code myself, or help > do it. We are still working on the development version. It will be put up for testing when it is ready. Mo DeJong Red Hat Inc From irox@redhat.com Sun Oct 22 17:21:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Sun, 22 Oct 2000 17:21:00 -0000 Subject: Build SN452 on Windows NT4 References: Message-ID: <39F3846A.2E873294@redhat.com> Mo DeJong wrote: > > > Hi, > > > > Please help: > > "If you must rebuild Source-Navigator on Windows, > > please send mail to sourcenav@sourceware.cygnus.com, > > and we will attempt to post rational build > > instructions." > > > > I'm trying to build SN452 on Windows NT4 so I can > > patch the integration with Clearcase 4.x. I have all > > the PC's, OS's, tools, debuggers, etc., except the > > build environment for Windows NT. > > > > I can stop by your offices in Durham if you can show > > me. Or, can you post something? > > See: > > http://sources.redhat.com/ml/sourcenav/2000-q3/msg00375.html > > Mo DeJong > Red Hat Inc In the line that reads: CC=cl CFLAGS="-0d -Z7" ../SN452-src/configure Add -DALWAYS_SHOW_SELECTION so that it reads: CC=cl CFLAGS="-0d -Z7 -DALWAYS_SHOW_SELECTION" ../SN452-src/configure This should fix the Search highlighting bug: http://sources.redhat.com/ml/sourcenav/2000-q4/msg00070.html Ian. From spolk@redhat.com Mon Oct 23 09:51:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 09:51:00 -0000 Subject: How to use Source Navigator with Clearcase version 4? References: <39F06653.2BC113FD@makesys.com> Message-ID: <4.2.0.58.20001023095140.01e98360@pop.cygnus.com> At 11:35 AM 10/20/00 -0400, Michael Thomas wrote: >Hi, > >None of the basic clearcase commands work using Source Navigator on >Windows, with ClearCase version 4 on Windows NT. >I don't know what properties to set in >"...Cygnus\SN452\share\etc\sn_prop.cfg" or what to change to make it >work. > >Please advise. > >Regards, > >Mike Thomas We have never had the resources to test this. The company we bought this software from had an expired ClearCase demo license that they tested with, and we could not get it renewed. We would be happy to accept patches. More information about the revision control system is in the Programmer's Reference Guide. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Mon Oct 23 09:52:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 09:52:00 -0000 Subject: Build SN452 on Windows NT4 References: <20001021191906.27861.qmail@web702.mail.yahoo.com> Message-ID: <4.2.0.58.20001023095315.01e99550@pop.cygnus.com> At 12:19 PM 10/21/00 -0700, Michael Thomas wrote: >Hi, > >Please help: >"If you must rebuild Source-Navigator on Windows, >please send mail to sourcenav@sourceware.cygnus.com, >and we will attempt to post rational build >instructions." > >I'm trying to build SN452 on Windows NT4 so I can >patch the integration with Clearcase 4.x. I have all >the PC's, OS's, tools, debuggers, etc., except the >build environment for Windows NT. > >I can stop by your offices in Durham if you can show >me. Or, can you post something? Well, you can stop by, but as we are in San Francisco, it would do you no good. You need to have Microsoft Visual C++ 5.0 and the latest cygwin installed on your system. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Mon Oct 23 09:54:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 09:54:00 -0000 Subject: itcl compatibility problem References: <39F2F801.310677FA@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001023095514.01e98960@pop.cygnus.com> At 05:21 PM 10/22/00 +0300, Eray Ozkural wrote: >I've built sourcenav debian packages with the tcl/tk 8.3 >as I did for insight, but it seems that because of (at >least) itcl compatibility it's going to fail in runtime. > >Meanwhile, I'm going to have to make huge debs with tcl/tk >in 'em. > >In debian, we're using itcl/itk 3.1 so is the next version >going to be using that? It's going to be using a later version, 3.2. >Another thing, I haven't found any development snapshots on the >ftp site. Is that a temporary situation? Being able to look >at some development snapshots would be very helpful at this >point, because I might try to port the code myself, or help >do it. > >Thanks! SN452.tar.gz is a source snapshot. This problem is one of the reasons we have not done our own packages. We *require* our own versions of tcl, tk, itcl, tix, libgui, and grep. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Mon Oct 23 09:55:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 09:55:00 -0000 Subject: Parser, interested developpers References: <39F2F801.310677FA@cs.bilkent.edu.tr> <39F3301B.7C47672C@infonet.ca> Message-ID: <4.2.0.58.20001023095622.01e87220@pop.cygnus.com> At 02:21 PM 10/22/00 -0400, Gilles J. Seguin wrote: >Eray Ozkural wrote: > > > > I've built sourcenav debian packages with the tcl/tk 8.3 > > as I did for insight, but it seems that because of (at > > least) itcl compatibility it's going to fail in runtime. > > > > Meanwhile, I'm going to have to make huge debs with tcl/tk > > in 'em. > > > > In debian, we're using itcl/itk 3.1 so is the next version > > going to be using that? > > > > Another thing, I haven't found any development snapshots on the > > ftp site. Is that a temporary situation? Being able to look > > at some development snapshots would be very helpful at this > > point, because I might try to port the code myself, or help > > do it. > >Development version, > >- Where is the development CVS ? > Target/commitment was end October. It won't be for a while. >- My undersanding is that the target date for stable release > with the tcl/tk 8.3 upgrade is March 2001. > Why are deloppers out of this upgrade ? > Can we be of some help ? Or why can we ?? Because we have to untangle the Cygnus development tree. >- What are the commodity/options that are in revision/enhancement. > Parsers need improvement. Parsers need new languages. > >I will like to know who is having interest on parser. >- Documents/tools for parsing. >- Can we agree on one parser, >- Works that need to be done. >- Peoples who can be tester. > >Gilles J. Seguin >segg@infonet.ca Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From berek@usa.net Mon Oct 23 10:20:00 2000 From: berek@usa.net (Berek) Date: Mon, 23 Oct 2000 10:20:00 -0000 Subject: How do you delete a Project Editor view? References: Message-ID: <001c01c03d15$8ee95dc0$216410ac@concord.com> See subject. From erayo@cs.bilkent.edu.tr Mon Oct 23 10:56:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Mon, 23 Oct 2000 10:56:00 -0000 Subject: Parser, interested developpers References: <39F2F801.310677FA@cs.bilkent.edu.tr> <4.2.0.58.20001023095622.01e87220@pop.cygnus.com> Message-ID: <39F47C38.E95CEFB2@cs.bilkent.edu.tr> Syd Polk wrote: > > Because we have to untangle the Cygnus development tree. I built insight like that and up'ped already. You can have a look at it when it comes through, or I can post diffs to insight list. I also did sourcenav the same way, but there're a lot of incompatibilities with itcl. Is there any chance I could get versions of tcl code that's been converted to itcl3.2? If that's not available I can do it on my own time, learning a bit of itcl/itk in the meantime. I'm writing this because I wouldn't like duplication of work. Thanks, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From spolk@redhat.com Mon Oct 23 11:05:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 11:05:00 -0000 Subject: Parser, interested developpers References: <39F2F801.310677FA@cs.bilkent.edu.tr> <4.2.0.58.20001023095622.01e87220@pop.cygnus.com> <39F47C38.E95CEFB2@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001023110541.01e839d0@pop.cygnus.com> At 08:58 PM 10/23/00 +0300, Eray Ozkural wrote: >Syd Polk wrote: > > > > Because we have to untangle the Cygnus development tree. > >I built insight like that and up'ped already. You can have >a look at it when it comes through, or I can post diffs to insight >list. I also did sourcenav the same way, but there're a lot >of incompatibilities with itcl. > >Is there any chance I could get versions of tcl code that's >been converted to itcl3.2? If that's not available I can do >it on my own time, learning a bit of itcl/itk in the meantime. >I'm writing this because I wouldn't like duplication of work. We have done extensive work in upgrading to itcl 3.x. We will be releasing it at some point this year. I am sorry we don't have a better answer for you, but we are unlikely to pre-release this. >Thanks, > >-- >Eray (exa) Ozkural >Comp. Sci. Dept., Bilkent University, Ankara >e-mail: erayo@cs.bilkent.edu.tr >www: http://www.cs.bilkent.edu.tr/~erayo Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Mon Oct 23 11:06:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 11:06:00 -0000 Subject: Parser, interested developpers References: <39F2F801.310677FA@cs.bilkent.edu.tr> <4.2.0.58.20001023095622.01e87220@pop.cygnus.com> <39F47C38.E95CEFB2@cs.bilkent.edu.tr> Message-ID: <4.2.0.58.20001023110708.01e89600@pop.cygnus.com> At 08:58 PM 10/23/00 +0300, Eray Ozkural wrote: >Syd Polk wrote: > > > > Because we have to untangle the Cygnus development tree. > >I built insight like that and up'ped already. You can have >a look at it when it comes through, or I can post diffs to insight >list. I also did sourcenav the same way, but there're a lot >of incompatibilities with itcl. > >Is there any chance I could get versions of tcl code that's >been converted to itcl3.2? If that's not available I can do >it on my own time, learning a bit of itcl/itk in the meantime. >I'm writing this because I wouldn't like duplication of work. > >Thanks, Insight Source-Navigator tcl 8.0.5 (sort of) tcl 8.1b1 (sort of) tk 8.0.5 (sort of) tk 8.1b1 (sort of) itcl 3.0 itcl 1.5 (sort of) This is the crux of the problem, and why we currently need our own copies of everything. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From spolk@redhat.com Mon Oct 23 13:01:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 13:01:00 -0000 Subject: Insight, Source-Navigator, Tcl/Tk, and you Message-ID: <4.2.0.58.20001023122236.01e83e50@pop.cygnus.com> There has been much discussion about Insight, Source-Navigator, Tcl/Tk and packaging on both the Insight and Source-Navigator lists recently. I thought I would lay out the state of the world to the list, and solicit feedback. Introduction Both Insight and Source-Navigator started life integrated in the old Cygnus build tree. This build tree has local versions of Tcl, Tk, Tix, incr Tcl, dejagnu, and expect. One of the primary reasons is that the Cygnus build tree is designed for a cross-development environment. This means that it can compile binaries for another OS than the one the compile is running on. So, I can build a compiler and debugger which run on linux but generate code for an arm processor. I can build them to run on Windows and generate code for an arm processor. I can generate a compiler which runs on Linux and compiles for Windows. I can even generate a compiler which runs on windows, generates code for arm, and built it on Linux. Tcl/Tk, as it is distributed, does not have the necessary changes in configure to allow this to happen. Also, previous Cygnus products (which are no longer being supported) added features to Tcl/Tk which the maintainers did not accept, for whatever reasons. So we ended up with a divergent tree. Tcl/Tk and gcc gcc is dependent on tcl because it uses dejagnu and expect to run the test suite. Tcl/Tk and gdb gdb is dependent on tcl for the same reason. Tcl/Tk and Insight or Source-Navigator This should be obvious. The mainline tcl code in our build tree is Tcl 8.0.5, Tk 8.0.5, and incr Tcl 3.0. There are many modifications to the code, particularly in the configure files and Makefiles to support automated testing and building on various platforms, including cross-compiling. Source-Navigator diverges from this because we did a Japanese release in early 1999. We had to upgrade our tcl version to some derivative of Tcl 8.1, which we did. This caused an internal divergence, but we were able to handle it. At the time, we used Tcl 8.1b1. At this time also, however, the mainline Tcl/Tk tree was using incrTcl 1.5, hacked by us to support namespaces. Source-Navigator 4.5.2 still uses incr Tcl 1.5. It was decided at the time that upgrading the rest of the Tcl versions to 8.1 would have to wait until there was more time. At some point, gdb's cvs archive was put onto sources.redhat.com. The version of tcl, tk, and incr Tcl that Insight uses was checked into this archive. So there are several goals that we have for this stuff: - We want to use one version of Tcl/Tk for everything. - We would prefer to use Tcl/Tk that is included with the user's setup if at all possible, but we have to be prepared to provide our own versions if the user does not have one or if it is the incorrect version. This is actually a customer requirement; we cannot force the customer to upgrade the version if he or she has the wrong one. - We have to be able to build and run the test suites for gcc, gdb, and Source-Navigator from one source tree. This is not trivial because regular expressions changed between Tcl 8.1 and 8.2, and we have to update the more than 100000+ tests to the new syntax. While it is expected that only 1 in 1000 tests might be affected, that still leaves 100 tests. And they have to be tested on all popular platforms, including embedded targets. What are we doing about it? - What I am doing about it is I am upgrading our build tree to be unified. - I will then submit patches to the Tcl maintainers so that their tree and ours have a minimal number of differences. - After that, I will update sources.redhat.com - After that, I will put the Source-Navigator cvs tree on sources.redhat.com. Why I will not support packages generated by the community: Neither gdb nor source-navigator has been adequately tested on any new versions of Tcl/Tk. They both were designed to encapsulate their own versions of Tcl/Tk. We are trying to change that, but it is slow. All of the open source packaging schemes I have seen rely on the distribution version of Tcl/Tk. While things may appear to work initially with these version of Tcl/Tk, I really prefer to have at least the base gdb test suites run with this version of Tcl/Tk and against our version and any differences between the two accounted for. At some point when all of this work is complete, we will be posting RPMs for Red Hat Linux, and we may post packages in other formats as well. This is a lot of work, and it is really hard to delegate out. What can you do? Submit patches back to these lists which help make the job easier. If you have managed to get things to compile and work on later versions of Tcl, try to make it work on the distribution version of Tcl and the one we ship simultaneously, with #ifdef in C code, or comparisons to the tcl_version variable in Tcl code. Once the unification is done, we can remove the support for older versions. I am sorry this is so long-winded, but it is important, and these are real issues which have to be dealt with. I am sorry I have not adequately explained the position before now. Syd Polk Insight and Source-Navigator maintainer From gcbma@hotmail.com Mon Oct 23 15:09:00 2000 From: gcbma@hotmail.com (G Colby) Date: Mon, 23 Oct 2000 15:09:00 -0000 Subject: Problems downloading Source Navigator Message-ID: I've been attempting to get to the download area for Source Navigator all afternoon without success. Your Web site gives the impression that the software is available for public download, but a user/password pop-up window intercedes and prevents any progress. The user/password screen is not accepting anonymous login and there is no indication anywhere as to what a correct user/password might be. Please assist. _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com . Share information about yourself, create your own public profile at http://profiles.msn.com . From gcbma@hotmail.com Mon Oct 23 15:09:00 2000 From: gcbma@hotmail.com (G Colby) Date: Mon, 23 Oct 2000 15:09:00 -0000 Subject: Problems downloading Source Navigator Message-ID: I've been attempting to get to the download area for Source Navigator all afternoon without success. Your Web site gives the impression that the software is available for public download, but a user/password pop-up window intercedes and prevents any progress. The user/password screen is not accepting anonymous login and there is no indication anywhere as to what a correct user/password might be. Please assist. gcbma@hotmail.com _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com . Share information about yourself, create your own public profile at http://profiles.msn.com . From spolk@redhat.com Mon Oct 23 15:24:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 23 Oct 2000 15:24:00 -0000 Subject: Problems downloading Source Navigator References: Message-ID: <4.2.0.58.20001023152539.01e874c0@pop.cygnus.com> At 10:08 PM 10/23/00 +0000, G Colby wrote: >I've been attempting to get to the download area for Source Navigator all >afternoon without success. Your Web site gives the impression that the >software is available for public download, but a user/password pop-up window >intercedes and prevents any progress. The user/password screen is not >accepting anonymous login and there is no indication anywhere as to what a >correct user/password might be. > >Please assist. I would go to one of the mirror sites listed at http://sources.redhat.com/mirrors.html and try there. The error you are seeing is misleading; it means that the server is full and not accepting connections. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mdejong@cygnus.com Mon Oct 23 15:32:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 23 Oct 2000 15:32:00 -0000 Subject: Problems downloading Source Navigator References: Message-ID: On Mon, 23 Oct 2000, G Colby wrote: > I've been attempting to get to the download area for Source Navigator all > afternoon without success. Your Web site gives the impression that the > software is available for public download, but a user/password pop-up window > intercedes and prevents any progress. The user/password screen is not > accepting anonymous login and there is no indication anywhere as to what a > correct user/password might be. I think that is a bug in your browser, you are using IE right? When the site has too many users, it does that. Download it from a mirror: ftp://ftp.freesoftware.com/pub/sourceware/sourcenav/releases/SN452.tar.gz Mo DeJong Red Hat Inc From bje@cygnus.com Mon Oct 23 15:49:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Mon, 23 Oct 2000 15:49:00 -0000 Subject: Problems downloading Source Navigator References: Message-ID: Hi, I've been attempting to get to the download area for Source Navigator all afternoon without success. Your Web site gives the impression that the software is available for public download, but a user/password pop-up window intercedes and prevents any progress. The user/password screen is not accepting anonymous login and there is no indication anywhere as to what a correct user/password might be. I just checked it myself, visiting http://sources.redhat.com/ and following the links all the way to: ftp://sources.redhat.com/pub/sourcenav/releases/binaries/ What URLs are you trying to use? Or are you connecting to a mirror? Ben From erayo@cs.bilkent.edu.tr Tue Oct 24 06:34:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural) Date: Tue, 24 Oct 2000 06:34:00 -0000 Subject: Insight, Source-Navigator, Tcl/Tk, and you References: <4.2.0.58.20001023122236.01e83e50@pop.cygnus.com> Message-ID: <39F582AC.BCE1BD38@cs.bilkent.edu.tr> Here are the experimental sourcenav package's changelog and diffs it will compile but crash when run. -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo sourcenav (4.5.2-1) unstable; urgency=low * Initial Release. * added bogus config/tixConfig.sh * commented out guitcl.h include from tclsql.c * hyper/Makefile.am hyper/Makefile.in: changed libitcl.a path * hyper: -litcl changed to -litcl3.1 :( * hyper/recudir.c: added Tcl_SetVar2Ex for tcl8.3 * ITCL_SH is itclsh * added itcl_mkindex.tcl due to deficient itcl packaging :( * Makefile.in: added DESTDIR to FLAGS_TO_PASS * snavigator/db/Makefile.in: softcoded sdkdir to use make var * snavigator/Makefile.in: removed install from SUBDIRS * snavigator/Makefile.in: commented out install-exec-am: install-exec-local * softcoded @SN_SDK_EXEC_PREFIX@ to $(SN_SDK_EXEC_PREFIX), multiple * snavigator/snavigator/unix/snavigator.in: TCL_LIBRARY=/usr/lib/tcl8.3 * snavigator/hyper/tclsql.c: changed %s/share/tcl -> %s/lib/tcl * snavigator/hyper/tclsql.c: itcl_lib = "itcl3.1" -- Eray Ozkural Wed, 18 Oct 2000 21:32:13 +0300 Local variables: mode: debian-changelog End: -------------- next part -------------- A non-text attachment was scrubbed... Name: sourcenav_4.5.2-1.diff.gz Type: application/x-gzip Size: 354076 bytes Desc: not available URL: From berek@usa.net Tue Oct 24 10:02:00 2000 From: berek@usa.net (Berek) Date: Tue, 24 Oct 2000 10:02:00 -0000 Subject: SN Project Editor Views References: <4.2.0.58.20001023122236.01e83e50@pop.cygnus.com> <39F582AC.BCE1BD38@cs.bilkent.edu.tr> Message-ID: <006d01c03ddc$3c4ccee0$0b6410ac@concord.com> I've created a number of views in Project Editor. I'd like to delete a couple of them. Can anyone tell me how to do it? There doesn't appear to be any way to do it short of deleting the project and starting over again. I've read the documentation from cover-to-cover and found nothing. From DBriggs@dotcast.com Tue Oct 24 15:18:00 2000 From: DBriggs@dotcast.com (Deborah Briggs) Date: Tue, 24 Oct 2000 15:18:00 -0000 Subject: SourceNavigator crashes when adding directory to an existing project? Message-ID: <52C41B218DE28244B071A1B96DD474F63F5048@DC-SRVR1.dotcast.com> Hi - I'm running SourceNavigator 4.5.2 on Linux, it crashes (screen freezes under X windows, can't redraw, doesn't respond to commands but doesn't exit the app either) when I try to add a directory to an existing project but doesn't crash if I add files individually. Any advice appreciated, adding files by hand is incredibly tedious. From mdejong@cygnus.com Tue Oct 24 15:59:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 24 Oct 2000 15:59:00 -0000 Subject: SourceNavigator crashes when adding directory to an existing proj ect? References: <52C41B218DE28244B071A1B96DD474F63F5048@DC-SRVR1.dotcast.com> Message-ID: On Tue, 24 Oct 2000, Deborah Briggs wrote: > Hi - I'm running SourceNavigator 4.5.2 on Linux, it crashes (screen freezes > under X windows, can't redraw, doesn't respond to commands but doesn't exit > the app either) when I try to add a directory to an existing project but > doesn't crash if I add files individually. Any advice appreciated, adding > files by hand is incredibly tedious. Boy, I don't know what to tell you. I am also running 4.5.2 on Linux and I have no problem, I can add directories just fine. Could you recompile with the -g flag (export CFLAGS=-g ; ./configure ...) and then run sn under gdb (to do that, you just open up the install/bin/snavigator file in an editor and uncomment the bit about "run SN in GDB"). If this crashes on you, get a stack trace and send that in. Also, a complete description of your Linux box would help (hardware, distro and version, what kernel and so on). Mo DeJong Red Hat Inc From berek@usa.net Wed Oct 25 07:53:00 2000 From: berek@usa.net (Berek) Date: Wed, 25 Oct 2000 07:53:00 -0000 Subject: Deleting Project Editor Views References: Message-ID: <003d01c03e93$5983d320$0b6410ac@concord.com> Is there any way I can delete a project view? I've asked this question a couple of times in the past few days and have not seen a response from anyone. By a "project view" I mean a "view" created in Project Editor that allows you to "see" only selected parts of the overall project. Is anyone reading this email? From mdejong@cygnus.com Wed Oct 25 08:11:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 25 Oct 2000 08:11:00 -0000 Subject: Deleting Project Editor Views References: <003d01c03e93$5983d320$0b6410ac@concord.com> Message-ID: On Wed, 25 Oct 2000, Berek wrote: > Is there any way I can delete a project view? There does not seem to be. The whole interface to adding a new view is also fairly lame. This looks like something that needs to be fixed for 5.0. > I've asked this question a > couple of times in the past few days and have not seen a response from > anyone. By a "project view" I mean a "view" created in Project Editor that > allows you to "see" only selected parts of the overall project. Is anyone > reading this email? Calm down there, we will get to mail when we have a chance. Mo DeJong Red Hat Inc From Sandra.McDaniel@cisf.af.mil Wed Oct 25 08:12:00 2000 From: Sandra.McDaniel@cisf.af.mil (McDaniel Sandra C Contr SMC DET 11 /MCM) Date: Wed, 25 Oct 2000 08:12:00 -0000 Subject: Deleting Project Editor Views Message-ID: <07961E923F99D311AA7C0008C7167802014C0587@cisf2.cisf.af.mil> I'm not a member of the Source Navigator staff, but here's a suggestion. Have you tried going into the .snprj directory and deleting the files project_name.# where # is an number greater than 1. I don't think you want to delete your initial view. Sandra > -----Original Message----- > From: Berek [SMTP:berek@usa.net] > Sent: Wednesday, October 25, 2000 8:54 AM > To: sourcenav@sources.redhat.com > Subject: Deleting Project Editor Views > > Is there any way I can delete a project view? I've asked this question a > couple of times in the past few days and have not seen a response from > anyone. By a "project view" I mean a "view" created in Project Editor that > allows you to "see" only selected parts of the overall project. Is anyone > reading this email? From irox@redhat.com Wed Oct 25 13:07:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Wed, 25 Oct 2000 13:07:00 -0000 Subject: Deleting Project Editor Views References: <003d01c03e93$5983d320$0b6410ac@concord.com> Message-ID: <39F73D56.943815C7@redhat.com> Berek wrote: > > Is there any way I can delete a project view? I've asked this question a > couple of times in the past few days and have not seen a response from > anyone. By a "project view" I mean a "view" created in Project Editor that > allows you to "see" only selected parts of the overall project. Is anyone > reading this email? Unfortunately, there doesn't seem to be a way to delete a "view". you can unhide all the files in a "view" so it is the same as the default "view". Ian. From spolk@redhat.com Wed Oct 25 13:35:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 25 Oct 2000 13:35:00 -0000 Subject: Deleting Project Editor Views References: <003d01c03e93$5983d320$0b6410ac@concord.com> Message-ID: <4.2.0.58.20001025133636.00a34dc0@pop.cygnus.com> At 10:53 AM 10/25/00 -0400, Berek wrote: >Is there any way I can delete a project view? I've asked this question a >couple of times in the past few days and have not seen a response from >anyone. By a "project view" I mean a "view" created in Project Editor that >allows you to "see" only selected parts of the overall project. Is anyone >reading this email? It does not appear to be possible to delete project editor views at this time. We are planning to completely overhaul the Project Editor in the future, moving all of its functionality in a new integrated Symbol Browser. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mcousins@dialisdn.com Fri Oct 27 05:22:00 2000 From: mcousins@dialisdn.com (Merlyn (root)) Date: Fri, 27 Oct 2000 05:22:00 -0000 Subject: include file help (anyone?) Message-ID: <00102708214600.00557@MoonBase1> Perhaps someone can help me with a problem I am having using snavigator. when I'm trying to work with include files that use the form #include for example (where a directory below the include directory is used ..i.e let's say /usr/include) snavigator says (file not found) now the file is there...and /usr/include is specified in the path for the include file search. so why doesn't it work? if anyone can help I would great accpreciate it thanks merlyn From irox@redhat.com Fri Oct 27 09:15:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 27 Oct 2000 09:15:00 -0000 Subject: include file help (anyone?) References: <00102708214600.00557@MoonBase1> Message-ID: <39F9AA08.29FD4823@redhat.com> Hi, I'm assuming that you are double clicking on the include file in the include browser. (?) What OS are you running on? I've heard of this bug before but haven't managed to reproduce it on my Red Hat 6.2 box. (This is meant to problems on Windows version also.) Ian. "Merlyn(root)" wrote: > > Perhaps someone can help me with a problem I am having using snavigator. > when I'm trying to work with include files that use the form > #include for example (where a directory below the include > directory is used ..i.e let's say /usr/include) snavigator says (file not > found) now the file is there...and /usr/include is specified in the path for > the include file search. so why doesn't it work? > > if anyone can help I would great accpreciate it > > thanks > > merlyn From Fabrice_Gautier@sdesigns.com Fri Oct 27 17:54:00 2000 From: Fabrice_Gautier@sdesigns.com (Fabrice Gautier) Date: Fri, 27 Oct 2000 17:54:00 -0000 Subject: SN memory leak. Message-ID: <5AF611D83988D4118ACF00E0B10400540557EC@mr-intranet.sdesigns.net> Hi, Using SN 4.5.2 with WinNT4, I constated that memory occupation never decrease even when I close a project. I have too quit and restart quite often if I don't want to be Out of Memory. Thanks -- Fabrice Gautier fabrice_gautier@sdesigns.com From mcousins@dialisdn.com Sat Oct 28 10:26:00 2000 From: mcousins@dialisdn.com (Merlyn Cousins) Date: Sat, 28 Oct 2000 10:26:00 -0000 Subject: include file problems Message-ID: <39FB0C26.70804@dialisdn.com> Ok..first i'd like to thank you for responding. In answer to your questions...I'm a linux man I'm using SuSE 6.1 (with compiled version of snavigator) I had no errors during the compilation. everything else seems to work but the include stuff. Which really gives me problems for obvious reasons. as far as double clicking on include files..yes i do that and the Prog responds with (file not found) but it's there. It seems to do this anytime the syntax of #include is used. also a quick question for you... Why do you have to click on the include files to bring them into the project? this can be a real pain!! shouldn't there be a option of some sort that Snav does a recursive search and auto loads the #include files? I mean if these files were in the directory or a subdirectory of the current project and a new project was created, would not Snav auto bring them in? having to double click through the tree is a real time waster. So anyway thanxs again.. If i can get these points points worked out Snav along with gvim will be a GREAT aid to me in helping to Understand (reverse engineer) source code. Hi, I'm assuming that you are double clicking on the include file in the include browser. (?) What OS are you running on? I've heard of this bug before but haven't managed to reproduce it on my Red Hat 6.2 box. (This is meant to problems on Windows version also.) Ian. "Merlyn(root)" wrote: > > Perhaps someone can help me with a problem I am having using snavigator. > when I'm trying to work with include files that use the form > #include for example (where a directory below the include > directory is used ..i.e let's say /usr/include) snavigator says (file not > found) now the file is there...and /usr/include is specified in the path for > the include file search. so why doesn't it work? > > if anyone can help I would great accpreciate it > > thanks > > merlyn From irox@redhat.com Sat Oct 28 13:22:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Sat, 28 Oct 2000 13:22:00 -0000 Subject: include file problems References: <39FB0C26.70804@dialisdn.com> Message-ID: <39FB3577.B6F30744@redhat.com> Hi, Merlyn Cousins wrote: > everything else seems to work but the include stuff. > Which really gives me problems for obvious reasons. > as far as double clicking on include files..yes i do that > and the Prog responds with (file not found) but it's there. > It seems to do this anytime the syntax of #include > is used. Where are your system include files? Source-Navigator at the moment only knows to look into /usr/include, which means that it could be miss a lot of things (?). In the preferences dialog there is an "Includes" tab, this lets you add more include directories for systems include etc. (Of course this might be a bug in the parser given #inclucde "" vs. #include <>) > also a quick question for you... > Why do you have to click on the include files to bring them > into the project? > this can be a real pain!! shouldn't there be a option of > some sort that Snav does a recursive search and auto loads > the #include files? > I mean if these files were in the directory or a subdirectory > of the current project and a new project was created, would > not Snav auto bring them in? having to double click through > the tree is a real time waster. One option could be adding /usr/include or whatever to the project. Maybe we need an "include system headers" option when creating projects. Ian. > So anyway thanxs again.. > If i can get these points points worked out Snav along with gvim > will be a GREAT aid to me in helping to Understand (reverse engineer) source code. > > Hi, > > I'm assuming that you are double clicking on the include > file in the include browser. (?) > > What OS are you running on? > > I've heard of this bug before but haven't managed to > reproduce it on my Red Hat 6.2 box. (This is meant to > problems on Windows version also.) > > Ian. > > "Merlyn(root)" wrote: > > > > Perhaps someone can help me with a problem I am having using snavigator. > > when I'm trying to work with include files that use the form > > #include for example (where a directory below the include > > directory is used ..i.e let's say /usr/include) snavigator says (file not > > found) now the file is there...and /usr/include is specified in the path for > > the include file search. so why doesn't it work? > > > > if anyone can help I would great accpreciate it > > > > thanks > > > > merlyn From mcousins@dialisdn.com Sat Oct 28 14:32:00 2000 From: mcousins@dialisdn.com (Merlyn (root)) Date: Sat, 28 Oct 2000 14:32:00 -0000 Subject: include file problems References: <39FB3577.B6F30744@redhat.com> Message-ID: <00102817320000.00462@MoonBase1> Hi there....thanxs again for quick reply. my include files are in /usr/include and directories below (/usr/include) I tried changing the directory search path in INCLUDE preferences (no luck) To give you an example of exactly what is going on. consider this. one of my include files is #include this should resolve to /usr/include/sys/times.h snav with or without directory /usr/include/sys specified in preferences does not find the file...(and it's there). as far as a option to bring in the headers into the current project without the double clicking stuff... I think that would be great! Just have an option which resolves all (external yuk) (i.e. external from current project) include files. thanks alot Merlyn!! On Sat, 28 Oct 2000, you wrote: > Hi, > > Merlyn Cousins wrote: > > everything else seems to work but the include stuff. > > Which really gives me problems for obvious reasons. > > as far as double clicking on include files..yes i do that > > and the Prog responds with (file not found) but it's there. > > It seems to do this anytime the syntax of #include > > is used. > > Where are your system include files? Source-Navigator > at the moment only knows to look into /usr/include, which > means that it could be miss a lot of things (?). > In the preferences dialog there is an "Includes" tab, this > lets you add more include directories for systems include etc. > > (Of course this might be a bug in the parser given #inclucde "" > vs. #include <>) > > > also a quick question for you... > > Why do you have to click on the include files to bring them > > into the project? > > this can be a real pain!! shouldn't there be a option of > > some sort that Snav does a recursive search and auto loads > > the #include files? > > I mean if these files were in the directory or a subdirectory > > of the current project and a new project was created, would > > not Snav auto bring them in? having to double click through > > the tree is a real time waster. > > One option could be adding /usr/include or whatever to the > project. Maybe we need an "include system headers" option > when creating projects. > > Ian. > > > So anyway thanxs again.. > > If i can get these points points worked out Snav along with gvim > > will be a GREAT aid to me in helping to Understand (reverse engineer) source code. > > > > Hi, > > > > I'm assuming that you are double clicking on the include > > file in the include browser. (?) > > > > What OS are you running on? > > > > I've heard of this bug before but haven't managed to > > reproduce it on my Red Hat 6.2 box. (This is meant to > > problems on Windows version also.) > > > > Ian. > > > > "Merlyn(root)" wrote: > > > > > > Perhaps someone can help me with a problem I am having using snavigator. > > > when I'm trying to work with include files that use the form > > > #include for example (where a directory below the include > > > directory is used ..i.e let's say /usr/include) snavigator says (file not > > > found) now the file is there...and /usr/include is specified in the path for > > > the include file search. so why doesn't it work? > > > > > > if anyone can help I would great accpreciate it > > > > > > thanks > > > > > > merlyn From erayo@cs.bilkent.edu.tr Sun Oct 29 04:46:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Sun, 29 Oct 2000 04:46:00 -0000 Subject: Sourcenav color icon image Message-ID: <39FC0E3B.105AA73E@cs.bilkent.edu.tr> Is there a color icon image for sourcenav? Thanks, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From bje@cygnus.com Sun Oct 29 13:51:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Sun, 29 Oct 2000 13:51:00 -0000 Subject: Sourcenav color icon image References: <39FC0E3B.105AA73E@cs.bilkent.edu.tr> Message-ID: Is there a color icon image for sourcenav? If you're referring to the icon you see when S-N is minimised, then there isn't one. This is because Tk only allows this icon to be in X bitmap format -- thus, two colours. Ben From bje@cygnus.com Sun Oct 29 13:57:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Sun, 29 Oct 2000 13:57:00 -0000 Subject: include file problems References: <39FB0C26.70804@dialisdn.com> Message-ID: Why do you have to click on the include files to bring them into the project? this can be a real pain!! shouldn't there be a option of some sort that Snav does a recursive search and auto loads the #include files? Because often, parsing these files takes a lot of time and imports a lot of detail into your project you don't care about. It's unusual to really want to know where strcmp() is declared, since you either know, or don't really care. Don't you agree? Ben From mdejong@cygnus.com Sun Oct 29 14:16:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sun, 29 Oct 2000 14:16:00 -0000 Subject: Sourcenav color icon image References: Message-ID: On Mon, 30 Oct 2000, Ben Elliston wrote: > Is there a color icon image for sourcenav? > > If you're referring to the icon you see when S-N is minimised, then there > isn't one. This is because Tk only allows this icon to be in X bitmap > format -- thus, two colours. > > Ben This has been fixed in newer versions of Tk. You can now use an XPM instead of an XBM for the icon image. Of course, this only works with newer window managers but if it does not work it will fallback to the XBM. So, it should be possible to have a color icon image. Mo DeJong Red Hat Inc From erayo@cs.bilkent.edu.tr Sun Oct 29 15:34:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Sun, 29 Oct 2000 15:34:00 -0000 Subject: Sourcenav color icon image References: Message-ID: <39FCA612.57A96B5B@cs.bilkent.edu.tr> Mo DeJong wrote: > > This has been fixed in newer versions of Tk. You can now > use an XPM instead of an XBM for the icon image. Of course, > this only works with newer window managers but if it does > not work it will fallback to the XBM. So, it should be > possible to have a color icon image. > Thanks a lot, but could you please tell me which artwork I'm supposed to use for source navigator icon? ;) I don't trust my drawing skills because I've been programming for too long and I might have lost something... :) I know this is a Red Hat project, but using the red hat logo, be it monochrome or true-color, wouldn't be very informative for the common user. ;) Do you want me to draw an icon for you? [ you might have to bear the results! ] Or there's the artwork at sourcenav home page and it's already in the debian package: orion:exa$ see /usr/lib/sourcenav/share/bitmaps/hourglass.gif shows me a handsome hourglass. I think I should use that as the icon rather than drawing one myself. I'll see if I have to shrink that and report the results here. Cheers, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From bje@cygnus.com Sun Oct 29 15:50:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Sun, 29 Oct 2000 15:50:00 -0000 Subject: Sourcenav color icon image References: <39FCA612.57A96B5B@cs.bilkent.edu.tr> Message-ID: Mo, > This has been fixed in newer versions of Tk. You can now > use an XPM instead of an XBM for the icon image. Of course, > this only works with newer window managers but if it does > not work it will fallback to the XBM. So, it should be > possible to have a color icon image. I must say I'm not surprised to hear this was improved in recent versions of Tk -- I wasn't aware of it, though. Thanks! Ben From mdejong@cygnus.com Sun Oct 29 16:47:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sun, 29 Oct 2000 16:47:00 -0000 Subject: Sourcenav color icon image References: <39FCA612.57A96B5B@cs.bilkent.edu.tr> Message-ID: On Mon, 30 Oct 2000, Eray Ozkural (exa) wrote: > Mo DeJong wrote: > > > > This has been fixed in newer versions of Tk. You can now > > use an XPM instead of an XBM for the icon image. Of course, > > this only works with newer window managers but if it does > > not work it will fallback to the XBM. So, it should be > > possible to have a color icon image. > > > > Thanks a lot, but could you please tell me which artwork I'm > supposed to use for source navigator icon? ;) Sure, how do you like the attached shadowman.xpm? Is the size OK for what you want to do? Mo DeJong Red Hat Inc /* XPM */ static char * shadowman_xpm[] = { "70 56 17 1", " c None", ". c #D50101", "+ c #A10000", "@ c #5D5E5E", "# c #620000", "$ c #B6B5B5", "% c #D8D1D1", "& c #F4F3F3", "* c #8E8E8E", "= c #200404", "- c #E99191", "; c #E8E4E4", "> c #DE5858", ", c #323333", "' c #000000", ") c #FFFFFF", "! c #CC0000", "''''''''''''''''''''''''''''''''''''''''''''''''''''''!!!!!!!!!!!!!!!!", "''''''''''#.!#'#!.......!!!!!!!...!!!!!!!='''''''''''''#!!!!!!!!!!!!!!", "''''''''''+.!=''=##====#+!!!!!!!!!!!!!!!.#''''''''''''''#!!!!!!!!!!!!!", "'''''''''=+..#''''''''''#!!!!!!!!!!!!!!!.+'''''''''''''''+!!!!!!!!!!!!", "'''''''''=!!!!#''''=####+!!!!!!!!!!!!!!!!!#''''''''''''''#!!!!!!!!!!!!", "'''''''''#.!!.!#'#+!.....!!!!!!!!!!!!!!!!.+'''''''''''''''+!!!!!!!!!!!", "'''''''''#.!!!.!+!..!!!!!!!!!!!!!!!!!!!!!.!=''''''''''''''=!!!!!!!!!!!", "'''''''''#.!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!.#'''''''''''''''#!!!!!!!!!!", "'''''''''=!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.+'''''''''''''''=!!!!!!!!!!", "#++!!!!+='#..!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!='''''''''''''''+!!!!!!!!!", "....!!.!=''#+..!!!!!!!!!!!!!!!!!!!!!!!!!!!!.#'''''''''''''''#!!!!!!!!!", "!!!!!!.+''''=#+!...!!!!!!!!!!!!!!!!!!!!!!!!.+'''''''''''''''=+!!!!!!!!", "!!!!!!!!=''''''=#++!...!!!!!!!!!!!!!!!!!!!!.+''''''''''''''''#!!!!!!!!", "!!!!!!!.#'''''''''==##+!..!!!!!!!!!!!!!!!!!.+''''''''''''''''#!!!!!!!!", "!!!!!!!!!#''''''''''''==#+.!!!!!!!!!!!!!!!!.+'#+#='''''''''''=+!!!!!!!", "!!!!!!!!!.#='''''''''''''=+.!!!!!!!!!!!!!!!.#'+..!+#''''''''''+!!!!!!!", "!!!!!!!!!!.+#'''''''''''''=!!!!!!!!!!!!!!!!!=#!!!!.!+=''''''''#!!!!!!!", "!!!!!!!!!!!.!+#='''''''''''+.!!!!!!!!!!!!!.+#!!!!!!!.!='''''''#!!!!!!!", "!!!!!!!!!!!!!..+#='''''''''#.!!!!!!!!!!!!!!!!!!!!!!!!.!=''''''=!!!!!!!", ".!!!!!!!!!!!!!!..!+#=''''''=+.!!!!!!!!!!!!!!!!!!!!!!!!.+''''''=!!!!!!!", "+.!!!!!!!!!!!!!!!!..!+##='''=!.!!!!!!!!!!!!!!!!!!!!!!!!!='''''=+!!!!!!", "=#!.!!!!!!!!!!!!!!!!...!!+###!!!!!!!!!!!!!!!!!!!!!!!!!!.#'''''=+!!!!!!", "''#!.!!!!!!!!!!!!!!!!!!!!....!!!!!!!!!!!!!!!!!!!!!!!!!!.#''''''+!!!!!!", "'''=+!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!=''''''+!!!!!!", "'''''#+..!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!+=''''''+!!!!!!", "''''''=#!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.#'''''''+!!!!!!", "''''''''=#!..!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!=''''''=+!!!!!!", "''''',**@==+!..!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.#'''''''=+!!!!!!", "'''',&)))%,'=#+..!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.#''''''''=!!!!!!!", "''''$)))))%=''=#+!...!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!#'''''''''=!!!!!!!", "''''$))))))*'''''=#+!..!!!!!!!!!!!!!!!!!!!!!!!!!..+=''''''''''#!!!!!!!", "''''$))))))&@@$*''''=#+!...!!!!!!!!!!!!!!!!!!...+#''''''''''''#!!!!!!!", "''''@)))))))&&))@'''''==#++!!!.............!!+#=''''''''''''''+!!!!!!!", "=='''*))))))))));,'''''''''=###+++++++++++##='''''''''''''''''+!!!!!!!", "%$@'''$))))))))));,''''''''''''''=''''=''''''''''''''''''''''=!!!!!!!!", ")))@'',%))))))))))&@'''''''''''''''''''''''''''''''''''''''''#!!!!!!!!", ")));,''=*)))))))))))$@,==,,@****,,@$$@,,@$$'''''''''''''''''=+!!!!!!!!", "))))%@,=*)))))))))))));$%;&)))))&;))))&&))*'''''''''''''''''=!!!!!!!!!", "))))));%)))))))))))))))))))))))))))))))))&@'''''''''''''''''#!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))))))%'''''''''''''''''=!!!!!!!!!!", "))))))))))))))))))))))))))))&&)))&&))))))$'''''''''''''''''#!!!!!!!!!!", ")))))))))))))))))))))))))))$,,*&%,$))))))$''''''''''''''''=+!!!!!!!!!!", ")))))))))))))))))))))))))))*''',='$))))))*''''''''''''''''#!!!!!!!!!!!", ")))))))))))))))))))))))))));**,''*&))))))@''''''',,,@,,,'=!!!!!!!!!!!!", "))))))))))))))))))))))))))))))%$%)))))))$''''=@$%;;&)&;;$+!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))));,''''*))))))))))-!!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))))@''''=$)))))))))%.!!!!!!!!!!!!!", "))))))))))))))))))))))))))))))))))))))*''',*%)))))))));.!!!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))%='=*&))))))))))&>!!!!!!!!!!!!!!!", "))))))))))))))))))))))))))))))))))))&@'=$))))))))))))>!!!!!!!!!!!!!!!!", "))))))))))))))))))))))))))))))))))))*''$))))))))))))>!!!!!!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))$'=*)))))))))))&>!!!!!!!!!!!!!!!!!!", "))))))))))))))))))))))))))))))))))&**%)))))))))));>!!!!!!!!!!!!!!!!!!!", "))))))))))))))))))))))))))))))))))))))))))))))))%>!!!!!!!!!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))))))))))))-.!!!!!!!!!!!!!!!!!!!!!", ")))))))))))))))))))))))))))))))))))))))))))))%>!!!!!!!!!!!!!!!!!!!!!!!"}; From erayo@cs.bilkent.edu.tr Sun Oct 29 17:24:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Sun, 29 Oct 2000 17:24:00 -0000 Subject: Sourcenav color icon image References: Message-ID: <39FCBFE7.A4CBB302@cs.bilkent.edu.tr> Mo DeJong wrote: > Sure, how do you like the attached shadowman.xpm? Is > the size OK for what you want to do? > Nice, but this one or the hourglass? Thanks, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From erayo@cs.bilkent.edu.tr Sun Oct 29 17:26:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Sun, 29 Oct 2000 17:26:00 -0000 Subject: include file problems References: Message-ID: <39FCC06A.A6E5A771@cs.bilkent.edu.tr> Ben Elliston wrote: > > Because often, parsing these files takes a lot of time and imports a lot of > detail into your project you don't care about. It's unusual to really want > to know where strcmp() is declared, since you either know, or don't really > care. Don't you agree? A friend of mine used to make include directories into separate projects for this purpose in VC++. Perhaps it's a useful function after all. Though I don't know how it would be possible in sourcenav. Thanks, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From erayo@cs.bilkent.edu.tr Sun Oct 29 17:39:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Sun, 29 Oct 2000 17:39:00 -0000 Subject: RFC: Description for sourcenav Message-ID: <39FCC387.D36FF5FB@cs.bilkent.edu.tr> Here is the current description for debian sourcenav package. Description: Source code analysis, editor, browser and build tool. Source-Navigator is a source code analysis tool which lets users to edit, browse and build their projects. With it, you can edit your source code, display relationships between classes and functions and members, and display call trees. You can also build your projects, either with your own makefile, or by using Source-Navigator's build system to automatically generate a makefile. Source-Navigator works with the Insight GUI interface for GDB. As you can see I just stole it from your site and edited it a bit. :I I pasted it as it is, so I know that it should be "Insight GUI for gdb" or something like that ;) Should I write Description: Integrated Development Environment (IDE) for various languages for one-line description? And what other changes would be appropriate for the paragraph? (This typically shouldn't be much larger) This is the text with which Debian users search for or read about packages. Thanks! -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From Fabrice_Gautier@sdesigns.com Mon Oct 30 12:00:00 2000 From: Fabrice_Gautier@sdesigns.com (Fabrice Gautier) Date: Mon, 30 Oct 2000 12:00:00 -0000 Subject: Difficulties while running insight and SN Message-ID: <5AF611D83988D4118ACF00E0B10400540557F3@mr-intranet.sdesigns.net> Hi, I'm using Insight 5 and SN 4.52 on WinNT. I often experiments locks with Insight while doing remote debugging and when this occurs SN seems locked too until i kill Insight: I can't display the SN windows but everything return fine once i killed Insight. Is it possible that it is some kind of lock in a shared library ? It seems like a GUI thing more than anything else, at least for SN. Thanks -- Fabrice Gautier fabrice_gautier@sdesigns.com From mdejong@cygnus.com Mon Oct 30 13:16:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 30 Oct 2000 13:16:00 -0000 Subject: SN memory leak. References: <5AF611D83988D4118ACF00E0B10400540557EC@mr-intranet.sdesigns.net> Message-ID: On Fri, 27 Oct 2000, Fabrice Gautier wrote: > Hi, > > Using SN 4.5.2 with WinNT4, I constated that memory occupation never > decrease even when I close a project. I have too quit and restart quite > often if I don't want to be Out of Memory. > > Thanks I have been able to reproduce the problem you decribe on a Linux box running 5.0. I am going to have to add this to our bug list. Mo DeJong Red Hat Inc From edp@shell1.eznet.net Thu Nov 2 00:44:00 2000 From: edp@shell1.eznet.net (Ed P.) Date: Thu, 02 Nov 2000 00:44:00 -0000 Subject: sourcenav 4.52 cparser.c fix for MANY function args Message-ID: <20001102084443.13950.qmail@eznet.net> I fixed a bug in the source navigator cbrowser's cparser.c. Regards, Ed Pendzik edp@eznet.net -------------------------------------------------- Text description -------------------------------------------------- snavigator/parsers/cpp/cparser.c: increased 2 fixed string buffers to prevent core dump when parsing functions with many (67+) arguments. -------------------------------------------------- ChangeLog entry -------------------------------------------------- 2000-11-02 Ed Pendzik * snavigator/parsers/cpp/cparser.c: increased 2 fixed string buffers to prevent core dump when parsing functions with many (67+) arguments. -------------------------------------------------- diff -c3p SN452-source/snavigator/parsers/cpp/cparser.c.orig SN452-source/snavigator/parsers/cpp/cparser.c -------------------------------------------------- *** SN452-source/snavigator/parsers/cpp/cparser.c.orig Thu Nov 2 01:51:25 2000 --- SN452-source/snavigator/parsers/cpp/cparser.c Thu Nov 2 01:52:45 2000 *************** extern void put_cross1( int type, char * *** 6166,6172 **** if( cross_ref_fp ) { Tcl_DString utfString; ! char tmp[1000]; sprintf(tmp,"%d;%d;%s;%d;%d;%d;%d;%ld;%s;%s;%s;%s;%s;%d;\n" , PAF_CROSS_REF_CPP --- 6166,6172 ---- if( cross_ref_fp ) { Tcl_DString utfString; ! char tmp[4096]; sprintf(tmp,"%d;%d;%s;%d;%d;%d;%d;%ld;%s;%s;%s;%s;%s;%d;\n" , PAF_CROSS_REF_CPP *************** extern int f_charno( int i ) *** 7053,7059 **** extern char *identstring( int i ) { sString_t sString; ! static char ac[1000]; sString = ident( i ); --- 7053,7059 ---- extern char *identstring( int i ) { sString_t sString; ! static char ac[4096]; sString = ident( i ); -------------------------------------------------- From mdejong@cygnus.com Thu Nov 2 00:53:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Thu, 02 Nov 2000 00:53:00 -0000 Subject: sourcenav 4.52 cparser.c fix for MANY function args References: <20001102084443.13950.qmail@eznet.net> Message-ID: On Thu, 2 Nov 2000, Ed P. wrote: > I fixed a bug in the source navigator cbrowser's cparser.c. > > Regards, > Ed Pendzik > edp@eznet.net Funny, I just fixed that but the other day :) Your fix makes the buffers bigger. That might make the crashes go away but if you had a function with a ton of arguments, it would come back. My solution to this problem was to use snprintf() and pass in the size of the buffer. What do you think of that approach? Mo DeJong Red Hat Inc From ac131313@cygnus.com Thu Nov 2 02:27:00 2000 From: ac131313@cygnus.com (Andrew Cagney) Date: Thu, 02 Nov 2000 02:27:00 -0000 Subject: sourcenav 4.52 cparser.c fix for MANY function args References: Message-ID: <3A01401E.8C64DEAC@cygnus.com> Mo DeJong wrote: > > On Thu, 2 Nov 2000, Ed P. wrote: > > > I fixed a bug in the source navigator cbrowser's cparser.c. > > > > Regards, > > Ed Pendzik > > edp@eznet.net > > Funny, I just fixed that but the other day :) > > Your fix makes the buffers bigger. That might make > the crashes go away but if you had a function with > a ton of arguments, it would come back. > > My solution to this problem was to use snprintf() > and pass in the size of the buffer. What do you > think of that approach? Check asprintf(). Andrew From erayo@cs.bilkent.edu.tr Thu Nov 2 06:48:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Thu, 02 Nov 2000 06:48:00 -0000 Subject: Call Graph Generator Message-ID: <3A017F1A.A11D066A@cs.bilkent.edu.tr> Hi, I've hacked some tcl code to generate graphviz files for the call graph of a project. I don't know if such a thing was available elsewhere but it works for me (TM) Makes classes into clusters and places their methods in them. Each arc represents function use. The small hack is based on call-freq code... Even the name is call-graph.tk although it doesn't do GUI yet... It's not even alpha, so I await your comments. I've already used it to generate the call graph of a large C++ project and I expect that it will be very useful for others, too. If you have requests for different kinds of graphs, feel free to ask for them as they'll be probably useful for all of us! Thanks! -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo call-graph.tk.gz -------------- next part -------------- A non-text attachment was scrubbed... Name: call-graph.tk.gz Type: application/x-gzip Size: 1828 bytes Desc: not available URL: From bje@cygnus.com Thu Nov 2 13:23:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Thu, 02 Nov 2000 13:23:00 -0000 Subject: Call Graph Generator References: <3A017F1A.A11D066A@cs.bilkent.edu.tr> Message-ID: Hi, I've hacked some tcl code to generate graphviz files for the call graph of a project. I don't know if such a thing was available elsewhere but it works for me (TM) Makes classes into clusters and places their methods in them. Each arc represents function use. I'm really interested to see some screen shots -- can you post a URL or two? Ben From erayo@cs.bilkent.edu.tr Thu Nov 2 14:36:00 2000 From: erayo@cs.bilkent.edu.tr (Eray Ozkural (exa)) Date: Thu, 02 Nov 2000 14:36:00 -0000 Subject: Call Graph Generator References: <3A017F1A.A11D066A@cs.bilkent.edu.tr> Message-ID: <20001103003824.A7748@orion.exa.homeip.net> On Fri, Nov 03, 2000 at 08:22:44AM +1100, Ben Elliston wrote: > I'm really interested to see some screen shots -- can you post a URL or two? Yeah, why not? Here's the huge (I mean large...) ps file. gzipped. and the dot input files, (vanilla and edited) http://pccluster.cs.bilkent.edu.tr/~exa/ I could show it on tk window if tcldot worked with the old tcl/tk version that sourcenav uses. :( When you guys finish porting that to tcl8.3 we can make this into another navigator window!!! Happy hacking, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From wgacquer@ubisoft.fr Thu Nov 2 18:24:00 2000 From: wgacquer@ubisoft.fr (William Gacquer) Date: Thu, 02 Nov 2000 18:24:00 -0000 Subject: Call Graph Generator Message-ID: <5D2A198567EED311BD0F009027A25C874E3502@SRVMAIL-ANN> quite good.. in fact, better than the actual ones -----Original Message----- From: Eray Ozkural [ mailto:erayo@cs.bilkent.edu.tr ] Sent: jeudi 2 novembre 2000 23:38 To: Ben Elliston Cc: sourcenav@sources.redhat.com Subject: Re: Call Graph Generator On Fri, Nov 03, 2000 at 08:22:44AM +1100, Ben Elliston wrote: > I'm really interested to see some screen shots -- can you post a URL or two? Yeah, why not? Here's the huge (I mean large...) ps file. gzipped. and the dot input files, (vanilla and edited) http://pccluster.cs.bilkent.edu.tr/~exa/ I could show it on tk window if tcldot worked with the old tcl/tk version that sourcenav uses. :( When you guys finish porting that to tcl8.3 we can make this into another navigator window!!! Happy hacking, -- Eray (exa) Ozkural Comp. Sci. Dept., Bilkent University, Ankara e-mail: erayo@cs.bilkent.edu.tr www: http://www.cs.bilkent.edu.tr/~erayo From bje@cygnus.com Fri Nov 3 01:48:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Fri, 03 Nov 2000 01:48:00 -0000 Subject: Call Graph Generator References: <20001103003824.A7748@orion.exa.homeip.net> Message-ID: I could show it on tk window if tcldot worked with the old tcl/tk version that sourcenav uses. :( When you guys finish porting that to tcl8.3 we can make this into another navigator window!!! Very nice! I think this would be a nice contribution to S-N, even if it's just a tool that emits the dot files. As for how it could be included in the product, I'll leave that up to the S-N team to contemplate. Ben From Brecht.Vermeulen@rug.ac.be Sat Nov 4 17:26:00 2000 From: Brecht.Vermeulen@rug.ac.be (Brecht Vermeulen) Date: Sat, 04 Nov 2000 17:26:00 -0000 Subject: Call Graph Generator References: <20001103003824.A7748@orion.exa.homeip.net> Message-ID: Hi, some other quick hacking (it was my first time in tcl ;-) ) on the Call Graph Generator which makes it still more interesting to me, so maybe also to others : * the edges are drawn in different colors * calls within the same class are drawn dashed * constructors are trapeziums * as third command line argument a file can be optionally given, which must contain one line with all files (with full path as in the sn project) separated with a space out of which the functions have to be drawn. This was handy because otherwise if you have e.g. a server and client with each a main function, it was drawn as one main function. Still a problem : if a function of one class calls a function with the same name on a pointer which refers to another class, then it is drawn as a loopback, but this seems a problem with the parser of SN, I think. It would be nice if the instance variables were also drawn for each class (.iv database) and if optionally also references to these could be drawn. best regards, and some cool tools ;-) Brecht On Fri, 3 Nov 2000, Eray Ozkural wrote: > On Fri, Nov 03, 2000 at 08:22:44AM +1100, Ben Elliston wrote: > > I'm really interested to see some screen shots -- can you post a URL or two? > > Yeah, why not? Here's the huge (I mean large...) ps file. gzipped. > > and the dot input files, (vanilla and edited) > > http://pccluster.cs.bilkent.edu.tr/~exa/ > > I could show it on tk window if tcldot worked with the old tcl/tk version > that sourcenav uses. :( When you guys finish porting that to tcl8.3 > we can make this into another navigator window!!! > > Happy hacking, > > -- > Eray (exa) Ozkural > Comp. Sci. Dept., Bilkent University, Ankara > e-mail: erayo@cs.bilkent.edu.tr > www: http://www.cs.bilkent.edu.tr/~erayo > --------------------------------- Brecht Vermeulen Department of Information Technology (INTEC) Ghent University - IMEC Sint-Pietersnieuwstraat 41 B-9000 Gent Belgium Phone : +32 9 267 35 86 Fax : +32 9 267 35 99 Email : brecht.vermeulen@intec2.rug.ac.be -------------- next part -------------- A non-text attachment was scrubbed... Name: call-graph2.tk.gz Type: application/x-gzip Size: 2288 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: example_NS.dot.gz Type: application/x-gzip Size: 1500 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: example_NS.ps.gz Type: application/x-gzip Size: 6847 bytes Desc: not available URL: From kline@thought.org Mon Nov 6 13:37:00 2000 From: kline@thought.org (Gary Kline) Date: Mon, 06 Nov 2000 13:37:00 -0000 Subject: Testing + a first question... Message-ID: <20001106133731.B23848@tao.thought.org> Assuming this gets thru to the source navigator list, I'll throw out my question. Is there a way of making an vi or vi-like editor the default editor with sourcenav? After using this environment for several hours over the weekend, it seems to fit my needs well with the exception of the editor. I've added "xterm -e vim" to the preferences in one tcl popup, but find this does not affect the default editor which displays files in color within a popup frame. On my main system I have xemacs with viper, vi, vim, gvim, and xvile and probably other vi-clones. Anybody else use this tool whose fingers are most comfortable with vi? thanks much, gary kline -- Gary D. Kline kline@tao.thought.org Public service Unix From echristo@cygnus.com Mon Nov 6 13:42:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Mon, 06 Nov 2000 13:42:00 -0000 Subject: Testing + a first question... References: <20001106133731.B23848@tao.thought.org> Message-ID: <3A072878.A20F823B@cygnus.com> At one point there were hooks to use (x)emacs as an external editor for source nav. You could run that via viper mode. -eric Gary Kline wrote: > > > Assuming this gets thru to the source navigator list, I'll > throw out my question. > > Is there a way of making an vi or vi-like editor the default > editor with sourcenav? After using this environment for several > hours over the weekend, it seems to fit my needs well with the > exception of the editor. > > I've added "xterm -e vim" to the preferences in one tcl popup, > but find this does not affect the default editor which displays > files in color within a popup frame. On my main system I have > xemacs with viper, vi, vim, gvim, and xvile and probably other > vi-clones. > > Anybody else use this tool whose fingers are most comfortable > with vi? > > thanks much, > > gary kline > > -- > Gary D. Kline kline@tao.thought.org Public service Unix From mdejong@cygnus.com Mon Nov 6 13:43:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 06 Nov 2000 13:43:00 -0000 Subject: Testing + a first question... References: <20001106133731.B23848@tao.thought.org> Message-ID: On Mon, 6 Nov 2000, Gary Kline wrote: > > Assuming this gets thru to the source navigator list, I'll > throw out my question. > > Is there a way of making an vi or vi-like editor the default > editor with sourcenav? After using this environment for several > hours over the weekend, it seems to fit my needs well with the > exception of the editor. > > I've added "xterm -e vim" to the preferences in one tcl popup, > but find this does not affect the default editor which displays > files in color within a popup frame. On my main system I have > xemacs with viper, vi, vim, gvim, and xvile and probably other > vi-clones. > > Anybody else use this tool whose fingers are most comfortable > with vi? > > thanks much, > > gary kline The documentation has an example of how to set up the editor prefs so that vi is opened when you want to edit something. http://sources.redhat.com/sourcenav/online-docs/userguide/editor.html#pgfId-896855 Mo DeJong Red Hat Inc From kline@thought.org Mon Nov 6 14:34:00 2000 From: kline@thought.org (Gary Kline) Date: Mon, 06 Nov 2000 14:34:00 -0000 Subject: Testing + a first question... References: <20001106133731.B23848@tao.thought.org> <3A072878.A20F823B@cygnus.com> Message-ID: <20001106143351.A24808@tao.thought.org> On Mon, Nov 06, 2000 at 01:54:00PM -0800, Eric Christopher wrote: > At one point there were hooks to use (x)emacs as an external editor for > source nav. You could run that via viper mode. > Yes, and directly above were the hooks for vi, which I did use. It works well, but there is another window that you can popup where is displayed with sourcenav's own editor. I'd like to use only a vi editor consistantly. I just finished reading more of the redhat docs and have dropped in some tcl functions into my ~/.sn directory. Be great if this suite let the user choose his own editor... gary -- Gary D. Kline kline@tao.thought.org Public service Unix From kline@thought.org Tue Nov 7 12:36:00 2000 From: kline@thought.org (Gary Kline) Date: Tue, 07 Nov 2000 12:36:00 -0000 Subject: Testing + a first question... References: <3A072878.A20F823B@cygnus.com> Message-ID: <200011072036.eA7KaNd11990@thought.org> I added some of the tcl from the redhat documentation website and am able to invoke either xterm + vim or [x]emacs (which works with viper). So for the most part, I can use this IDE tool with a vi-like editor. Still, if I click on Windows->new->Editor I get the SN default editor. Are there any plans for letting the user choose his own editor and having that editor invoked in whatever window or frame that SN uses? The nice things about this suite are the cross-reference frames that show the relationships of the user's code; and other features. Still, having to use an external editor is a bit clumsy. gary kline -- Gary D. Kline kline@tao.thought.org Public service Unix From Reiner.Suikat@dlr.de Wed Nov 8 00:27:00 2000 From: Reiner.Suikat@dlr.de (Reiner Suikat) Date: Wed, 08 Nov 2000 00:27:00 -0000 Subject: Bug?: Selection of linker in Build Rules Dialog Message-ID: Hello, I'm new to this list and to sourcenav. I just installed it yesterday (binary release on RedHat 6.2) and I'm VERY impressed! One thing I don't get to work though (is it a bug?): The manual says to select the linker program to be used (from the Link Rules dialog): click on the three dots to the left of the linker command entry field. Too bad, there are no three dots, its a plain text field and it cannot be edited. Any hints, why I don't get the selection dialog? I want to use gcc as linker command to build shared libraries, but if I select library as target, it only offers ar, if I select executable, it uses g++, which does not behave well in linking shared libraries. Thanks Reiner -- Reiner Suikat Institute of Flight Guidance German Aerospace Research Center Reiner.Suikat@dlr.de From mdejong@cygnus.com Wed Nov 8 01:16:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 08 Nov 2000 01:16:00 -0000 Subject: Bug?: Selection of linker in Build Rules Dialog References: Message-ID: On Wed, 8 Nov 2000, Reiner Suikat wrote: > Hello, > I'm new to this list and to sourcenav. I just installed it yesterday (binary > release on RedHat 6.2) and I'm VERY impressed! > One thing I don't get to work though (is it a bug?): The manual says to > select the linker program to be used (from the Link Rules dialog): > click on the three dots to the left of the linker command entry field. > Too bad, there are no three dots, its a plain text field and it cannot be > edited. Yeah, that is a bug. The little ... button is actually there, but you just can't see it. This button was removed in prep for the next version. It is a good thing you pointed it out because we need to update the docs to account for that. > Any hints, why I don't get the selection dialog? I want to use gcc as linker > command to build shared libraries, but if I select library as target, it > only offers ar, if I select executable, it uses g++, which does not behave > well in linking shared libraries. I just tried it and it looks like there is a bug in the "Linker" field. I set it to ld but when I closed the dialog it gets reset to gcc. I also tested this with 5.0 and the bug is still in there. I am going to file a bug report on this so it will get fixed in 5.0. Mo DeJong Red Hat Inc From Jean-Hugues.Zorio-NOSPAM@NOSPAM-der.edf.fr-NOSPAM Wed Nov 8 02:42:00 2000 From: Jean-Hugues.Zorio-NOSPAM@NOSPAM-der.edf.fr-NOSPAM (jean-hugues zorio) Date: Wed, 08 Nov 2000 02:42:00 -0000 Subject: Are there Source Navigator Binaries for AIX and/or HP-UX ? Message-ID: <3A092D44.BFAE0094@NOSPAM-der.edf.fr-NOSPAM> Hi, I'm seeking binaries for both AIX and HP-UX (I need HP-UX now, and AIX in the future). I've briefly tried to build the distribution under HP-UX (10.2) but encountered many problems (while building Tcl, oddly) and I had not the time required to solve the matter. Before I try again, I'd like to find an already built binary if someone has managed to compile one, and is willing to share the result of his work. Thanks in advance, Jean-Hugues Zorio From spolk@redhat.com Wed Nov 8 11:08:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 08 Nov 2000 11:08:00 -0000 Subject: Are there Source Navigator Binaries for AIX and/or HP-UX ? References: <3A092D44.BFAE0094@NOSPAM-der.edf.fr-NOSPAM> Message-ID: <4.2.0.58.20001108110932.01a635a0@pop.cygnus.com> At 11:39 AM 11/8/00 +0100, jean-hugues zorio wrote: >Hi, > >I'm seeking binaries for both AIX and HP-UX (I need HP-UX now, and AIX >in the future). > >I've briefly tried to build the distribution under HP-UX (10.2) but >encountered many problems (while building Tcl, oddly) and I had not the >time required to solve the matter. We have built on both of these systems in the past. On HPUX, however, the "make install" is crashing /bin/sh on me. I usually have no problems compiling, however. Are you using gcc, or HP's compiler? >Before I try again, I'd like to find an already built binary if someone >has managed to compile one, and is willing to share the result of his >work. Well, next time I do a net release, I will include an HPUX binary. >Thanks in advance, > > >Jean-Hugues Zorio > Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From bje@cygnus.com Wed Nov 8 13:27:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Wed, 08 Nov 2000 13:27:00 -0000 Subject: Are there Source Navigator Binaries for AIX and/or HP-UX ? References: <3A092D44.BFAE0094@NOSPAM-der.edf.fr-NOSPAM> Message-ID: I've briefly tried to build the distribution under HP-UX (10.2) but encountered many problems (while building Tcl, oddly) and I had not the time required to solve the matter. If you post (brief) error messages from the compilation step, I'm sure someone on the list can help you. Ben From tobias.mkortkamp@cas-gmbh.de Thu Nov 9 04:56:00 2000 From: tobias.mkortkamp@cas-gmbh.de (Tobias Muller-Kortkamp) Date: Thu, 09 Nov 2000 04:56:00 -0000 Subject: Database Export Message-ID: <01C04A55.0406BD40.tobias.mkortkamp@cas-gmbh.de> Hi, I tried some database export. Now there are fields in various tables named "attributes". They are coded in hexadecimal digits. is there a reference what these codings mean? i.e (excerpt from table cl (classes)) name start_position end_position attributes ActionHandler 27.007 37.2 0x1 JavaDraw 15.013 302.1 0x4 JavaDrawFrame 10.013 33.1 0x4 Shape 8.019 20.1 0x24 UltraColor 9.013 83.1 0x4 WidgetPanel 14.013 236.1 0x4 Thanks in advance Tobias Tobias Mueller-Kortkamp mailto:tobias.mkortkamp@cas-gmbh.de CAS GmbH http://www.cas-gmbh.de Friedensstrasse 20 Phone: +49 (0) 61 04 - 98 08 - 0 D-63179 Obertshausen Fax: +49 (0) 61 04 - 98 08 - 30 ***************************************************** From ceo@shore.net Thu Nov 9 12:13:00 2000 From: ceo@shore.net (Chip Olson) Date: Thu, 09 Nov 2000 12:13:00 -0000 Subject: Building on FreeBSD 3.2 Message-ID: <3A0B0553.9938D8D3@shore.net> Hi; I'm trying to build SN 4.5.2 on FreeBSD 3.2, and am failing miserably. [ceo ~/me/SN452-source/snavigator]$ make make all-recursive Making all in misc Making all in libutils Making all in db Making all in ../parsers/cpp/cpplib make: don't know how to make ../../db/PORT/libdb.a. Stop *** Error code 1 Stop. *** Error code 1 Stop. *** Error code 1 Stop. It looks like the make process is in SN452-source/snavigator/db, builds SN452-source/snavigator/parsers/cpp/cpplib from there, and then goes to SN452-source/db/PORT, which is in the original source tree rather than the parallel build tree, SN452-source/snavigator/db/PORT. Since configure was run in the build tree, there's no Makefile there and nothing was built. Is there something subtle wrong with my system (which wouldn't surprise me in the least), or am I just missing something? Upgrading to FreeBSD 4.0 isn't an option, unfortunately. -- -Chip Olson. | ceo at shore dot net "You ask me why we celebrate, when nothing has been won; We take dark hours, we make them great, that's all we've ever done..." -Oysterband, "This Is The Voice". From spolk@redhat.com Thu Nov 9 12:25:00 2000 From: spolk@redhat.com (Syd Polk) Date: Thu, 09 Nov 2000 12:25:00 -0000 Subject: Building on FreeBSD 3.2 References: <3A0B0553.9938D8D3@shore.net> Message-ID: <3A0B0921.589A902A@redhat.com> Chip Olson wrote: > > Hi; > > I'm trying to build SN 4.5.2 on FreeBSD 3.2, and am failing miserably. > > [ceo ~/me/SN452-source/snavigator]$ make > make all-recursive > Making all in misc > Making all in libutils > Making all in db > Making all in ../parsers/cpp/cpplib > make: don't know how to make ../../db/PORT/libdb.a. Stop > *** Error code 1 > > Stop. > *** Error code 1 > > Stop. > *** Error code 1 > > Stop. > > It looks like the make process is in SN452-source/snavigator/db, builds > SN452-source/snavigator/parsers/cpp/cpplib from there, and then goes to > SN452-source/db/PORT, which is in the original source tree rather than the > parallel build tree, SN452-source/snavigator/db/PORT. Since configure was > run in the build tree, there's no Makefile there and nothing was built. It looks like configure bombed out in this directory. Dump your complete configure log and look for errors. From irox@redhat.com Thu Nov 9 12:42:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Thu, 09 Nov 2000 12:42:00 -0000 Subject: Building on FreeBSD 3.2 Message-ID: <200011092036.MAA01922@manx.soma.redhat.com> On Thu, 09 Nov 2000, Chip Olson wrote: > Date: Thu, 09 Nov 2000 15:13:07 -0500 > To: sourcenav@sourceware.cygnus.com > From: Chip Olson > Subject: Building on FreeBSD 3.2 > > > Hi; > > I'm trying to build SN 4.5.2 on FreeBSD 3.2, and am failing miserably. > > [ceo ~/me/SN452-source/snavigator]$ make Are all the other parts build (tk,tcl,db,grep,etc)? Otherwise you probably want to do something like: [ceo ~/me/SN452-source]$ make all-snavigator Which sould build db/PORT/* and friends before buildin the main snavigator directory. Afterwards, just do a make install-snavigator. (My making a couple of assumptions like you have a build dir like ~/me/SN452-build and so on.) Ian. > make all-recursive > Making all in misc > Making all in libutils > Making all in db > Making all in ../parsers/cpp/cpplib > make: don't know how to make ../../db/PORT/libdb.a. Stop > *** Error code 1 > > Stop. > *** Error code 1 > > Stop. > *** Error code 1 > > Stop. > > It looks like the make process is in SN452-source/snavigator/db, builds > SN452-source/snavigator/parsers/cpp/cpplib from there, and then goes to > SN452-source/db/PORT, which is in the original source tree rather than > the > parallel build tree, SN452-source/snavigator/db/PORT. Since configure was > run in the build tree, there's no Makefile there and nothing was built. > > Is there something subtle wrong with my system (which wouldn't surprise > me > in the least), or am I just missing something? Upgrading to FreeBSD 4.0 > isn't an option, unfortunately. > > -- > -Chip Olson. | ceo at shore dot net > "You ask me why we celebrate, when nothing has been won; > We take dark hours, we make them great, that's all we've ever done..." > -Oysterband, "This Is The Voice". > From bje@cygnus.com Thu Nov 9 13:38:00 2000 From: bje@cygnus.com (Ben Elliston) Date: Thu, 09 Nov 2000 13:38:00 -0000 Subject: Database Export References: <01C04A55.0406BD40.tobias.mkortkamp@cas-gmbh.de> Message-ID: Hi, is there a reference what these codings mean? i.e (excerpt from table cl (classes)) name start_position end_position attributes ActionHandler 27.007 37.2 0x1 JavaDraw 15.013 302.1 0x4 JavaDrawFrame 10.013 33.1 0x4 Shape 8.019 20.1 0x24 UltraColor 9.013 83.1 0x4 WidgetPanel 14.013 236.1 0x4 If I recall correctly, these attributes are formed from the set of constants defined in sn.h -- such as: #define PAF_PRIVATE 0x000001 These constants are OR'ed together to form attribute values. Ben From ceo@shore.net Thu Nov 9 14:02:00 2000 From: ceo@shore.net (Chip Olson) Date: Thu, 09 Nov 2000 14:02:00 -0000 Subject: Building on FreeBSD 3.2 References: <200011092036.MAA01922@manx.soma.redhat.com> Message-ID: <3A0B1EB8.814CDF75@shore.net> Wow, that was fast. OK, I'm a moron; I wasn't building out of a parallel build directory. Now I'm having another problem. After lintcl8.1.a successfully builds: gcc -c -g -O2 -fwritable-strings -I../../../SN452-source/tcl8.1/unix/../generic -I../../../SN452-source/tcl8.1/unix -DHAVE_GETCWD=1 -DHAVE_OPENDIR=1 -DHAVE_STRSTR=1 -DHAVE_STRTOL=1 -DHAVE_TMPNAM=1 -DHAVE_WAITPID=1 -DNO_VALUES_H=1 -DHAVE_UNISTD_H=1 -DUSE_TERMIOS=1 -DHAVE_SYS_TIME_H=1 -DTIME_WITH_SYS_TIME=1 -DHAVE_TM_ZONE=1 -DHAVE_TM_GMTOFF=1 -DHAVE_ST_BLKSIZE=1 -DSTDC_HEADERS=1 -DNEED_MATHERR=1 -DHAVE_SIGNED_CHAR=1 -DHAVE_SYS_IOCTL_H=1 -DHAVE_SYS_FILIO_H=1 -DTCL_SHLIB_EXT=\"\" ../../../SN452-source/tcl8.1/unix/../unix/tclAppInit.c gcc tclAppInit.o -L/usr/home/ceo/SN/SN452-build/tcl8.1/unix -ltcl8.1 -lm -lc \ -o tclsh /usr/home/ceo/SN/SN452-build/tcl8.1/unix/libtcl8.1.a(tclLoad.o): In function `Tcl_LoadObjCmd': /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c:332: undefined reference to `TclpLoadFile' /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c:342: undefined reference to `TclpUnloadFile' /usr/home/ceo/SN/SN452-build/tcl8.1/unix/libtcl8.1.a(tclLoad.o): In function `LoadExitProc': /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c(.text+0x841): undefined reference to `TclpUnloadFile' gmake[2]: *** [tclsh] Error 1 gmake[2]: Leaving directory `/usr/home/ceo/SN/SN452-build/tcl8.1/unix' gmake[1]: *** [all] Error 2 gmake[1]: Leaving directory `/usr/home/ceo/SN/SN452-build/tcl8.1' gmake: *** [all-tcl8.1] Error 2 I have perfectly functional builds of Tcl 8.0 and 8.2 on my system already. -- -Chip Olson. | ceo at shore dot net "You ask me why we celebrate, when nothing has been won; We take dark hours, we make them great, that's all we've ever done..." -Oysterband, "This Is The Voice". From bje@redhat.com Thu Nov 9 14:37:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 09 Nov 2000 14:37:00 -0000 Subject: Building on FreeBSD 3.2 References: <3A0B1EB8.814CDF75@shore.net> Message-ID: Now I'm having another problem. After lintcl8.1.a successfully builds: /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c:332: undefined reference to `TclpLoadFile' /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c:342: undefined reference to `TclpUnloadFile' /usr/home/ceo/SN/SN452-build/tcl8.1/unix/libtcl8.1.a(tclLoad.o): In function `LoadExitProc': /usr/home/ceo/SN/SN452-build/tcl8.1/unix/../../../SN452-source/tcl8.1/unix/../generic/tclLoad.c(.text+0x841): undefined reference to `TclpUnloadFile' At a glance, this looks like the code that loads and unloads dynamic modules in Tcl -- that is, code that would use dlopen() or similar. Perhaps Tcl's configury is unable to determine which mechanism is needed on FreeBSD? Again, check Tcl's config.log for the gory details. It's probably been known about and fixed in the official Tcl distribution for some time. Ben From ceo@shore.net Thu Nov 9 18:48:00 2000 From: ceo@shore.net (Chip Olson) Date: Thu, 09 Nov 2000 18:48:00 -0000 Subject: Building on FreeBSD 3.2 References: Message-ID: <3A0B61D6.2F988250@shore.net> Two steps forward and one back. Indeed, config.log showed me the answer; it wasn't able to find libdl.so. This turned out to be in /compat/linux/lib, so I added a -L line pointing thereto at the point in the config script where it compiles its little bit of test code to see if dlopen() exists. Now, the same compilation gives the result "/compat/linux/lib/libdl.so: undefined reference to `_IO_stderr_". So it looks like I have some poking around with the FreeBSD Linux compatibility libraries ahead of me; oh joy. I have a question, however, that might enable me to bypass all that: Is there a way to build SN in such a way that it uses the already-installed versions of tcl, grep, db etc. rather than the ones that come with the source distribution? -- -Chip Olson. | ceo at shore dot net "You ask me why we celebrate, when nothing has been won; We take dark hours, we make them great, that's all we've ever done..." -Oysterband, "This Is The Voice". From bje@redhat.com Thu Nov 9 18:56:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 09 Nov 2000 18:56:00 -0000 Subject: Building on FreeBSD 3.2 References: <3A0B61D6.2F988250@shore.net> Message-ID: Two steps forward and one back. Indeed, config.log showed me the answer; it wasn't able to find libdl.so. This turned out to be in /compat/linux/lib, so I added a -L line pointing thereto at the point in the config script where it compiles its little bit of test code to see if dlopen() exists. Now, the same compilation gives the result "/compat/linux/lib/libdl.so: undefined reference to `_IO_stderr_". So it looks like I have some poking around with the FreeBSD Linux compatibility libraries ahead of me; oh joy. Why are you trying to use the Linux compatibility? Why not try and compile the whole package to run natively on FreeBSD? If that's the case, where do you normally link dl_open() from? Ben From spolk@redhat.com Thu Nov 9 20:27:00 2000 From: spolk@redhat.com (Syd Polk) Date: Thu, 09 Nov 2000 20:27:00 -0000 Subject: Building on FreeBSD 3.2 References: <3A0B61D6.2F988250@shore.net> Message-ID: <3A0B79FB.573AFA8B@redhat.com> Chip Olson wrote: > > Two steps forward and one back. Indeed, config.log showed me the answer; it > wasn't able to find libdl.so. This turned out to be in /compat/linux/lib, > so I added a -L line pointing thereto at the point in the config script > where it compiles its little bit of test code to see if dlopen() exists. > Now, the same compilation gives the result "/compat/linux/lib/libdl.so: > undefined reference to `_IO_stderr_". So it looks like I have some poking > around with the FreeBSD Linux compatibility libraries ahead of me; oh joy. > > I have a question, however, that might enable me to bypass all that: Is > there a way to build SN in such a way that it uses the already-installed > versions of tcl, grep, db etc. rather than the ones that come with the > source distribution? > Not currently. We are working hard for that to happen. From mdejong@cygnus.com Fri Nov 10 00:08:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Fri, 10 Nov 2000 00:08:00 -0000 Subject: Bug?: Selection of linker in Build Rules Dialog References: Message-ID: On Wed, 8 Nov 2000, Reiner Suikat wrote: > I want to use gcc as linker command to build shared libraries, but > if I select library as target, it only offers ar, if I select > executable, it uses g++, which does not behave well in linking > shared libraries. > > Thanks > Reiner Try out the attached patch. It should apply cleanly to the 4.5.2 release and fix the problem where the linker does not reflect changes made in the link rules dialog. Mo DeJong Red Hat Inc Index: gui/buildtarget.tcl =================================================================== RCS file: /cvs/cvsfiles/devo/snavigator/gui/buildtarget.tcl,v retrieving revision 1.1.2.27.2.9 diff -u -r1.1.2.27.2.9 buildtarget.tcl --- buildtarget.tcl 2000/02/11 23:48:53 1.1.2.27.2.9 +++ buildtarget.tcl 2000/11/10 08:04:58 @@ -429,6 +429,10 @@ # Get the link entry point. set LinkerEntryPoint [paf_db_proj get -key "IDELinkEntryPoint$TargetName"] + # Get the link command/executable + set LinkerApplication [paf_db_proj get\ + -key "IDELinkerApplication${TargetName}"] + # Get the defines list # TODO: Get rule type from toolchain foreach ruletype "CC CPP JAVA" { @@ -493,6 +497,9 @@ # Store the link entry point. paf_db_proj put "IDELinkEntryPoint$TargetName" $LinkerEntryPoint + # Store the link command/executable + paf_db_proj put "IDELinkerApplication${TargetName}" $LinkerApplication + # Store the defines list # TODO: Get rule type from toolchain foreach ruletype "CC CPP JAVA" { @@ -554,6 +561,9 @@ # delete the link entry point. paf_db_proj delete "IDELinkEntryPoint$TargetName" + + # delete the link command/executable + paf_db_proj delete "IDELinkerApplication${TargetName}" # delete the defines list # TODO: Get rule type from toolchain Index: gui/targetedit.tcl =================================================================== RCS file: /cvs/cvsfiles/devo/snavigator/gui/targetedit.tcl,v retrieving revision 1.1.2.49.2.14 diff -u -r1.1.2.49.2.14 targetedit.tcl --- targetedit.tcl 2000/04/03 21:41:40 1.1.2.49.2.14 +++ targetedit.tcl 2000/11/10 08:04:58 @@ -786,7 +786,7 @@ $displaycmdline configure -state disabled # Get linker - SetLinkerExecutable + SetLinkerExecutable [$linkExecLEB.txt get] # Update the linker command line set cmdline [GetLinkerCommandLine $tchain_obj] From tobias.mkortkamp@cas-gmbh.de Fri Nov 10 00:19:00 2000 From: tobias.mkortkamp@cas-gmbh.de (Tobias Muller-Kortkamp) Date: Fri, 10 Nov 2000 00:19:00 -0000 Subject: AW: Database Export Message-ID: <01C04AF7.47229480.tobias.mkortkamp@cas-gmbh.de> Hi, Ben, thank you for your prompt answer. You are right, it is in "sn.h". I?ll now make a list of sensible combinations. Best wishes Tobias Hi, is there a reference what these codings mean? i.e (excerpt from table cl (classes)) name start_position end_position attributes ActionHandler 27.007 37.2 0x1 JavaDraw 15.013 302.1 0x4 JavaDrawFrame 10.013 33.1 0x4 Shape 8.019 20.1 0x24 UltraColor 9.013 83.1 0x4 WidgetPanel 14.013 236.1 0x4 If I recall correctly, these attributes are formed from the set of constants defined in sn.h -- such as: #define PAF_PRIVATE 0x000001 These constants are OR'ed together to form attribute values. Ben From Reiner.Suikat@dlr.de Fri Nov 10 01:45:00 2000 From: Reiner.Suikat@dlr.de (Reiner Suikat) Date: Fri, 10 Nov 2000 01:45:00 -0000 Subject: Bug?: Selection of linker in Build Rules Dialog References: Message-ID: Thanks a lot, it works fine! regards Reiner -----Original Message----- From: sourcenav-owner@sources.redhat.com [ mailto:sourcenav-owner@sources.redhat.com]On Behalf Of Mo DeJong Sent: Friday, November 10, 2000 9:09 AM To: sourcenav@sourceware.cygnus.com Subject: Re: Bug?: Selection of linker in Build Rules Dialog On Wed, 8 Nov 2000, Reiner Suikat wrote: > I want to use gcc as linker command to build shared libraries, but > if I select library as target, it only offers ar, if I select > executable, it uses g++, which does not behave well in linking > shared libraries. > > Thanks > Reiner Try out the attached patch. It should apply cleanly to the 4.5.2 release and fix the problem where the linker does not reflect changes made in the link rules dialog. Mo DeJong Red Hat Inc From asincero@arcadio.net Fri Nov 10 06:45:00 2000 From: asincero@arcadio.net (Arcadio A. Sincero Jr.) Date: Fri, 10 Nov 2000 06:45:00 -0000 Subject: Unable to select "Library" as Target Type! Message-ID: Hello list, I'm trying to select "Library" as the Target Type for one of my Build Targets for a project. However, its not letting me select anything other than "Executable". Is this a bug or am I doing something wrong? I've used both the precompiled version of Source Navigator (v4.5.2) and I've also compiled it myself. Both have the same problem. Thanks! --- =============================================================================== Arcadio A. Sincero Jr. Computer Science Major at the University of Maryland Baltimore County, Linux systems administrator, wanna-be Windows and Linux software developer, amateur competitive bodybuilder and all-around nice guy. WWW: http://www.arcadio.net/~asincero (COMING REAL SOON NOW(tm)!) E-mail: mailto:asincero@arcadio.net "If we weren't supposed to eat animals, then why are they made with meat?" From Dave@thomases.com Fri Nov 10 13:25:00 2000 From: Dave@thomases.com (Dave Thomas) Date: Fri, 10 Nov 2000 13:25:00 -0000 Subject: Adding Ruby support to Source Navigator Message-ID: I've written a Source Navigator parser, and I'm in the middle of writing a highlight parser, for the Ruby language. What is the recommended way of distributing this? I can see how I can ship the parser separately, but unless I'm misunderstanding something the highlighter has to be linked in to the product. Would you folks be interested in adding this Ruby support to the main product? Regards Dave From ceo@shore.net Fri Nov 10 14:19:00 2000 From: ceo@shore.net (Chip Olson) Date: Fri, 10 Nov 2000 14:19:00 -0000 Subject: [Fwd: Re: Building on FreeBSD 3.2] Message-ID: <3A0C7464.7A306A15@shore.net> Ben Elliston wrote: > Why are you trying to use the Linux compatibility? Why not try and compile > the whole package to run natively on FreeBSD? If that's the case, where do > you normally link dl_open() from? Why, that's an excellent idea. (The answer to your question is "because I don't know what part of this is Linux-specific and what isn't"). As it turns out, dlopen() in FreeBSD is in libc rather than in its own library, so I took the -ldl off of the line in the config script. Now it doesn't complain about that when I run configure, but I'm still getting the same error when I try to build, and a check of config.log reveals a number of similar issues (i.e. trying to link in libraries that don't exist in FreeBSD). I'd think that, if this is a standard Tcl distribution, it'd be set up to deal with FreeBSD as well as Linux, so I'm not quite sure what's going on here. At this point, I'm not sure if the issue lies with SN, the 3rd-party modules it includes, FreeBSD, or my setup. But I *MASSIVELY* appreciate everyone's help. -- -Chip Olson. | ceo at shore dot net "You ask me why we celebrate, when nothing has been won; We take dark hours, we make them great, that's all we've ever done..." -Oysterband, "This Is The Voice". From bje@redhat.com Fri Nov 10 16:15:00 2000 From: bje@redhat.com (Ben Elliston) Date: Fri, 10 Nov 2000 16:15:00 -0000 Subject: [Fwd: Re: Building on FreeBSD 3.2] References: <3A0C7464.7A306A15@shore.net> Message-ID: Hi Chip, turns out, dlopen() in FreeBSD is in libc rather than in its own library, so I took the -ldl off of the line in the config script. Now it doesn't complain about that when I run configure, but I'm still getting the same error when I try to build, and a check of config.log reveals a number of similar issues (i.e. trying to link in libraries that don't exist in FreeBSD). I'd think that, if this is a standard Tcl distribution, it'd be set up to deal with FreeBSD as well as Linux, so I'm not quite sure what's going on here. OK -- let's tackle this from the beginning. If you configure the entire tree, what breaks first? Do you know why? Can you send more details? I'm prepared to help you nut through it. Cheers, Ben From asincero@arcadio.net Fri Nov 10 16:21:00 2000 From: asincero@arcadio.net (Arcadio A. Sincero Jr.) Date: Fri, 10 Nov 2000 16:21:00 -0000 Subject: Unable to select "Library" as Target Type! References: Message-ID: <00c801c04b75$560ea0f0$060a0a0a@arcadio.net> Never mind, I think I discovered what the problem was. The problem apparently was with the X server I was using (XWin32 on a Windows 2000 box) because when I switched to eXceed, the Target type combo box allowed me to select Library :-/. So either this is a bug in XWin32 or a bug in Tk. --- ====================================================================== Arcadio A. Sincero Jr. Computer Science Major at the University of Maryland Baltimore County, Linux systems administrator, wanna-be Windows and Linux software developer, amateur competitive bodybuilder and all-around nice guy. WWW: http://www.arcadio.net/~asincero (COMING REAL SOON NOW(tm)!) E-mail: mailto:asincero@arcadio.net "If we weren't supposed to eat animals, then why are they made with meat?" ----- Original Message ----- From: "Arcadio A. Sincero Jr." To: Sent: Friday, November 10, 2000 10:41 AM Subject: Unable to select "Library" as Target Type! > Hello list, > > I'm trying to select "Library" as the Target Type for one of my > Build Targets for a project. However, its not letting me select anything > other than "Executable". Is this a bug or am I doing something > wrong? I've used both the precompiled version of Source Navigator > (v4.5.2) and I've also compiled it myself. Both have the same problem. > > Thanks! > > --- > ============================================================================ === > Arcadio A. Sincero Jr. > Computer Science Major at the University of Maryland Baltimore County, > Linux systems administrator, wanna-be Windows and Linux software developer, > amateur competitive bodybuilder and all-around nice guy. > > WWW: http://www.arcadio.net/~asincero (COMING REAL SOON NOW(tm)!) > E-mail: mailto:asincero@arcadio.net > > "If we weren't supposed to eat animals, then why are they made with meat?" > > > > From mdejong@cygnus.com Sat Nov 11 18:55:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sat, 11 Nov 2000 18:55:00 -0000 Subject: Adding Ruby support to Source Navigator References: Message-ID: On 10 Nov 2000, Dave Thomas wrote: > I've written a Source Navigator parser, and I'm in the middle of > writing a highlight parser, for the Ruby language. > > What is the recommended way of distributing this? I can see how I can > ship the parser separately, but unless I'm misunderstanding something > the highlighter has to be linked in to the product. > > Would you folks be interested in adding this Ruby support to the main > product? Well, are you going to be around to help keep it working? I don't know that would have a problem with adding support for ruby, I just want to make sure we don't end up with an abandoned parser. I know Ben was thinking about what changes would be needed to move the syntax highlighter out of the core, but he would know more about that. Mo DeJong Red Hat Inc From Reiner.Suikat@dlr.de Mon Nov 13 00:38:00 2000 From: Reiner.Suikat@dlr.de (Reiner Suikat) Date: Mon, 13 Nov 2000 00:38:00 -0000 Subject: Suggestions for future releases Message-ID: Hello, after working for a few days with sourcenav I'm very impressed with it! This is really a great tool to manage our software project consisting of some hundreds of files. It only took me a day to get the entire project with all build targets set up. However, as with every neat thing, there is room for improvement. Here are some suggestions for things that I consider extremely useful: 1) Import and export of build targets. I have found no way to exchange build target definitions between users. Since our project consists of about 20 shared libs and 10 executables it is extremely annoying, if every user has to set up all targets himself. The only way to copy target definitions at this time appears to copy the .proj file. Unfortunately this also contains other things like user specific settings, which one might not want to overwrite. I think an import/exprot facility should be fairly easy to add, as writing project definitions to and from the .proj file is implemented anyway. All it needs are menu entries to import or export selected (or all) targets to an external file (in .proj format). Of course, this requires also some overwrite protection mechanism. 2) Meta or super targets. With this I mean targets that define a set of other targets to be build. As said before, we have about 30 targets in our project, so rebuilding the whole thing requires 30 separate builds. With a super target one could do this in one step. Obviously it would be neat if one could define different such super targets, like build all shared libs in the project. 3) Projectwide Defines/Compiler flags/Include paths. This would allow to change defines for the entire project easily. Such changes would be useful when moving the software to another machine or architecture. As a workaround it is of course possible to use environment variables, but that is less elegant. 4) A version control view of the entire project. Sourcenav's version control support is great for single files, there is nothing missing here. But there is no support for the project as a whole, like a list of files that are modified locally or that need to be updated from the repository. This would be a great addition also. (Of course, one could use standalone version control software for that, but that means working with two different tools rather than one integrated tool.) These are my suggestions for things I consider useful or missing in sourcenav. Again, I am very impressed with sourcenav! I would like to see other peoples opinions on these and especially of course, if any of these are already in the works ... Greetings Reiner -- Reiner Suikat Institute of Flight Guidance German Aerospace Center (DLR) reiner.suikat@dlr.de From irox@redhat.com Mon Nov 13 01:26:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Mon, 13 Nov 2000 01:26:00 -0000 Subject: Suggestions for future releases References: Message-ID: <3A0FB3A4.D6E5C32C@redhat.com> Reiner Suikat wrote: > 1) Import and export of build targets. I have found no way to exchange build > target definitions between users. Since our project consists of about 20 > shared libs and 10 executables it is extremely annoying, if every user has > to set up all targets himself. The only way to copy target definitions at > this time appears to copy the .proj file. Unfortunately this also contains > other things like user specific settings, which one might not want to > overwrite. > I think an import/exprot facility should be fairly easy to add, as writing > project definitions to and from the .proj file is implemented anyway. All it > needs are menu entries to import or export selected (or all) targets to an > external file (in .proj format). Of course, this requires also some > overwrite protection mechanism. I've been thinking about this for a while now. One idea I'm kind of being to like is that of exporting/storing build target infos in XML format. Ideally these XML files will be able to replace Makefiles to some extent (or Makefile.am). SN would be able to load up these XML files and build the projects they represent. But, using a tool like sn2make you could generate regular makefiles (automated build systems, builds on headless or remote boxes. $ sn2make -f mybuildtarget.sn | make all (or whatever) What I'd really like is a system that can not only replace makefiles but also handle tricks and coolness of automake and autoconf. But without forcing developers into a GUI environment (it'll be easier in Source-Navigator, but you can still easily do builds, etc. from the command line). > 2) Meta or super targets. With this I mean targets that define a set of > other targets to be build. As said before, we have about 30 targets in our > project, so rebuilding the whole thing requires 30 separate builds. With a > super target one could do this in one step. Obviously it would be neat if > one could define different such super targets, like build all shared libs in > the project. I've also thought about this to some extent. Actually, this is what I wanted to implement the first time around, but I called them abstract targets. Unfortunately, it was going to take more time than I had to implement. > 3) Projectwide Defines/Compiler flags/Include paths. This would allow to > change defines for the entire project easily. Such changes would be useful > when moving the software to another machine or architecture. As a workaround > it is of course possible to use environment variables, but that is less > elegant. Also, Defines/Compiler flags/Include paths/etc., on a per file basis within a build target. > 4) A version control view of the entire project. Sourcenav's version control > support is great for single files, there is nothing missing here. But there > is no support for the project as a whole, like a list of files that are > modified locally or that need to be updated from the repository. This would > be a great addition also. (Of course, one could use standalone version > control software for that, but that means working with two different tools > rather than one integrated tool.) This is a "historic" problem in SN. The version control interfaces was designed to use many different type of version control in a kind of plug-n-play way (add the specs to your config file a Robert's your dad's brother, it works!). Of course it doesn't work so great due to the conceptual differences between different version control systems. This is the main reason it works for single files only... So, for things to work at the "entire project" level, we need to rewrite the version control UI for each version control system we want to support. One feature I'd like to add (post 5.0 release) is to see which files in your project are current, changed locally and changed in the repository. > These are my suggestions for things I consider useful or missing in > sourcenav. Again, I am very impressed with sourcenav! I would like to see > other peoples opinions on these and especially of course, if any of these > are already in the works ... The first three items, I believe, are essential if we want to be a serious IDE. Anyway, I'm glad to here you like Source-Navigator and thanks for the feedback. Ian. From Reiner.Suikat@dlr.de Mon Nov 13 02:44:00 2000 From: Reiner.Suikat@dlr.de (Reiner Suikat) Date: Mon, 13 Nov 2000 02:44:00 -0000 Subject: Building on SGI with IRIX 6.2 Message-ID: Hello, I'm trying to build Sourcenav 4.5.2 on an SGI Indigo 2 workstation running IRIX 6.2. I had no success using SGI's compiler, but did get a lot farther with gcc (version 2.8.1). Unfortunately, it doesn't work all the way, it breaks here: gcc -DHAVE_CONFIG_H -I. -I../../../SN452-source/snavigator/hyper -I.. -Wall -fwritable-strings -I../../../SN452-source/snavigator/misc/libutils -I/usr/ people/suikat/SN452-source/tk8.1/generic -I/usr/people/suikat/SN452-source/t k8.1/generic/../unix -I/usr/people/suikat/SN452-source/tk8.1/generic/../win -I/usr/people/suikat/SN452-source/tcl8.1/generic -I../../../SN452-source/sn avigator/../db/include -I../../../SN452-source/snavigator/db -I../../../SN45 2-source/snavigator/../libgui/src -g -O2 -c tcl_high.c ../../../SN452-source/snavigator/hyper/tcl_high.l: In function `tcl_highlight_lex': ../../../SN452-source/snavigator/hyper/tcl_high.l:55: warning: subscript has type `char' ../../../SN452-source/snavigator/hyper/tcl_high.l: In function `flush_lex_scanner': ../../../SN452-source/snavigator/hyper/tcl_high.l:280: `YY_FLUSH_BUFFER' undeclared (first use in this function) ../../../SN452-source/snavigator/hyper/tcl_high.l:280: (Each undeclared identifier is reported only once ../../../SN452-source/snavigator/hyper/tcl_high.l:280: for each function it appears in.) ../../../SN452-source/snavigator/hyper/tcl_high.l: At top level: ../../../SN452-source/snavigator/hyper/tcl_high.l:561: warning: `yyunput' defined but not used ../../../SN452-source/snavigator/hyper/tcl_high.l:812: warning: `yy_push_state' defined but not used ../../../SN452-source/snavigator/hyper/tcl_high.l:839: warning: `yy_pop_state' defined but not used ../../../SN452-source/snavigator/hyper/tcl_high.l:848: warning: `yy_top_state' defined but not used tcl_high.c:239: warning: `yy_flex_strcpy' declared `static' but never defined make[3]: *** [tcl_high.o] Error 1 make[3]: Leaving directory `/usr/people/suikat/tmp/snavigator/hyper' make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory `/usr/people/suikat/tmp/snavigator' make[1]: *** [all-recursive-am] Error 2 make[1]: Leaving directory `/usr/people/suikat/tmp/snavigator' make: *** [all-snavigator] Error 2 Any idea? Do I need a more recent version of gcc? Even better, is there a binary for SGI available somewhere? Thanks Reiner From endre@cmr.no Mon Nov 13 04:16:00 2000 From: endre@cmr.no (Endre M. Lidal) Date: Mon, 13 Nov 2000 04:16:00 -0000 Subject: Building on SGI with IRIX 6.2 References: Message-ID: On Mon, 13 Nov 2000, Reiner Suikat wrote: > Hello, > > I'm trying to build Sourcenav 4.5.2 on an SGI Indigo 2 workstation running > IRIX 6.2. I had no success using SGI's compiler, but did get a lot farther > with gcc (version 2.8.1). > > Unfortunately, it doesn't work all the way, it breaks here: [... errors ...] > > Any idea? Do I need a more recent version of gcc? Even better, is there a > binary for SGI available somewhere? > When I compiled Sourcenav 4.5.2 for Irix (Irix 6.5 on an Onyx2) I had to use the most recent gcc (version 2.95.2), SGI's CC and gcc v.2.8.1 gave errors. I do not now of any SGI binaries of Sourcenav, but You can find the latest precompiled (and install packed) gcc for SGI at http://freeware.sgi.com/ Hope this helps you. endre -- Endre M. Lidal (endre@cmr.no) Scientist, Advanced Computing Christian Michelsen Research AS (www.cmr.no) From bje@redhat.com Mon Nov 13 04:19:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 13 Nov 2000 04:19:00 -0000 Subject: Building on SGI with IRIX 6.2 References: Message-ID: I'm trying to build Sourcenav 4.5.2 on an SGI Indigo 2 workstation running IRIX 6.2. I had no success using SGI's compiler, but did get a lot farther with gcc (version 2.8.1). Is the configure script locating SGI's shipping version of lex? If you install GNU flex does the problem go away? Ben From bje@redhat.com Mon Nov 13 06:05:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 13 Nov 2000 06:05:00 -0000 Subject: Adding Ruby support to Source Navigator References: Message-ID: Yo Mo, Well, are you going to be around to help keep it working? I don't know that would have a problem with adding support for ruby, I just want to make sure we don't end up with an abandoned parser. I know Ben was thinking about what changes would be needed to move the syntax highlighter out of the core, but he would know more about that. I've already started -- I have the configury in place to build the syntax highlighters as shared libraries with libtool. The next step is to just add in the hooks to load these libraries with Tcl's "load" command at start-up. I'm trying to decide whether all libraries matching a specific pattern (say, "*high*.so" should be loaded, or whether they should be named explicitly in the configuration file. Thoughts? At any rate, this'll be a vast improvement. Ben From bje@redhat.com Mon Nov 13 06:31:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 13 Nov 2000 06:31:00 -0000 Subject: [Patch] Remove warning in dbsym.c Message-ID: <200011131431.eADEVD526656@scooby.cygnus.com> The following patch eliminates a warning from GCC about `ret' being possibly unused. The warning is correct. 2000-11-14 Ben Elliston * dbsym.c (dbisempty): Initialise local variable to allay compiler warnings. *** dbsym.c 2000/04/20 00:38:38 1.31 --- dbsym.c 2000/11/13 14:28:17 *************** *** 413,419 **** DB *db_exclude = pars->db_exclude; DBT data; DBT key; ! int ret; int flag; int cou; int len; --- 413,419 ---- DB *db_exclude = pars->db_exclude; DBT data; DBT key; ! int ret = 0; int flag; int cou; int len; From spolk@redhat.com Mon Nov 13 08:42:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 13 Nov 2000 08:42:00 -0000 Subject: Building on SGI with IRIX 6.2 References: Message-ID: <3A101AD3.8AD140A1@redhat.com> It looks like you need GNU flex and bison as well, instead of the lex and yacc that ship with SGI. Reiner Suikat wrote: > > Hello, > > I'm trying to build Sourcenav 4.5.2 on an SGI Indigo 2 workstation running > IRIX 6.2. I had no success using SGI's compiler, but did get a lot farther > with gcc (version 2.8.1). > > Unfortunately, it doesn't work all the way, it breaks here: > gcc -DHAVE_CONFIG_H -I. -I../../../SN452-source/snavigator/hyper -I.. -Wall > -fwritable-strings -I../../../SN452-source/snavigator/misc/libutils -I/usr/ > people/suikat/SN452-source/tk8.1/generic -I/usr/people/suikat/SN452-source/t > k8.1/generic/../unix -I/usr/people/suikat/SN452-source/tk8.1/generic/../win > -I/usr/people/suikat/SN452-source/tcl8.1/generic -I../../../SN452-source/sn > avigator/../db/include -I../../../SN452-source/snavigator/db -I../../../SN45 > 2-source/snavigator/../libgui/src -g -O2 -c tcl_high.c > ../../../SN452-source/snavigator/hyper/tcl_high.l: In function > `tcl_highlight_lex': > ../../../SN452-source/snavigator/hyper/tcl_high.l:55: warning: subscript has > type `char' > ../../../SN452-source/snavigator/hyper/tcl_high.l: In function > `flush_lex_scanner': > ../../../SN452-source/snavigator/hyper/tcl_high.l:280: `YY_FLUSH_BUFFER' > undeclared (first use in this function) > ../../../SN452-source/snavigator/hyper/tcl_high.l:280: (Each undeclared > identifier is reported only once > ../../../SN452-source/snavigator/hyper/tcl_high.l:280: for each function it > appears in.) > ../../../SN452-source/snavigator/hyper/tcl_high.l: At top level: > ../../../SN452-source/snavigator/hyper/tcl_high.l:561: warning: `yyunput' > defined but not used > ../../../SN452-source/snavigator/hyper/tcl_high.l:812: warning: > `yy_push_state' defined but not used > ../../../SN452-source/snavigator/hyper/tcl_high.l:839: warning: > `yy_pop_state' defined but not used > ../../../SN452-source/snavigator/hyper/tcl_high.l:848: warning: > `yy_top_state' defined but not used > tcl_high.c:239: warning: `yy_flex_strcpy' declared `static' but never > defined > make[3]: *** [tcl_high.o] Error 1 > make[3]: Leaving directory `/usr/people/suikat/tmp/snavigator/hyper' > make[2]: *** [all-recursive] Error 1 > make[2]: Leaving directory `/usr/people/suikat/tmp/snavigator' > make[1]: *** [all-recursive-am] Error 2 > make[1]: Leaving directory `/usr/people/suikat/tmp/snavigator' > make: *** [all-snavigator] Error 2 > > Any idea? Do I need a more recent version of gcc? Even better, is there a > binary for SGI available somewhere? > > Thanks > > Reiner From spolk@redhat.com Mon Nov 13 08:47:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 13 Nov 2000 08:47:00 -0000 Subject: Adding Ruby support to Source Navigator References: Message-ID: <3A101BE4.129FADE4@redhat.com> Ben Elliston wrote: > > Yo Mo, > > Well, are you going to be around to help keep it working? I don't know > that would have a problem with adding support for ruby, I just want to > make sure we don't end up with an abandoned parser. I know Ben was > thinking about what changes would be needed to move the syntax > highlighter out of the core, but he would know more about that. > > I've already started -- I have the configury in place to build the syntax > highlighters as shared libraries with libtool. The next step is to just add > in the hooks to load these libraries with Tcl's "load" command at start-up. > > I'm trying to decide whether all libraries matching a specific pattern (say, > "*high*.so" should be loaded, or whether they should be named explicitly in > the configuration file. Thoughts? Hmm. I hate configuration files. I would vote for a pattern. Something like: sn*high*.so on unix and sn*high*.dll on windows. > > At any rate, this'll be a vast improvement. > > Ben From bje@redhat.com Mon Nov 13 14:11:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 13 Nov 2000 14:11:00 -0000 Subject: Adding Ruby support to Source Navigator References: <3A101BE4.129FADE4@redhat.com> Message-ID: > I'm trying to decide whether all libraries matching a specific pattern (say, > "*high*.so" should be loaded, or whether they should be named explicitly in > the configuration file. Thoughts? Hmm. I hate configuration files. I would vote for a pattern. Something like: sn*high*.so on unix and sn*high*.dll on windows. What about the problems whereby users get confused by implict behaviour? I have implicit behaviour. :-) Ben From irox@redhat.com Mon Nov 13 14:28:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Mon, 13 Nov 2000 14:28:00 -0000 Subject: Adding Ruby support to Source Navigator Message-ID: <200011132221.OAA12918@manx.soma.redhat.com> On Tue, 14 Nov 2000, Ben Elliston wrote: > Hmm. I hate configuration files. I would vote for a pattern. > Something like: > sn*high*.so on unix > and > sn*high*.dll on windows. > > What about the problems whereby users get confused by implict behaviour? > I have implicit behaviour. :-) I'm kind of a fan of having a directory for putting things like highing libs in. SN can then just load all valid shared libraries in this directory. This is how the new 5.0 toolchain support will work. Ian. From emthornber@iee.org Tue Nov 14 08:55:00 2000 From: emthornber@iee.org (Mark Thornber) Date: Tue, 14 Nov 2000 08:55:00 -0000 Subject: xref and referred by? Message-ID: <3A117C62.23ACA905@iee.org> Following on from Dave Banham's e-mail and the request for a test case from Mo DeJong - the c++_demo project in the sn distribution exhibits the problem. xref of 'int Agent(mi) Agent(Sequencer *)' shows 'w agent_value(ud)' which cannot be found in the symbol table but is in glish/Agent.h Line 123 "Value* agent_value;". (There are other undefined symbols in the same xref display.) --Mark Thornber From dave.banham@tde.alstom.com Wed Nov 15 04:57:00 2000 From: dave.banham@tde.alstom.com (dave.banham@tde.alstom.com) Date: Wed, 15 Nov 2000 04:57:00 -0000 Subject: xref and referred by? Message-ID: Mark, Thanks for finding an example test case for this problem - I have been on holiday for the last three weeks and now I'm burried under a mountain of stuff to be done by yesterday! I trust that those of you with the dark knowledge of SN will be able to find and fix this problem. Regards Dave Banham Mark Thornber on 14/11/2000 17:54:42 To: sourcenav@sources.redhat.com cc: (bcc: Dave BANHAM/PCS/PCG/TDE/GECALSTHOM) Subject: Re: xref and referred by? Following on from Dave Banham's e-mail and the request for a test case from Mo DeJong - the c++_demo project in the sn distribution exhibits the problem. xref of 'int Agent(mi) Agent(Sequencer *)' shows 'w agent_value(ud)' which cannot be found in the symbol table but is in glish/Agent.h Line 123 "Value* agent_value;". (There are other undefined symbols in the same xref display.) --Mark Thornber From fnasser@cygnus.com Thu Nov 16 05:37:00 2000 From: fnasser@cygnus.com (Fernando Nasser) Date: Thu, 16 Nov 2000 05:37:00 -0000 Subject: debug error, enable-external-editor Message-ID: <3A13E2DD.8EB67F43@cygnus.com> (I am posting this message in behalf of Chris Ruth) Hi, I'm new to this world. Ex-windows programmer thrown into Linux-land. Please help. I installed source-navigator4.5.2 and insight-5.0. I am trying to use the debugger from the 'Build' window in source-navigator. When I run 'Debug' from that window I get the following error from gdb, 'unrecognized option --enable-external-editor'. Do you know how to fix this? Yes, I read the previous postings on this but I'm still confused as to what i need to download to get it working. thanks, -- Christopher Ruth, Ph.D. Hologic, Inc. 35 Crosby Drive Bedford, MA 01730 Phone: (781) 999-7575 Fax: (781) 280-0614 email: cruth@hologic.com From mdejong@cygnus.com Thu Nov 16 13:26:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Thu, 16 Nov 2000 13:26:00 -0000 Subject: debug error, enable-external-editor References: <3A13E2DD.8EB67F43@cygnus.com> Message-ID: On Thu, 16 Nov 2000, Fernando Nasser wrote: > (I am posting this message in behalf of Chris Ruth) > > Hi, > I'm new to this world. Ex-windows programmer thrown into Linux-land. > Please help. > I installed source-navigator4.5.2 and insight-5.0. I am trying to use > the > debugger from the 'Build' window in source-navigator. When I run 'Debug' > > from that window I get the following error from gdb, 'unrecognized > option --enable-external-editor'. Do you know how to fix this? Yes, I > read the previous postings on this but I'm still confused as to what i > need to download to get it working. > thanks, You don't have insight on your PATH. SN is finding and execing the regular gdb and that is causing the error. Make sure insight appears on the PATH before /usr/bin/gdb. Mo DeJong Red Hat Inc From morreale@indstorage.com Thu Nov 16 14:08:00 2000 From: morreale@indstorage.com (Peter W. Morreale) Date: Thu, 16 Nov 2000 14:08:00 -0000 Subject: SN and kernel 2.4.0-test1 Message-ID: <3A145A76.21ECAA64@indstorage.com> Anybody use sn-4.5.2 against kernel 2.4.0-test1? While building the dbs, it appears to go into an infinite loop on .../linux/arch/arm/kernel/arch.c my cpu./mem applet jumps to 98% user cpu time for as long as I let cbrowser live. arch.s is only 10k lines.... oh yeah, RH 7.0, stoc ... Thanks, -PWM From echristo@cygnus.com Thu Nov 16 14:20:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Thu, 16 Nov 2000 14:20:00 -0000 Subject: SN and kernel 2.4.0-test1 References: <3A145A76.21ECAA64@indstorage.com> Message-ID: <3A146061.A3314EE3@cygnus.com> "Peter W. Morreale" wrote: > > Anybody use sn-4.5.2 against kernel 2.4.0-test1? > > While building the dbs, it appears to go into an infinite loop on > .../linux/arch/arm/kernel/arch.c > > my cpu./mem applet jumps to 98% user cpu time for as long as I let > cbrowser live. arch.s is only > 10k lines.... > > oh yeah, RH 7.0, stoc ... > arch.c or arch.s? If it is arch.s I doubt it will work since (afair) SN doesn't have an arm assembly parser. You'll probably need to unload those files. -eric From david.autry@marconimed.com Thu Nov 16 15:26:00 2000 From: david.autry@marconimed.com (David Autry) Date: Thu, 16 Nov 2000 15:26:00 -0000 Subject: SourceNav with Windows 2K problems Message-ID: <012101c05024$4d57fa70$be633690@autry> Hi, I have found a few problems using SourceNav Version 4.5.2 with Windows 2K. Im using Exceed for Win32, Version 6.2.0.16 as my X environment. 1. Whenever SourceNav is working with a large number of files, such as scanning a project or greping, the mouse cursor must be in a SourceNav window or it will stop processing files. I can watch the CPU usage drop to zero and hear the hard drive stop. Give focus to a SourceNav window and everything starts up again. 2. In the Symbols window, sometimes (most) I cannot resize the File, Directory, and Type columns. The cursor never changes to the double arrow. When I can resize them, they revert to their default size the next time SourceNav is started. Is there a way to save these settings? 3. There is no control of font size/type when printing from Windows. Is this feature planned for a future release? Thanks. David Autry ---- +------------------------------+--------------------------------------+ | David Autry | Internet: dautry@mr.marconimed.com | | Marconi Medical Systems | Voice: (440) 483-4195 | | 595 Miner Road | Fax: (440) 483-5728 | | Cleveland, Ohio, USA 44143 | | +------------------------------+--------------------------------------+ From mdejong@cygnus.com Thu Nov 16 15:43:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Thu, 16 Nov 2000 15:43:00 -0000 Subject: SourceNav with Windows 2K problems References: <012101c05024$4d57fa70$be633690@autry> Message-ID: On Thu, 16 Nov 2000, David Autry wrote: > Hi, > > I have found a few problems using SourceNav Version 4.5.2 with Windows 2K. > Im using Exceed for Win32, Version 6.2.0.16 as my X environment. Are you running SN on some remote system and displaying it on you Windows box with Exceed? You do know that there is a Windows native version of SN, right? If this is your configuration, I suggest that you try to remote display to a Linux or other Unix box and see if that fixes your problem. It might just be a problem with your Exceed server. Mo DeJong Red Hat Inc From morri70@attglobal.net Thu Nov 16 22:19:00 2000 From: morri70@attglobal.net (c.f. morrison) Date: Thu, 16 Nov 2000 22:19:00 -0000 Subject: Solaris 8(Intel) Build Message-ID: <3A14CE96.773FA7B1@attglobal.net> Hi, Is it possible to build Source Navigator on Solaris8/Intel. I built it with no problems on Linux but 'configure' in Solaris/Intel fails with the error ' can not find sources in /opt/sfw/SN452-source/snavigator. I love SN & would dearly love to have it on my Solaris/Intel box. Thanks in advance cf morrison From echristo@cygnus.com Thu Nov 16 22:30:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Thu, 16 Nov 2000 22:30:00 -0000 Subject: Solaris 8(Intel) Build References: <3A14CE96.773FA7B1@attglobal.net> Message-ID: <3A14D347.3635EB20@cygnus.com> "c.f. morrison" wrote: > > Hi, > Is it possible to build Source Navigator on Solaris8/Intel. I built > it with no problems on Linux but 'configure' in Solaris/Intel fails with > the error > ' can not find sources in /opt/sfw/SN452-source/snavigator. > I love SN & would dearly love to have it on my Solaris/Intel box. How interesting. Perhaps that is a symlink to somewhere else? Can you try the configure and build on the same disk/partition? I don't have a Solaris8/Intel box afaik to attempt this on, so I can only help via remote :) -eric From bje@redhat.com Thu Nov 16 23:08:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 16 Nov 2000 23:08:00 -0000 Subject: Solaris 8(Intel) Build References: <3A14CE96.773FA7B1@attglobal.net> Message-ID: Is it possible to build Source Navigator on Solaris8/Intel. I built it with no problems on Linux but 'configure' in Solaris/Intel fails with the error ' can not find sources in /opt/sfw/SN452-source/snavigator. There's no reason why it should not work. How about doing this from your build directory: sh -v -x /path/to/sn/src/configure --any-options And look to see why it's failing. You might also like to try: /path/to/src/configure --srcdir=/path/to/src/ Ben From david.autry@marconimed.com Fri Nov 17 06:56:00 2000 From: david.autry@marconimed.com (David Autry) Date: Fri, 17 Nov 2000 06:56:00 -0000 Subject: SourceNav with Windows 2K problems Message-ID: <002001c050a6$2753f220$be633690@autry> I am using the native Windows version of SourceNav on Win 2K. I just mentioned having Exceed for Win32 as part of my system configuration and that maybe it might be contributing to the window focus problem. As a test, I disabled Exceed and then did a grep. Same problem. Move the cursor out of a SN window, and all file activity stops. Move the cursor back in and SN starts processing files again. It is interesting to note that the SN window doesn't have to be selected to start processing files, just having the cursor is enough. David Autry Marconi Medical Systems ----- Original Message ----- From: Mo DeJong To: sourcenav@sourceware.cygnus.com Sent: Thursday, November 16, 2000 6:43 PM Subject: Re: SourceNav with Windows 2K problems On Thu, 16 Nov 2000, David Autry wrote: > Hi, > > I have found a few problems using SourceNav Version 4.5.2 with Windows 2K. > Im using Exceed for Win32, Version 6.2.0.16 as my X environment. Are you running SN on some remote system and displaying it on you Windows box with Exceed? You do know that there is a Windows native version of SN, right? If this is your configuration, I suggest that you try to remote display to a Linux or other Unix box and see if that fixes your problem. It might just be a problem with your Exceed server. Mo DeJong Red Hat Inc From BBurton@3dfx.com Fri Nov 17 09:21:00 2000 From: BBurton@3dfx.com (Brent Burton) Date: Fri, 17 Nov 2000 09:21:00 -0000 Subject: SourceNav with Windows 2K problems Message-ID: <2C8579D633A0D3119C28009027AF523302BF31D8@exchange1.stb.com> I have seen this too, also using native 4.5.2 on Win2k. But what I noticed is that not only must the pointer be in the window, it must move for SN to update any indicators. There's a project that's used daily, so I'll refresh it. I selected Refresh Project from the Tools menu, and select another app. Here's a description of what occurs: Symbol Window: the "Deleting" progress bar appears, disk activity is high, but progress bar doesn't move. Moving the pointer into the window allows the bar to zip across, completing. The Rescanning Project dialog appears: As David mentions, file activity stops when the pointer is outside the window. I noticed that it also stopped if the pointer was in the window but didn't' move. As soon as it moves, progress continues. Symbol Window: the "Deleting" progress bar reappears, disk activity is high, and the progress bar doesn't move. I let it sit and after about 5 minutes it jumped to 36%, disk activity was again high, then paused until the pointer entered. The "xref" progress bar, when it appeared, behaved similarly. I had noticed this some time ago, even with versions 4.1 and 4.2. Figuring it to be a quirk with Tcl/Tk on Windows, I lived with it. Maybe Ben or Mo knows how these are updated (update idletasks?) and maybe the issue will go away with Tk 8.x. Hope this info helps. -Brent -----Original Message----- From: David Autry [ mailto:david.autry@marconimed.com ] Sent: Friday, November 17, 2000 8:54 AM To: sourcenav@sourceware.cygnus.com Subject: Re: SourceNav with Windows 2K problems I am using the native Windows version of SourceNav on Win 2K. I just mentioned having Exceed for Win32 as part of my system configuration and that maybe it might be contributing to the window focus problem. As a test, I disabled Exceed and then did a grep. Same problem. Move the cursor out of a SN window, and all file activity stops. Move the cursor back in and SN starts processing files again. It is interesting to note that the SN window doesn't have to be selected to start processing files, just having the cursor is enough. David Autry Marconi Medical Systems ----- Original Message ----- From: Mo DeJong To: sourcenav@sourceware.cygnus.com Sent: Thursday, November 16, 2000 6:43 PM Subject: Re: SourceNav with Windows 2K problems On Thu, 16 Nov 2000, David Autry wrote: > Hi, > > I have found a few problems using SourceNav Version 4.5.2 with Windows 2K. > Im using Exceed for Win32, Version 6.2.0.16 as my X environment. Are you running SN on some remote system and displaying it on you Windows box with Exceed? You do know that there is a Windows native version of SN, right? If this is your configuration, I suggest that you try to remote display to a Linux or other Unix box and see if that fixes your problem. It might just be a problem with your Exceed server. Mo DeJong Red Hat Inc From morri70@ieee.org Sat Nov 18 12:48:00 2000 From: morri70@ieee.org (c.f. morrison) Date: Sat, 18 Nov 2000 12:48:00 -0000 Subject: Solaris 8(Intel) Build References: <3A14CE96.773FA7B1@attglobal.net> <3A14D347.3635EB20@cygnus.com> Message-ID: <3A16EBC7.CBDE831F@ieee.org> Eric, I should have made clear that I use /opt/sfw/SN452-source as my source tree and /opt/sfw/SN452-build for my build. Thanks for response cfm Eric Christopher wrote: > "c.f. morrison" wrote: > > > > Hi, > > Is it possible to build Source Navigator on Solaris8/Intel. I built > > it with no problems on Linux but 'configure' in Solaris/Intel fails with > > the error > > ' can not find sources in /opt/sfw/SN452-source/snavigator. > > I love SN & would dearly love to have it on my Solaris/Intel box. > > How interesting. Perhaps that is a symlink to somewhere else? Can you > try the configure and build on the same disk/partition? > > I don't have a Solaris8/Intel box afaik to attempt this on, so I can > only help via remote :) > > -eric From morri70@ieee.org Sun Nov 19 17:23:00 2000 From: morri70@ieee.org (c.f. morrison) Date: Sun, 19 Nov 2000 17:23:00 -0000 Subject: Solaris 8(Intel) Build References: <3A14CE96.773FA7B1@attglobal.net> Message-ID: <3A187DB4.937F6886@ieee.org> I would like to put on record that Source Navigator builds fine on Solaris8/Intel, It does demand using the GNU tar on the source package AND gmake (the Solaris make had problems with one of the makefiles). I'm now rather happy & thanks to Ben & Eric for their responses. cfm "c.f. morrison" wrote: > Hi, > Is it possible to build Source Navigator on Solaris8/Intel. I built > it with no problems on Linux but 'configure' in Solaris/Intel fails with > the error > ' can not find sources in /opt/sfw/SN452-source/snavigator. > I love SN & would dearly love to have it on my Solaris/Intel box. From mdejong@cygnus.com Sun Nov 19 17:28:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sun, 19 Nov 2000 17:28:00 -0000 Subject: Solaris 8(Intel) Build References: <3A187DB4.937F6886@ieee.org> Message-ID: On Sun, 19 Nov 2000, c.f. morrison wrote: > I would like to put on record that Source Navigator builds fine on > Solaris8/Intel, It does demand using the GNU tar on the source package AND > gmake (the Solaris make had problems with one of the makefiles). I'm now > rather happy & thanks to Ben & Eric for their responses. > cfm > > "c.f. morrison" wrote: As far as I know, having both GNU tar and GNU make is a requirement for building Source-Navigator on any platform (including Windows). I just wanted to state that for the record so that nobody else tries to use the vendor supplied (cough, always broken, cough) tools to build SN. Mo DeJong Red Hat Inc From bje@redhat.com Sun Nov 19 17:29:00 2000 From: bje@redhat.com (Ben Elliston) Date: Sun, 19 Nov 2000 17:29:00 -0000 Subject: Solaris 8(Intel) Build References: <3A187DB4.937F6886@ieee.org> Message-ID: I would like to put on record that Source Navigator builds fine on Solaris8/Intel, It does demand using the GNU tar on the source package AND gmake (the Solaris make had problems with one of the makefiles). I'm now rather happy & thanks to Ben & Eric for their responses. Building a tar file that can be unpacked with older Unix tar programs is a simple matter of packaging. I assume this didn't happen -- Syd? As for the make problem -- this should not happen, either. S-N uses Automake, which generates portable Makefiles. Can you post the error that you got so that we can fix such problems? Cheers, Ben From mdejong@cygnus.com Sun Nov 19 17:34:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Sun, 19 Nov 2000 17:34:00 -0000 Subject: Solaris 8(Intel) Build References: Message-ID: On Mon, 20 Nov 2000, Ben Elliston wrote: > I would like to put on record that Source Navigator builds fine on > Solaris8/Intel, It does demand using the GNU tar on the source package > AND gmake (the Solaris make had problems with one of the makefiles). > I'm now rather happy & thanks to Ben & Eric for their responses. > > Building a tar file that can be unpacked with older Unix tar programs is a > simple matter of packaging. I assume this didn't happen -- Syd? > > As for the make problem -- this should not happen, either. S-N uses > Automake, which generates portable Makefiles. Can you post the error that > you got so that we can fix such problems? It does? Last I heard, automake generated makefiles would not run with BSD make. I have not run test myself, this is just what I have seen users report on other projects. Mo DeJong Red Hat Inc From s.contini@fol.it Mon Nov 20 04:47:00 2000 From: s.contini@fol.it (Simone Contini) Date: Mon, 20 Nov 2000 04:47:00 -0000 Subject: future plans Message-ID: <20001120125555.A6847@radagast> Hi, Source Navigator is very Cool!! thank you for it! I'd like to know how i can read about future plans and if the projects will take advantages of GTK/GNOME. thank you Simone Contini Games & Tools Programmer Florence - Italy Email: s.contini@fol.it WWW: http://3dfx4ever.oltrelinux.com/ http://scontini.xoasis.com/ ICQ # : 25879128 From tromey@cygnus.com Mon Nov 20 12:01:00 2000 From: tromey@cygnus.com (Tom Tromey) Date: Mon, 20 Nov 2000 12:01:00 -0000 Subject: Solaris 8(Intel) Build References: Message-ID: <87itpjnrn9.fsf@creche.cygnus.com> >>>>> "Mo" == Mo DeJong writes: Mo> It does? Last I heard, automake generated makefiles would not run Mo> with BSD make. False. It is true if you enable dependency tracking with automake 1.4. However, automake 1.4 was designed so that end users will never enable this feature. `make dist' disables it. Tom From spolk@redhat.com Mon Nov 20 12:02:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 20 Nov 2000 12:02:00 -0000 Subject: future plans References: <20001120125555.A6847@radagast> Message-ID: <4.2.0.58.20001120120413.00cb5720@pop.cygnus.com> At 12:55 PM 11/20/00 +0100, Simone Contini wrote: >Hi, > > Source Navigator is very Cool!! thank you for it! I'd like to know >how i can read about future plans and if the projects will take advantages >of GTK/GNOME. We are in the process of evaluating our future strategy. However, it is highly unlikely we will ever do Source-Navigator in GTK or GNOME. We might add the capability of communicating with the GTK/GNOME desktop services, but these would be Tcl/Tk modifications. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mdejong@cygnus.com Mon Nov 20 15:24:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Mon, 20 Nov 2000 15:24:00 -0000 Subject: future plans References: <20001120125555.A6847@radagast> Message-ID: On Mon, 20 Nov 2000, Simone Contini wrote: > Hi, > > Source Navigator is very Cool!! thank you for it! I'd like to know > how i can read about future plans and if the projects will take advantages > of GTK/GNOME. What features from Gtk or Gnome do you see as a "must have". The ability to do themes might be neat, but I don't see it as a critical component of a good IDE. Gnome has some configuration management stuff that might be handy if we needed it, but I don't think we do. If you have specific things you were thinking of, I would like to hear them. Keep in mind, that we can't depend on a Gnome feature because SN need to run under Windows too. Mo DeJong Red Hat Inc From timo@qeradiant.com Mon Nov 20 15:37:00 2000 From: timo@qeradiant.com (Timothee Besset) Date: Mon, 20 Nov 2000 15:37:00 -0000 Subject: future plans References: Message-ID: <3A19B59C.4070007@qeradiant.com> When can we expect an update to SN? The most important feature (imo) would be auto-completion. Auto-completion of symbols, listing the members when accessing a struct with -> or . Displaying the definition of the functions in a tooltip box when the mouse comes over them Displaying possible commentaries that are typed before struct definitions or function definitons Maybe even a window that would sit somewhere on your workspade and update to the class definition or inheritance of the class you are working in (well yeah, stuff you can get already, but it's not implemented in a very efficient way) I'm also lacking some docs or tips, maybe some of the stuff I want is there already. The behaviour of the windows is weird. If the file as been edited it would open a new window, but if not it would change the edit view. I didn't find the shortcuts to switch between editor views (like Alt+TAB on MSVC). I'm free to develop on win32 or linux. I prefer linux because on win98 my app kills the GDI resources and I have to reboot quite often. But I'm strongly tempted to go back to win32 cause I don't feel as efficient with sourcenav as I can be with MSDev Hope this helps.. TTimo Mo DeJong wrote: > On Mon, 20 Nov 2000, Simone Contini wrote: > >> Hi, >> >> Source Navigator is very Cool!! thank you for it! I'd like to know >> how i can read about future plans and if the projects will take advantages >> of GTK/GNOME. > > > What features from Gtk or Gnome do you see as a "must have". > The ability to do themes might be neat, but I don't see it > as a critical component of a good IDE. Gnome has some > configuration management stuff that might be handy if we > needed it, but I don't think we do. If you have specific > things you were thinking of, I would like to hear them. > Keep in mind, that we can't depend on a Gnome feature > because SN need to run under Windows too. > > Mo DeJong > Red Hat Inc > > > From tromey@cygnus.com Mon Nov 20 16:13:00 2000 From: tromey@cygnus.com (Tom Tromey) Date: Mon, 20 Nov 2000 16:13:00 -0000 Subject: future plans References: Message-ID: <871yw6i3ag.fsf@creche.cygnus.com> >>>>> "Mo" == Mo DeJong writes: Mo> What features from Gtk or Gnome do you see as a "must have". Mo> The ability to do themes might be neat, but I don't see it Mo> as a critical component of a good IDE. Gnome has some Mo> configuration management stuff that might be handy if we Mo> needed it, but I don't think we do. Themes are nice, but they are just candy. Most people would rather have a tool that works well than one that conforms to their environment. (Well, most people would prefer both, I'd imagine...) The config code is nice, and will get much nicer (with GConf), but unless you want network transparency and the use of weird back ends it isn't very important. There are two more important things that Gnome has that S-N currently can't do: * Drag and drop * Session management Both of these can be written as Tcl extensions though. The D&D protocol is (I believe) well-documented and in use by a few groups (both Gnome and KDE at least). (I know you've argued against D&D before, but I've never understood your argument.) Gnome's session management is just XSMP. There is an (ugly) X library to help you do this; Gnome just wraps it with a nicer API. (At some point session management and configuration storage might tie together somehow. That would be harder to emulate.) Gnome also has some other things, like the ability to play sounds in response to certain events (activate a button widget, get a beep). This is along the lines of themability though -- not deeply important. The really important Gnome stuff comes later: componentization of everything. Maybe it will be possible to use this from Tcl, too, with some work. I don't know. Maybe components aren't important for an IDE. I think they could be used though. For instance the debugger gui could be plugged in to S-N somehow. Tom From morri70@ieee.org Mon Nov 20 22:56:00 2000 From: morri70@ieee.org (c.f. morrison) Date: Mon, 20 Nov 2000 22:56:00 -0000 Subject: Solaris 8(Intel) Build References: Message-ID: <3A1A1CD4.E142AD16@ieee.org> Ben, Here is the error in Solaris make ... I hope I've provided enough! cfm Ben Elliston wrote: > As for the make problem -- this should not happen, either. S-N uses > Automake, which generates portable Makefiles. Can you post the error that > you got so that we can fix such problems? --/usr/local/ is linked to /opt/sfw/ --the makefile for opt/sfw/SN452-build/itcl/src contains this (see Line 75) ************************************************************ .............(lots of preceeding stuff) # Tcl library # (the configure script should find it; if not, edit below) TCL_LIB = -L/opt/sfw/SN452-build/tcl8.1/unix -ltcl8.1 -ldl -lsocket -lnsl -lm # Auxiliary stuff for the Tcl library. LIB_INSTALL_DIR = ${exec_prefix}/lib # Tk library # (the configure script should find it; if not, edit below) TK_LIB = -L/opt/sfw/SN452-build/tk8.1/unix -ltk8.1 # Linker switch(es) to use to link with the X11 library archive (the # configure script will try to set this value automatically, but you # can override it). X11_LIB = -L/usr/openwin/lib -lX11 # Libraries to use when linking: must include at least Tk, Tcl, Xlib, # and the math library (in that order). The "-lm " part will be # replaced (or has already been replaced) with relevant libraries as # determined by the configure script. ifeq ($(OBJEXT),obj) LIBS = $(TK_LIB) $(TCL_LIB) ********* LINE 75 else LIBS = $(TK_LIB) $(TCL_LIB) $(X11_LIB) -lm -lm endif ************************************************************************** -- when running Solaris make with -dd option I get this ************************************************************************** .............(lots of preceeding stuff) stat(Makefile) doname(Makefile) Looking for % rule for Makefile find_double_suffix_rule(Makefile) find_suffix_rule(Makefile,Makefile,) stat(Makefile) ROOTNAME= itcl VERSION= 1.5 PACKAGE= $(ROOTNAME)-$(VERSION) OBJEXT= o prefix= /usr/local exec_prefix= ${prefix} LIB_DIR= ${exec_prefix}/lib INCLUDE_DIR= ${prefix}/include BIN_DIR= ${exec_prefix}/bin TCL_INC= -I/opt/sfw/SN452-source/tcl8.1/generic TK_INC= -I/opt/sfw/SN452-source/tk8.1/generic X11_INC= -I/usr/openwin/include TCL_LIB= -L/opt/sfw/SN452-build/tcl8.1/unix -ltcl8.1 -ldl -lsocket -lnsl -lm LIB_INSTALL_DIR= ${exec_prefix}/lib TK_LIB= -L/opt/sfw/SN452-build/tk8.1/unix -ltk8.1 X11_LIB= -L/usr/openwin/lib -lX11 make: Fatal error in reader: Makefile, line 75: Unexpected end of line seen Current working directory /opt/sfw/SN452-build/itcl/src stat(.DONE) doname(.DONE) Looking for % rule for .DONE find_double_suffix_rule(.DONE) *** Error code 1 stat(.make.state) stat(../../SN452-source/itcl/.make.state) stat(.make.state) make: Fatal error: Command failed for target `all' Current working directory /opt/sfw/SN452-build/itcl stat(.DONE) stat(../../SN452-source/itcl/.DONE) doname(.DONE) Looking for % rule for .DONE find_double_suffix_rule(.DONE) *** Error code 1 stat(.make.state) stat(.make.state) make: Fatal error: From mdejong@cygnus.com Tue Nov 21 11:02:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 21 Nov 2000 11:02:00 -0000 Subject: future plans References: <3A19B59C.4070007@qeradiant.com> Message-ID: On Mon, 20 Nov 2000, Timothee Besset wrote: > When can we expect an update to SN? What kind of "update" were you looking for? We are working on getting a snapshot of 5.0 out sometime soon, but that will not even be alpha quality. It would be a "hackers only" pre-release. > The most important feature (imo) would be auto-completion. > Auto-completion of symbols, listing the members when accessing a struct > with -> or . That would be neat. Are you willing to work on implementing this feature? > Displaying the definition of the functions in a tooltip box when the > mouse comes over them This should be trivial, perhaps a day or two to implement and test. > Displaying possible commentaries that are typed before struct > definitions or function definitons Not sure what you mean by this. > Maybe even a window that would sit somewhere on your workspade and > update to the class definition or inheritance of the class you are > working in (well yeah, stuff you can get already, but it's not > implemented in a very efficient way) Is it so hard to switch to the Hier window or the class viewer? > I'm also lacking some docs or tips, maybe some of the stuff I want is > there already. > The behaviour of the windows is weird. If the file as been edited it > would open a new window, but if not it would change the edit view. I > didn't find the shortcuts to switch between editor views (like Alt+TAB > on MSVC). > I'm free to develop on win32 or linux. I prefer linux because on win98 > my app kills the GDI resources and I have to reboot quite often. But I'm > strongly tempted to go back to win32 cause I don't feel as efficient > with sourcenav as I can be with MSDev Well, feelings are a tricky thing. They might be good or they might have nothing to do with reality. When it comes to a GUI, folks start to get "bad feelings" over little things like a window taking too long to redraw. Don't forget that you do have the source, you can adapt it to better suite your "feelings". When was the last time you were able to reprogram MSDev to get around something you did not like? Mo DeJong Red Hat Inc From spolk@redhat.com Tue Nov 21 12:22:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 21 Nov 2000 12:22:00 -0000 Subject: future plans References: <"Mon,> <20> <2000> <15:23:58> <-0800> <(PST)> <871yw6i3ag.fsf@creche.cygnus.com> Message-ID: <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> >There are two more important things that Gnome has that S-N currently >can't do: > >* Drag and drop >* Session management > >Both of these can be written as Tcl extensions though. > >The D&D protocol is (I believe) well-documented and in use by a few >groups (both Gnome and KDE at least). (I know you've argued against >D&D before, but I've never understood your argument.) I don't remember arguing against the idea of drag-and-drop. It is a great feature which I would love to have. >Gnome's session management is just XSMP. There is an (ugly) X library >to help you do this; Gnome just wraps it with a nicer API. (At some >point session management and configuration storage might tie together >somehow. That would be harder to emulate.) I don't know what "session management" means. >Gnome also has some other things, like the ability to play sounds in >response to certain events (activate a button widget, get a beep). >This is along the lines of themability though -- not deeply important. > > >The really important Gnome stuff comes later: componentization of >everything. Maybe it will be possible to use this from Tcl, too, with >some work. I don't know. > >Maybe components aren't important for an IDE. I think they could be >used though. For instance the debugger gui could be plugged in to S-N >somehow. The problem is, of course, that this is a major rewrite of something. Either Tk has to be fitted with it, or we have to rewrite the entire GUI in GNOME. I would rather not spend several man-years doing the latter. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From mdejong@cygnus.com Tue Nov 21 12:31:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 21 Nov 2000 12:31:00 -0000 Subject: future plans References: <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> Message-ID: On Tue, 21 Nov 2000, Syd Polk wrote: > > >There are two more important things that Gnome has that S-N currently > >can't do: > > > >* Drag and drop > >* Session management > > > >Both of these can be written as Tcl extensions though. > > > >The D&D protocol is (I believe) well-documented and in use by a few > >groups (both Gnome and KDE at least). (I know you've argued against > >D&D before, but I've never understood your argument.) > > I don't remember arguing against the idea of drag-and-drop. It is a great > feature which I would love to have. Syd, I think he was talking about me. It is true, I have seen some drag and drop interfaces that are just stupid and a waste of time. On the other hand, I have seen some good uses of drag and drop too. For example, on the Mac you can drag a URL out of the web browser and put it on the desktop. It turns into a click-able thingy that will open up the web browser at the given URL later. Sure, it is just a bookmark but it was a reasonable interface. I have also seen some stupid uses of drap and drop, where it just gets in the way instead of helping. One thing that might be handy in SN is to drag a set of new files into the symbol browser, that could add the given files to the project. Of course, we already provide a way to add a directory or set of files to the project so I don't think anyone is going to be beating down the door for that feature. > >Gnome's session management is just XSMP. There is an (ugly) X library > >to help you do this; Gnome just wraps it with a nicer API. (At some > >point session management and configuration storage might tie together > >somehow. That would be harder to emulate.) > > I don't know what "session management" means. Well, for starters SN could open up with the windows in the same place and open on the same tools as when it shut down. Some folks really like this sort of interface, others could not care less. Mo DeJong Red Hat Inc From tromey@cygnus.com Tue Nov 21 12:32:00 2000 From: tromey@cygnus.com (Tom Tromey) Date: Tue, 21 Nov 2000 12:32:00 -0000 Subject: future plans References: <"Mon,> <20> <2000> <15:23:58> <-0800> <(PST)> <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> Message-ID: <87y9yddpof.fsf@creche.cygnus.com> >> Gnome's session management is just XSMP. There is an (ugly) X library >> to help you do this; Gnome just wraps it with a nicer API. (At some >> point session management and configuration storage might tie together >> somehow. That would be harder to emulate.) Syd> I don't know what "session management" means. When you log out, Gnome can save the state of all applications (all those which support session management) and then later restart the session where you left off. The idea is applications will remember what documents were open, cursor positions, etc. Tom From bruce@cenderis.demon.co.uk Tue Nov 21 15:18:00 2000 From: bruce@cenderis.demon.co.uk (Bruce Stephens) Date: Tue, 21 Nov 2000 15:18:00 -0000 Subject: future plans References: <"Mon,> <20> <2000> <15:23:58> <-0800> <(PST)> <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> <87y9yddpof.fsf@creche.cygnus.com> Message-ID: <87y9yc6hkc.fsf@cenderis.demon.co.uk> Tom Tromey writes: > >> Gnome's session management is just XSMP. There is an (ugly) X library > >> to help you do this; Gnome just wraps it with a nicer API. (At some > >> point session management and configuration storage might tie together > >> somehow. That would be harder to emulate.) > > Syd> I don't know what "session management" means. > > When you log out, Gnome can save the state of all applications (all > those which support session management) and then later restart the > session where you left off. > > The idea is applications will remember what documents were open, > cursor positions, etc. I don't think this requires anything that Tk doesn't already have, does it? Can't an application just use "wm protocol WM_SAVE_YOURSELF" and do something to save its state? Hmm, I guess there's a bit more to it than that: to restore, it would need to be able to identify an instance of information, but overall it shouldn't be *that* much more. From spolk@redhat.com Tue Nov 21 15:41:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 21 Nov 2000 15:41:00 -0000 Subject: future plans References: <"Tue,> <21> <2000> <12:24:50> <-0800> <"Mon,> <20> <2000> <15:23:58> <-0800> <(PST)> <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> <87y9yddpof.fsf@creche.cygnus.com> Message-ID: <4.2.0.58.20001121154334.01a83450@pop.cygnus.com> At 01:40 PM 11/21/00 -0700, Tom Tromey wrote: > >> Gnome's session management is just XSMP. There is an (ugly) X library > >> to help you do this; Gnome just wraps it with a nicer API. (At some > >> point session management and configuration storage might tie together > >> somehow. That would be harder to emulate.) > >Syd> I don't know what "session management" means. > >When you log out, Gnome can save the state of all applications (all >those which support session management) and then later restart the >session where you left off. > >The idea is applications will remember what documents were open, >cursor positions, etc. > >Tom Source-Navigator does this, to a limited extent. It's buggy as all get out. Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From tromey@cygnus.com Tue Nov 21 15:47:00 2000 From: tromey@cygnus.com (Tom Tromey) Date: Tue, 21 Nov 2000 15:47:00 -0000 Subject: future plans References: <"Mon,> <20> <2000> <15:23:58> <-0800> <(PST)> <4.2.0.58.20001121122249.01a67920@pop.cygnus.com> <87y9yddpof.fsf@creche.cygnus.com> <87y9yc6hkc.fsf@cenderis.demon.co.uk> Message-ID: <87g0kkdgnb.fsf@creche.cygnus.com> >>>>> "Bruce" == Bruce Stephens writes: Bruce> I don't think this requires anything that Tk doesn't already Bruce> have, does it? Can't an application just use "wm protocol Bruce> WM_SAVE_YOURSELF" and do something to save its state? Gnome can already interface with applications that use this method, via a proxy. However, XSMP has more sophisticated facilities than those available via the old ICCCM session management protocol. Tom From mdejong@cygnus.com Tue Nov 21 19:50:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 21 Nov 2000 19:50:00 -0000 Subject: future plans References: <4.3.1.2.20001121204418.00b9f480@mail.qeradiant.com> Message-ID: On Tue, 21 Nov 2000, Timothee Besset wrote: > At 11:02 AM 11/21/00 -0800, you wrote: > >On Mon, 20 Nov 2000, Timothee Besset wrote: > > > > > When can we expect an update to SN? > > > >What kind of "update" were you looking for? We are working on > >getting a snapshot of 5.0 out sometime soon, but that will > >not even be alpha quality. It would be a "hackers only" > >pre-release. > > > > > The most important feature (imo) would be auto-completion. > > > Auto-completion of symbols, listing the members when accessing a struct > > > with -> or . > > > >That would be neat. Are you willing to work on implementing > >this feature? > > Wish I had time to do that. I'm on the user side of sourcenav. I could test > stuff and report bugs but I doubt I would code actual new things. My main > concern is too see source navigator evolve and possibly in a more open way. > But I've been told it will happen with 5.0 Testing and bug reporting is really hard, we welcome any help you could provide. > TTimo > > PS: Source navigator used to be commercial software, at that time I was not > using linux. Now I am, I could get my boss to pay for a license. Instead > it's GPL and people tell me to implement stuff myself. heh .. funny There is another option. If you are interested in a specific feature, you can negotiate a custom engineering contract with Red Hat. We are currently involved in a number of custom contracts with Source-Navigator deliverables. Mo DeJong Red Hat Inc From rfweber@link.com Wed Nov 22 10:20:00 2000 From: rfweber@link.com (Richard F Weber) Date: Wed, 22 Nov 2000 10:20:00 -0000 Subject: Ada Parser Message-ID: <3A1C0E5B.ADF64C94@link.com> Do you know of any Ada parsers available for Source-Navigator? Thanks. --Rich From msimons@onevision.de Wed Nov 22 10:59:00 2000 From: msimons@onevision.de (Michael Simons) Date: Wed, 22 Nov 2000 10:59:00 -0000 Subject: Objective-C Parser Message-ID: <5.0.0.25.0.20001122195655.00b300b0@dali> Hello, Does anybody know of an Objective-C Parser for Source-Navigator? thanks, m!c From Martin.Fischer@de.bosch.com Thu Nov 23 05:10:00 2000 From: Martin.Fischer@de.bosch.com (Fischer Martin (AT/EES3)) Date: Thu, 23 Nov 2000 05:10:00 -0000 Subject: S-N 4.5.2 hangs while scanning header file Message-ID: <01CDE14BF507D2119A0C00805FBBB713016322B6@ermail3.er.bosch.de> Hi all, the file named c:/BC5/INCLUDE/OWL/timegadg.h is part of my installation of Borland C++ 5.02 / OWL 5.2 on NT 4.0, SP6. To get rid of all hanging processes, I have to shut down NT. All other header files of the OWL interface behave well. If somebody want's to look closer at the problem, let me know. fi From bje@redhat.com Thu Nov 23 05:14:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 23 Nov 2000 05:14:00 -0000 Subject: S-N 4.5.2 hangs while scanning header file References: <01CDE14BF507D2119A0C00805FBBB713016322B6@ermail3.er.bosch.de> Message-ID: the file named c:/BC5/INCLUDE/OWL/timegadg.h is part of my installation of Borland C++ 5.02 / OWL 5.2 on NT 4.0, SP6. If you create a project containing this single file, what happens? Ben From Martin.Fischer@de.bosch.com Thu Nov 23 08:56:00 2000 From: Martin.Fischer@de.bosch.com (Fischer Martin (AT/EES3)) Date: Thu, 23 Nov 2000 08:56:00 -0000 Subject: AW: S-N 4.5.2 hangs while scanning header file Message-ID: <01CDE14BF507D2119A0C00805FBBB713016322B7@ermail3.er.bosch.de> > -----Urspr????ngliche Nachricht----- > Von: Ben Elliston [ mailto:bje@redhat.com ] > Gesendet am: Donnerstag, 23. November 2000 14:14 > An: Fischer Martin (AT/EES3) > Cc: sourcenav@sources.redhat.com > Betreff: Re: S-N 4.5.2 hangs while scanning header file > > the file named c:/BC5/INCLUDE/OWL/timegadg.h is part of my > installation of Borland C++ 5.02 / OWL 5.2 on NT 4.0, SP6. > > If you create a project containing this single file, what happens? > the same > Ben > From bje@redhat.com Thu Nov 23 12:07:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 23 Nov 2000 12:07:00 -0000 Subject: AW: S-N 4.5.2 hangs while scanning header file References: <01CDE14BF507D2119A0C00805FBBB713016322B7@ermail3.er.bosch.de> Message-ID: the same That's a good thing. Now, the problem is, how are you going to contribute the file? It is no doubt under Borland copyright. Ben From nso@manbw.dk Fri Nov 24 01:03:00 2000 From: nso@manbw.dk (Niels Skou Olsen) Date: Fri, 24 Nov 2000 01:03:00 -0000 Subject: AW: S-N 4.5.2 hangs while scanning header file References: Message-ID: bje@redhat.com writes: > That's a good thing. Now, the problem is, how are you going to contribute > the file? It is no doubt under Borland copyright. One way to narrow it down is to do a "binary search": Cut the file in half an check each half. Then continue this process recursively on the failing halves until you have a small enough chunk of header file that fails. If everything goes well, you'll end up with a small enough chunk that you can "rewrite" it. This will not violate any copyrights.. Best regards -- Niels Skou Olsen, M.Sc. Intelligent Engine, R&D MAN B&W Diesel A/S From irox@redhat.com Fri Nov 24 12:56:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 24 Nov 2000 12:56:00 -0000 Subject: Ada Parser References: <3A1C0E5B.ADF64C94@link.com> Message-ID: <3A1ED609.96198D0C@redhat.com> Richard F Weber wrote: > > Do you know of any Ada parsers available for Source-Navigator? Thanks. Cas has an Ada95 parser, I'm gussing it's not open source or free, but they do have a trial download (14days) and there parsers should just plug straight into most resent versions of Source-Navigator: http://www.cas-gmbh.de/en/software/snav/main.html I've been looking at rewriting the Tcl and C/C++ parsers using PCCTS, I noticed that there was an Ada parser avaible for PCCTS which could probably be adapted to out put Source Navigator database info. Ian. From irox@redhat.com Fri Nov 24 13:06:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 24 Nov 2000 13:06:00 -0000 Subject: Objective-C Parser References: <5.0.0.25.0.20001122195655.00b300b0@dali> Message-ID: <3A1ED862.7376428B@redhat.com> Michael Simons wrote: > > Hello, > > Does anybody know of an Objective-C Parser for Source-Navigator? Not that I know of one. If you're feeling really ambitious you could have a shot at modifying the existing C/C++ parser to handle Objective-C, but the parser is kind of a mess. It might be easier to start from scratch. Ian. From paul@toolscenter.org Fri Nov 24 14:22:00 2000 From: paul@toolscenter.org (Paul Selormey) Date: Fri, 24 Nov 2000 14:22:00 -0000 Subject: Ada Parser References: <3A1C0E5B.ADF64C94@link.com> <3A1ED609.96198D0C@redhat.com> Message-ID: <001901c05665$2f552a10$2ca65a8f@paulselo> Hello Ian, > I've been looking at rewriting the Tcl and C/C++ parsers > using PCCTS, I noticed that there was an Ada parser > avaible for PCCTS which could probably be adapted to > out put Source Navigator database info. That will be a cool project. I have being playing with the Antlr too. But why PCCTS, since it is no longer being actively supported and developed like the new Antlr? A good and complete Java grammar by the authors of the Antlr is available already, as well as GNU C grammar and strangely :-) no one is done a C/C++ parser!!! There is a C/C++ parser for the PCCTS, but it not completely compatible with the Antlr. http://www.empathy.com/pccts/ Best regards, Paul. From bje@redhat.com Fri Nov 24 15:59:00 2000 From: bje@redhat.com (Ben Elliston) Date: Fri, 24 Nov 2000 15:59:00 -0000 Subject: AW: S-N 4.5.2 hangs while scanning header file References: <01CDE14BF507D2119A0C00805FBBB713016322B8@ermail3.er.bosch.de> Message-ID: Martin Fischer wrote: > the file named c:/BC5/INCLUDE/OWL/timegadg.h is part of my > installation of Borland C++ 5.02 / OWL 5.2 on NT 4.0, SP6. I have been able to reproduce the bug Martin reported with the following minimal test case: class foo { #ifdef WITH_INT foo (int, #else foo (char, #endif int); }; I'm going to have a crack at fixing it this morning. Ben From bje@redhat.com Fri Nov 24 19:57:00 2000 From: bje@redhat.com (Ben Elliston) Date: Fri, 24 Nov 2000 19:57:00 -0000 Subject: Helpful URL Message-ID: You might find this URL helpful for parser hacking. :-) http://www.freedict.com/onldict/hun.html Mo/Syd -- care to make a link from the S-N web pages? Ben From irox@redhat.com Sat Nov 25 02:28:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Sat, 25 Nov 2000 02:28:00 -0000 Subject: Helpful URL References: Message-ID: <3A1F942B.60713B0F@redhat.com> Ben Elliston wrote: > > You might find this URL helpful for parser hacking. :-) > > http://www.freedict.com/onldict/hun.html > > Mo/Syd -- care to make a link from the S-N web pages? err, HAHAHAHHAHAHAHAHHAHAHAHAHAHHA! You're not going to understand this unless you've read most of the source code. Ian. From irox@redhat.com Sat Nov 25 03:14:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Sat, 25 Nov 2000 03:14:00 -0000 Subject: Helpful URL References: Message-ID: <3A1F9F0A.C7EBFC01@redhat.com> Ben Elliston wrote: > > You might find this URL helpful for parser hacking. :-) > > http://www.freedict.com/onldict/hun.html > > Mo/Syd -- care to make a link from the S-N web pages? > > Ben /* Az elso nekifutasra False-szal terunk vissza, ami lehetoseget ad arra, ** hogy az ellentetes keyw_cpp ertekkel ujra megprobaljuk az egesz deklaraciot ** feldolgozni . ** A masodik nekifutasra is False-szal terunk vissza, ami lehetoseget ad arra, ** hogy visszaallitsuk a keyw_cpp-t az eredeti ertekre. ** A harmadik nekifutasra True-val terunk vissza. */ /* The rich drive against False red tape time reappoint, state to brush a side the out there, as fact primary enemy keyw_cpp long to fat problems escape goul radiotherapy rise work. First crank around false red tape time reappoint OK, I'm lost now.... Ian. From ceo@shore.net Sat Nov 25 17:55:00 2000 From: ceo@shore.net (ceo@shore.net) Date: Sat, 25 Nov 2000 17:55:00 -0000 Subject: Building on FreeBSD 3.2: Success!! Message-ID: <3A2065DA.A0DC3EA2@shore.net> After a lot of excellent help from Ben Elliston and other folks here, and some poking around in configure.in files, I finally figured it out. The problem was that the uname command on the machine I was building on produces highly nonstandard output, as it's my development machine at work, and my employer's products are based on a somewhat modified version of FreeBSD. Configure uses uname to figure out where various system calls live, including the ones dealing with dynamic libraries (dlopen() etc.). This meant it was unable to find those libraries, and hence the build failed. I hard-coded the $system variable in SN452-source/tcl8.1/unix/configure.in (line 728) to "FreeBSD-3.2", reran autoconf, deletd config.cache, and it configured and build perfectly. Haven't had a chance to play around with it much, as I'm logged in from home, and running large X apps over my 56K on a *very* good day wet string is not one of life's more rewarding experiences. Many thanks to all who helped. -- -Chip Olson. | ceo@shore.net From Reiner.Suikat@dlr.de Mon Nov 27 08:37:00 2000 From: Reiner.Suikat@dlr.de (Reiner Suikat) Date: Mon, 27 Nov 2000 08:37:00 -0000 Subject: Linking libraries with relative pathnames? Message-ID: Hello, I'm running into a problem trying to link with some shared libraries without having absolute paths defined. The idea is that if I had some means of sharing build target definitions among users, users could then link with their own local copies of certain libraries. Unfortunately, at this time that is not possible (I "share" build target information by copying the entire *.proj files, which works fine) since libraries can only be specified with full absolute pathname. My first idea fo just putting the libs I'm using in an environment variable and placing that in the Link Flags fields works fine on Linux, but not on SGI. The problem here is that the link line now first lists the libs and then the main .o file, in which case the SGI linker does not reolve any symbols from the main program. (It only ssearches each library once for the symbols that are unknown at that time). The gnu linker has no proble here. So, here is the question in short: Is there any way to get a link line similar to ld Flags -o something xx.o ${LIBS} so that the libs (with relative path) come after the main object file? Thanks for any help and keep up the good work, I love sourcenav!!! Greetings Reiner -- Reiner Suikat Institute of Flight Guidance German Aerospace Center (DLR) From irios@proin.es Tue Nov 28 09:52:00 2000 From: irios@proin.es (Nacho de los Ríos Tormo) Date: Tue, 28 Nov 2000 09:52:00 -0000 Subject: include file problems Message-ID: <00112818511300.00635@irios> I am also having this problem on a SuSE 7.0 installation. For example: I'm trying to call the library function: gettimeofday(struct timeval* tv, struct timezone * tz). According to man pages, I need #include (which should expand to /usr/include/sys/time.h) #include (which should expand to /usr/include/unistd.h) SNav is unable to find the correct files -- which are indeed there. I have toyed with the project settings include tab, as well as with the build settings. Depending on the combination, it either finds /usr/include/time.h (wrong file) or /usr/include/sys/unistd.h (also wrong). SNav only manages to find the correct files if I include them explicitly in the project (which calls for a trip to the project editor and to have the project reparsed). But these files themselves do include a variety of other files, which in their turn call others. Trying to manually include all the tree soon leads into confusion and dispair. And this is just for gettimeofday()! The thought of throwing in some threads and an ncurses interface sends shivers down my spine. Any ideas on what is wrong? -- Nacho de los R????os Tormo Procedimientos Integrados S.L. Sevillanos, 67 28609 Sevilla la Nueva, Madrid SPAIN From mdejong@cygnus.com Tue Nov 28 10:14:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Tue, 28 Nov 2000 10:14:00 -0000 Subject: include file problems References: <00112818511300.00635@irios> Message-ID: On Tue, 28 Nov 2000, Nacho de los =?iso-8859-1?q?R=EDos=20Tormo?= wrote: > I am also having this problem on a SuSE 7.0 installation. For example: > > I'm trying to call the library function: > > gettimeofday(struct timeval* tv, struct timezone * tz). > > According to man pages, I need > > #include (which should expand to /usr/include/sys/time.h) > #include (which should expand to /usr/include/unistd.h) > > SNav is unable to find the correct files -- which are indeed there. > > I have toyed with the project settings include tab, as well as with the build > settings. Depending on the combination, it either finds /usr/include/time.h > (wrong file) or /usr/include/sys/unistd.h (also wrong). > > SNav only manages to find the correct files if I include them explicitly in > the project (which calls for a trip to the project editor and to have the > project reparsed). But these files themselves do include a variety of other > files, which in their turn call others. > > Trying to manually include all the tree soon leads into confusion and > dispair. And this is just for gettimeofday()! The thought of throwing in some > threads and an ncurses interface sends shivers down my spine. > > Any ideas on what is wrong? Sounds like a bug. Why don't you take a shot at fixing it? Take at look at the Preferences&::AddInclude methods at line 879 of gui/preferences.tcl. That should get you started. Mo DeJong Red Hat Inc From bje@redhat.com Tue Nov 28 13:42:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 28 Nov 2000 13:42:00 -0000 Subject: include file problems References: <00112818511300.00635@irios> Message-ID: SNav only manages to find the correct files if I include them explicitly in the project (which calls for a trip to the project editor and to have the project reparsed). But these files themselves do include a variety of other files, which in their turn call others. I could be wrong, but my recollection is that S-N will only include files in your project that are explicitly listed at project creation time. That is, #include directives are not followed as you might expect. This is in fact a good idea, since it gives you more control over how big your database gets. Often, users don't care about header files or, say, the function prototype for strcmp(3). If you use the project editor to include all of /usr/include, does it work as expected? Ben From rhussong@ciena.com Tue Nov 28 17:09:00 2000 From: rhussong@ciena.com (Richard Hussong) Date: Tue, 28 Nov 2000 17:09:00 -0000 Subject: Java parser cannot handle foo[].class Message-ID: <2hl82t0gkumj7g2o4sidskgacu6pkeg8me@4ax.com> If a Java source contains the construct "foo[].class", where "foo" is a class name, the parser gets an "abnormal program termination", with text Error: put_symbol argument #3 must not be empty [...] Just thought you might like to know. I'm afraid I don't have time right now to investigate further or try to fix the bug. From rhussong@ciena.com Tue Nov 28 17:13:00 2000 From: rhussong@ciena.com (Richard Hussong) Date: Tue, 28 Nov 2000 17:13:00 -0000 Subject: parse errors: a suggestion Message-ID: This is a request for enhancement: might it be possible to allow the user instruct the project parser to continue silently on errors, rather than throw up two dialog boxes on each error, which must be dismissed by hand? I have some large program trees that trigger known errors in the parsers quite a few times, but I find SourceNav useful anyway, and I wish I could reparse projects without babysitting the parser. From bje@redhat.com Tue Nov 28 17:22:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 28 Nov 2000 17:22:00 -0000 Subject: Java parser cannot handle foo[].class References: <2hl82t0gkumj7g2o4sidskgacu6pkeg8me@4ax.com> Message-ID: If a Java source contains the construct "foo[].class", where "foo" is a class name, the parser gets an "abnormal program termination", with text Error: put_symbol argument #3 must not be empty [...] Can you provide a more complete test case, that can be fed straight into the parser? I'm not very fluent in Java. Ben From bje@redhat.com Tue Nov 28 22:40:00 2000 From: bje@redhat.com (Ben Elliston) Date: Tue, 28 Nov 2000 22:40:00 -0000 Subject: Java parser cannot handle foo[].class References: <87u28r4ko4.fsf@creche.cygnus.com> Message-ID: public class s { public static void main (String[] args) { System.out.println (String[].class); } } I was able to reproduce the problem with the program shown above, but haven't had time to investigate further. Ben From irios@proin.es Wed Nov 29 02:41:00 2000 From: irios@proin.es (Nacho de los Ríos Tormo) Date: Wed, 29 Nov 2000 02:41:00 -0000 Subject: include file problems References: Message-ID: <00112911400200.00884@irios> > > SNav only manages to find the correct files if I include them explicitly > > in the project (which calls for a trip to the project editor and to have > > the project reparsed). But these files themselves do include a variety of > > other files, which in their turn call others. > > > > Trying to manually include all the tree soon leads into confusion and > > dispair. And this is just for gettimeofday()! The thought of throwing in > > some threads and an ncurses interface sends shivers down my spine. > > > > Any ideas on what is wrong? > > Sounds like a bug. Why don't you take a shot at fixing it? > Take at look at the Preferences&::AddInclude methods at > line 879 of gui/preferences.tcl. That should get you started. > > Mo DeJong > Red Hat Inc I will try, but bearing in mind that I have no previous experience with either C++ or TCL/Tk don't expect results any time soon ... :-( Still, doesn't it seem unlikely that it be a problem in the source code? If it were, I'm sure everybody that had tried to build a program of any significance with SNav would have found the same problem. Most programs written nowadays must include files that are in subdirs of usr/include. Could it be something to do with environment variables or directory structure? After all, it seems that the problem occurs with SuSE but not with Red Hat ... Could you point me in some direction? Thanks for your support, Nacho de los R????os Tormo From GPante@ip-globalnet.com Wed Nov 29 03:18:00 2000 From: GPante@ip-globalnet.com (Geert Pante) Date: Wed, 29 Nov 2000 03:18:00 -0000 Subject: Stable Build for Win2000? Message-ID: <6198645AE14CD41189160001022A9B2111B8A4@IPG-LO-MAIL> Hello SourceNavigator people, I am working on an old 500.000 lines C++ project in Visual C++, and experienced serious limitations in VC++ 6.0's builtin source browsing capabilities. I discovered your free product and tried to use it as an alternative (under Win2000, SP1), but the application crashes from time to time, mostly when I try to browse through the core file, which is avg. 50.000 lines of code. In trying to get a more stable system, I tried to rebuild the SourceNavigator from the source code. I didn't find an easy way to do it in VC++, but could ./configure it under cygwin, Cygnus' port of GNU utilities to Win32. The configuration worked there, but the compilation failed at what seems to be some cygwin-specific details I couldn't figure out (some header files in the mingw32 include dir were incompatible, I think). So, what I want to ask to you is, how do I compile SN4.5.2 under Win2000 (preferably under VC++), or do you have a more recent, more stable version for me than the 4.5.2 version on your FTP-site? greetings, Geert Pante. From mdejong@cygnus.com Wed Nov 29 03:46:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 29 Nov 2000 03:46:00 -0000 Subject: Stable Build for Win2000? References: <6198645AE14CD41189160001022A9B2111B8A4@IPG-LO-MAIL> Message-ID: On Wed, 29 Nov 2000, Geert Pante wrote: > Hello SourceNavigator people, > > I am working on an old 500.000 lines C++ project in Visual C++, and > experienced serious limitations in VC++ 6.0's builtin source browsing > capabilities. I discovered your free product and tried to use it as an > alternative (under Win2000, SP1), but the application crashes from time to > time, mostly when I try to browse through the core file, which is avg. > 50.000 lines of code. > > In trying to get a more stable system, I tried to rebuild the > SourceNavigator from the source code. I didn't find an easy way to do it in > VC++, but could ./configure it under cygwin, Cygnus' port of GNU utilities > to Win32. The configuration worked there, but the compilation failed at > what seems to be some cygwin-specific details I couldn't figure out (some > header files in the mingw32 include dir were incompatible, I think). > > So, what I want to ask to you is, how do I compile SN4.5.2 under Win2000 > (preferably under VC++), or do you have a more recent, more stable version > for me than the 4.5.2 version on your FTP-site? Version 4.5.2 is the most recent one available. You could try debugging the crash you are running into. Posting s C stack trace of the spot where it crashed might help. You can find Windows build instructions in the mailing list archive. If the problem is in Tcl code, you don't need to recompile that, you can just edit it in place if you like. If you are uninterested in debugging the problem yourself, you might want to look into purchasing a support contract. cheers Mo DeJong Red Hat Inc From irios@proin.es Wed Nov 29 04:02:00 2000 From: irios@proin.es (Nacho de los Ríos Tormo) Date: Wed, 29 Nov 2000 04:02:00 -0000 Subject: include file problems References: <00112911400200.00884@irios> Message-ID: <00112913010300.01118@irios> Sorry for pestering you, but I would be happy to help have this bug fixed (if indeed a bug). My C++ knowledge is limited and my TCL/Tk knowledge NONEXISTANT, so any help from your part would greatly speed me up. > > > SNav only manages to find the correct files if I include them > > > explicitly in the project (which calls for a trip to the project editor > > > and to have the project reparsed). But these files themselves do > > > include a variety of other files, which in their turn call others. > > > > > > Trying to manually include all the tree soon leads into confusion and > > > dispair. And this is just for gettimeofday()! The thought of throwing > > > in some threads and an ncurses interface sends shivers down my spine. Dumb me! Why not add the whole /usr/include and all its subdirectories in one go with the project editor? .... Well, I've done so, so all involved files MUST now be included in the project. Still I get the same compilation errors: dos_in_lin.c: In function 'get_time_ms' dos_in_lin.c: 30: 'timeval' undeclared (first use in this function) .... make: *** [dos_in_lin.o] Error 1 +++ End +++ Thanks for your support, Nacho de los R????os Tormo Procedimientos Integrados S.L. Sevillanos, 67 28609 Sevilla la Nueva, Madrid SPAIN From bruce.stephens@messagingdirect.com Wed Nov 29 04:10:00 2000 From: bruce.stephens@messagingdirect.com (Bruce Stephens) Date: Wed, 29 Nov 2000 04:10:00 -0000 Subject: include file problems References: <00112911400200.00884@irios> <00112913010300.01118@irios> Message-ID: Nacho de los R????os Tormo writes: [...] > Dumb me! Why not add the whole /usr/include and all its > subdirectories in one go with the project editor? .... Well, I've > done so, so all involved files MUST now be included in the > project. Still I get the same compilation errors: > > dos_in_lin.c: In function 'get_time_ms' > dos_in_lin.c: 30: 'timeval' undeclared (first use in this function) > .... > make: *** [dos_in_lin.o] Error 1 > +++ End +++ That's a compiler error: it's got nothing to do with Source Navigator. SN doesn't tell the compiler what header files to use. Something like this ought to compile, regardless of what include files Source Navigator finds: #include void foo() { struct timeval tv; struct timezone tz; gettimeofday(&tv, &tz); } -- Bruce Stephens Bruce.Stephens@MessagingDirect.com MessagingDirect(UK) Ltd From lstabile@concord.com Wed Nov 29 05:30:00 2000 From: lstabile@concord.com (Larry Stabile) Date: Wed, 29 Nov 2000 05:30:00 -0000 Subject: [presales+tim.w@redhat.com: Re: Source Navigator] Message-ID: <200011291330.IAA27681@sandbox.concord.com> Is there anything you can tell me about the questions below? Thanks. Larry Stabile Concord Communications, Inc. 600 Nickerson Road Marlborough, MA 01752 lstabile@concord.com 508-303-4246 ------- Start of forwarded message ------- Sender: tim@redhat.com Date: Mon, 20 Nov 2000 06:54:56 -0800 Organization: Red Hat X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.2.16-22 i686) X-Accept-Language: en MIME-Version: 1.0 Subject: Re: Source Navigator References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit From: "tim.w" To: presales+lstabile+concord.com@redhat.com X-Mailgate-Version: Mailgate v1.3 Content-Length: 1520 > I tried Source Navigator 4.5.2 and it looks pretty good. > > We have a source base of 1M+ lines of C++ code, so it takes awhile to > create a project database. I would like to be able to operate in the > following way: > > 1. Make an initial source pass. > > 2. Schedule a nightly job to update the project database. > > 3. Allow developers to see this "master" project database > read-only from their desktop systems by accessing the shared > project database and source code. > > We use Clearcase as our source control system. > > The main question I have right now is whether the project database can > be updated incrementally -- i.e., by checking file modification times > or other methods. It would be expensive for us to scan the whole > source tree each night. > > Also, please comment on whether you think the above sharing > methodology is feasible. Note that I don't need a system right now > which can share a main database and incrementally update local > changes on developer's destktop machines. The main use of this will be > to give people a way to browse the general source structure, not as an > up-to-the-minute development tool. Larry, I have forwarded your concerns to the Source Navigator people. They can be contacted at... sourcenav@sources.redhat.com ...if you have any other questions about Source Navigator. Thank you, - -- Tim W. | "I don't care to belong to a club that Red Hat Presales | accepts people like me as members." | --Grocho Marx ------- End of forwarded message ------- From irios@proin.es Wed Nov 29 06:53:00 2000 From: irios@proin.es (Nacho de los Ríos Tormo) Date: Wed, 29 Nov 2000 06:53:00 -0000 Subject: include file problems References: Message-ID: <00112915522100.01532@irios> I am AWFULLY sorry; this was all my fault. ... I tried to compile the file from the command line and got the same errors. I tried gcc -E to do just the preprocessing, and the declaration was there OK. I checked the code, and there it was, glaring: struct timeval system_time; ... gettimeofday(&timeval, 0); ... &timeval????? SHAME on me! I'll try to get smarter ... Thanks again. Nacho de los R????os Tormo Procedimientos Integrados S.L. Sevillanos, 67 28609 Sevilla la Nueva, Madrid SPAIN From bje@redhat.com Wed Nov 29 15:06:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 29 Nov 2000 15:06:00 -0000 Subject: Include File Problem (fwd) Message-ID: ---------- Forwarded message ---------- Date: Wed, 29 Nov 2000 14:08:30 -0600 From: Rebecca.R.Krause@seagate.com To: bje@redhat.com Subject: Include File Problem I apologize for e-mailing you directly rather than going through the mailing list but my e-mail kept being returned by mailer-daemon so I thought I would give your e-mail address a shot. I am running RedHat Linux 7 and Source Navigator 4.5 2. I have a project set up that consists of 3 .cpp files and 7 .h files including proto.h, stat.h and dir.h. All these files are located at /home/myloginname/SNavProj/reports. These header files call other files that can be found in /usr/include or /usr/local/include. I went into the Project Editor window and added the directories for /usr/include and /usr/local/include. When I try to make my project I get an error because there is "no such file or directory" as proto.h, stat.h or dir.h. What have I done wrong? Thanks for the help in advance! Regards, Rebecca From spolk@redhat.com Wed Nov 29 18:45:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 29 Nov 2000 18:45:00 -0000 Subject: [presales+tim.w@redhat.com: Re: Source Navigator] References: <200011291330.IAA27681@sandbox.concord.com> Message-ID: <4.2.0.58.20001129184557.01b1c2e0@pop.cygnus.com> This is an often-requested feature from sites with large amounts of source code. Right now, we don't support it very well. It is on the (long) list of things to address. At 08:30 AM 11/29/00 -0500, Larry Stabile wrote: >Is there anything you can tell me about the questions below? Thanks. > >Larry Stabile >Concord Communications, Inc. >600 Nickerson Road >Marlborough, MA 01752 >lstabile@concord.com >508-303-4246 > >------- Start of forwarded message ------- >Sender: tim@redhat.com >Date: Mon, 20 Nov 2000 06:54:56 -0800 >Organization: Red Hat >X-Mailer: Mozilla 4.75 [en] (X11; U; Linux 2.2.16-22 i686) >X-Accept-Language: en >MIME-Version: 1.0 >Subject: Re: Source Navigator >References: >Content-Type: text/plain; charset=us-ascii >Content-Transfer-Encoding: 7bit >From: "tim.w" >To: presales+lstabile+concord.com@redhat.com >X-Mailgate-Version: Mailgate v1.3 >Content-Length: 1520 > > > > I tried Source Navigator 4.5.2 and it looks pretty good. > > > > We have a source base of 1M+ lines of C++ code, so it takes awhile to > > create a project database. I would like to be able to operate in the > > following way: > > > > 1. Make an initial source pass. > > > > 2. Schedule a nightly job to update the project database. > > > > 3. Allow developers to see this "master" project database > > read-only from their desktop systems by accessing the shared > > project database and source code. > > > > We use Clearcase as our source control system. Well, although there is something labelled "ClearCase" in Source-Navigator, it never worked for us, and we believe that it never worked for the original developers. > > > > The main question I have right now is whether the project database can > > be updated incrementally -- i.e., by checking file modification times > > or other methods. It would be expensive for us to scan the whole > > source tree each night. We already scan it incrementally. There are performance problems with deleting records from the underlying database that makes it horribly slow. > > > > Also, please comment on whether you think the above sharing > > methodology is feasible. Note that I don't need a system right now > > which can share a main database and incrementally update local > > changes on developer's destktop machines. The main use of this will be > > to give people a way to browse the general source structure, not as an > > up-to-the-minute development tool. > > >Larry, > >I have forwarded your concerns to the Source Navigator people. They can >be contacted at... > >sourcenav@sources.redhat.com > >...if you have any other questions about Source Navigator. > >Thank you, > >- -- >Tim W. | "I don't care to belong to a club that >Red Hat Presales | accepts people like me as members." > | --Grocho Marx >------- End of forwarded message ------- Syd Polk spolk@redhat.com Engineering Manager +1 415 777 9810 x 241 Red Hat, Inc. From lstabile@concord.com Wed Nov 29 19:59:00 2000 From: lstabile@concord.com (Larry Stabile) Date: Wed, 29 Nov 2000 19:59:00 -0000 Subject: [presales+tim.w@redhat.com: Re: Source Navigator] References: <4.2.0.58.20001129184557.01b1c2e0@pop.cygnus.com> Message-ID: <200011300358.WAA05234@sandbox.concord.com> Thanks for the info. It works ok in clearcase simply by running the navigator in a view. I don't think we need any other special features in that regard. I did try a rescan and even with the delete it was much quicker than a full project build. I will ask our admins to set this up on a build machine. If you like, I will report to you how it turned out. - Larry ------------------ Larry Stabile Concord Communications, Inc. 600 Nickerson Road Marlborough, MA 01752 lstabile@concord.com 508-303-4246 From girke@itu304.ut.TU-Berlin.DE Thu Nov 30 07:19:00 2000 From: girke@itu304.ut.TU-Berlin.DE (girke@itu304.ut.TU-Berlin.DE) Date: Thu, 30 Nov 2000 07:19:00 -0000 Subject: Problem under WinNT4 Server SP5 Message-ID: Hi, I have a problem with the xref part of SN. After parsing my project I got the following messages: 1. Dr.Watson: dbimp.exe Access violation (0xc0000005), adress 0x0043b267 2. Error: child process exited abnormally -> SN had a problem generating Cross-Reference information. CR information is not complete. 3. dbopen error: D:/Develop/db_transfer/.snprj/db_trans.by:Invalid argument My project is a MS VJ++ 6.0 with 28 source files and about 10.000 lines of code. Do you have any idea? Thanks in advance Mario Girke Institut f????r Technischen Umweltschutz Technische Universit????t Berlin ------------------------------------------- girke@itu304.ut.tu-berlin.de, girke@gmx.net From bje@redhat.com Thu Nov 30 15:57:00 2000 From: bje@redhat.com (Ben Elliston) Date: Thu, 30 Nov 2000 15:57:00 -0000 Subject: Problem under WinNT4 Server SP5 References: Message-ID: Do you have any idea? Are you able to hone in on the problem code by partitioning your project down until you've only got a few source files? When you do, can you post them somewhere? Ben From lstabile@concord.com Thu Nov 30 19:07:00 2000 From: lstabile@concord.com (Larry Stabile) Date: Thu, 30 Nov 2000 19:07:00 -0000 Subject: Question about batch mode Message-ID: <200012010306.WAA05980@sandbox.concord.com> Can batch mode be used to update a project as well as create a new one? Thanks. Larry Stabile Concord Communications, Inc. 600 Nickerson Road Marlborough, MA 01752 lstabile@concord.com 508-303-4246 From scherrey@innoverse.com Fri Dec 1 14:42:00 2000 From: scherrey@innoverse.com (Benjamin Scherrey) Date: Fri, 01 Dec 2000 14:42:00 -0000 Subject: SN on Sparc/Solaris editor weirdness. Message-ID: <00120117282907.20224@vai.innoverse.com> When I run SourceNavigator under SPARC/Solaris the editor inserts a '*' character everytime I hit the right shift key. The key is shifted properly so I get stuff like this: cout *<< *"Hello!*" *<< endl; when I typed this: cout << "Hello!" << endl; Any ideas? Ben Scherrey From irox@redhat.com Fri Dec 1 14:48:00 2000 From: irox@redhat.com (Ian Roxborough) Date: Fri, 01 Dec 2000 14:48:00 -0000 Subject: SN on Sparc/Solaris editor weirdness. References: <00120117282907.20224@vai.innoverse.com> Message-ID: <3A282ABF.D3635541@redhat.com> No, but it is the weirdest bug I've heard of in a while. When you say everytime, does this happen in file dialogs, filters, entry boxes the preferences dialog, etc.? IAn. Benjamin Scherrey wrote: > > When I run SourceNavigator under SPARC/Solaris the editor inserts a '*' > character everytime I hit the right shift key. The key is shifted properly so > I get stuff like this: > > cout *<< *"Hello!*" *<< endl; > > when I typed this: > > cout << "Hello!" << endl; > > Any ideas? > > Ben Scherrey From scherrey@innoverse.com Fri Dec 1 15:12:00 2000 From: scherrey@innoverse.com (Benjamin Scherrey) Date: Fri, 01 Dec 2000 15:12:00 -0000 Subject: SN on Sparc/Solaris editor weirdness. References: <00120117282907.20224@vai.innoverse.com> <3A282ABF.D3635541@redhat.com> Message-ID: <00120117572108.20224@vai.innoverse.com> Happens in all of SN. I'm accessing the Sun box from an Intel/Linux box via ssh. It's only SN that has this problem, however. No other program/shell does this. later, Ben Scherrey PS: Its SN452 built from source using gcc-2.95.2, FWIW. On Friday 01 December 2000 17:48, Ian Roxborough wrote: > No, but it is the weirdest bug I've heard of in a while. > > When you say everytime, does this happen in file dialogs, > filters, entry boxes the preferences dialog, etc.? From mdejong@cygnus.com Fri Dec 1 15:32:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Fri, 01 Dec 2000 15:32:00 -0000 Subject: SN on Sparc/Solaris editor weirdness. References: <00120117572108.20224@vai.innoverse.com> Message-ID: On Fri, 1 Dec 2000, Benjamin Scherrey wrote: > Happens in all of SN. I'm accessing the Sun box from an Intel/Linux box via > ssh. It's only SN that has this problem, however. No other program/shell does > this. > > later, > > Ben Scherrey > > PS: Its SN452 built from source using gcc-2.95.2, FWIW. Could you try running the following over your ssh X session? First, save these commands into a file called type.tcl. % cat type.tcl entry .e pack .e % wish type.tcl (I am assuming that you have wish installed somewhere on your server) That should bring up a little entry box that you can type into. Now run the version of wish that got installed in your SN install dir. $INSTALL/bin/wish type.tcl The same entry box should appear and you should be able to type into it with no problems. Try that and tell us what results you get. Mo DeJong Red Hat Inc From l.lopez@cstelecom.com Mon Dec 4 07:46:00 2000 From: l.lopez@cstelecom.com (Laurent Lopez) Date: Mon, 04 Dec 2000 07:46:00 -0000 Subject: snavigator 452 freeze while referenceing Linux kernel code Message-ID: <14891.47853.857633.853088@verif03.fty.cstelecom.com> In fact a single file make snavigator 4.5.2 freeze, this file is added to this mail. Just put this file in a directory and try to create a new projetc with this directory. Snavigator seems to enter in an endless loop while the "Scanning Project ..." window is open. -------------- next part -------------- A non-text attachment was scrubbed... Name: arch.c Type: text/x-c Size: 7593 bytes Desc: not available URL: From bje@redhat.com Mon Dec 4 16:22:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 04 Dec 2000 16:22:00 -0000 Subject: snavigator 452 freeze while referenceing Linux kernel code References: <14891.47853.857633.853088@verif03.fty.cstelecom.com> Message-ID: In fact a single file make snavigator 4.5.2 freeze, this file is added to this mail. Just put this file in a directory and try to create a new projetc with this directory. Snavigator seems to enter in an endless loop while the "Scanning Project ..." window is open. The following technique may help -- I use it often. $ echo arch.c > file.list $ gdb /path/to/cbrowser -y file.list -p /bin/cat (-y says where to find the file manifest, -p says which program will process the parser's output--which is usually dbimp). I would normally try to look into this myself, but I'm very busy right now. Good luck! From JLKasberger@lbl.gov Mon Dec 4 18:56:00 2000 From: JLKasberger@lbl.gov (Jay Kasberger) Date: Mon, 04 Dec 2000 18:56:00 -0000 Subject: You people rule. Message-ID: <3A2C5A11.387AB842@lbl.gov> Hi, I just wanted to let you know - as I'm sure you already do - that Source-Navigator is truly an excellent product. We use it here at Lawrence Berkeley National Laboratory for genome project code, and we couldn't be happier. It's very well thought out, powerful, and extremely easy to use. If you're interested, here's a feature suggestion: how about a way to export all your different build configurations into a single Makefile, with a build target for each configuration? Just an idea. Anyway, keep up the good work, and best of luck in future endeavors. From spolk@redhat.com Mon Dec 4 19:49:00 2000 From: spolk@redhat.com (Syd Polk) Date: Mon, 04 Dec 2000 19:49:00 -0000 Subject: You people rule. References: <3A2C5A11.387AB842@lbl.gov> Message-ID: <3A2C666F.F589BB44@redhat.com> Jay Kasberger wrote: > > Hi, > > I just wanted to let you know - as I'm sure you already do - that > Source-Navigator is truly an excellent product. We use it here at > Lawrence Berkeley National Laboratory for genome project code, and we > couldn't be happier. It's very well thought out, powerful, and > extremely easy to use. > > If you're interested, here's a feature suggestion: how about a way to > export all your different build configurations into a single Makefile, > with a build target for each configuration? Just an idea. Anyway, keep > up the good work, and best of luck in future endeavors. This is actually what autoconf/automake/configure is for. Maintaining different build configs in one Makefile is well nigh impossible. Instead, we maintain configure.in and all of the various Makefile.am's and we get all of the magic foo to make it build all over the place. From kiev78@ucla.edu Mon Dec 4 20:02:00 2000 From: kiev78@ucla.edu (PERELYUBSKIY, DENIS) Date: Mon, 04 Dec 2000 20:02:00 -0000 Subject: performance? Message-ID: hello, i am generally very happy with source navigator, but it is very slow when x-ref'ing fairly large code trees ( several hundred thousands lines ) i was just wondering if there were any plans to improve performance, or is it just the nature of the beast: it would take too long, and not be worth it at the end? also, i'd like to comment on one of the earlier threads which i saw in the archive: "SourceNav with Windows 2K problems". i too, see this problem, where the mouse cursor must be inside the window for the status bar to be updated, and processing to go on. please dont get me wrong, i love the program though. once its done parsing, it is very, very usefull thank you very much. denis From bje@redhat.com Mon Dec 4 20:04:00 2000 From: bje@redhat.com (Ben Elliston) Date: Mon, 04 Dec 2000 20:04:00 -0000 Subject: You people rule. References: <3A2C666F.F589BB44@redhat.com> Message-ID: This is actually what autoconf/automake/configure is for. Maintaining different build configs in one Makefile is well nigh impossible. More to the point, this implies that we know all of our build targets up front. We don't. Someone built S-N on FreeBSD last week -- until then, we had never done it. Ben From echristo@cygnus.com Tue Dec 5 00:15:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Tue, 05 Dec 2000 00:15:00 -0000 Subject: performance? References: Message-ID: <3A2C9901.388C3DBE@cygnus.com> > i am generally very happy with source navigator, > but it is very slow when x-ref'ing fairly > large code trees ( several hundred thousands lines ) > Umm.. several hundred thousand lines is quite a bit to x-ref if you think about it. One way of looking at it is to see how long it takes to compile that much code with optimization turned off - it is a similar exercise. You also may want to check that you don't have cross references to local variables turned on, this takes a _lot_ more time. -eric From spolk@redhat.com Tue Dec 5 08:15:00 2000 From: spolk@redhat.com (Syd Polk) Date: Tue, 05 Dec 2000 08:15:00 -0000 Subject: performance? References: <3A2C9901.388C3DBE@cygnus.com> Message-ID: <3A2D1560.ACDFD7B8@redhat.com> In general, X-Ref is O[n^2] where n is the number of variables in your source code. So large amounts of code produces a non-linear increase in the time it takes to X-Ref. Eric Christopher wrote: > > > i am generally very happy with source navigator, > > but it is very slow when x-ref'ing fairly > > large code trees ( several hundred thousands lines ) > > > > Umm.. several hundred thousand lines is quite a bit to x-ref if you > think about it. One way of looking at it is to see how long it takes to > compile that much code with optimization turned off - it is a similar > exercise. > > You also may want to check that you don't have cross references to local > variables turned on, this takes a _lot_ more time. > > -eric From mdejong@cygnus.com Wed Dec 6 15:00:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 06 Dec 2000 15:00:00 -0000 Subject: Running SN from Cygwin's bash shell. Message-ID: I have noticed some strange problems running Source-Navigator from Cygwin's bash shell on NT. The CPU usage of SN goes to 100% and it crashes on exit. SN works just fine if your launch it from the explorer of from a cmd.exe command line. At this point, I think we need to just suggest that folks not run Source-Navigator directly from bash until we can get this sorted out. cheers Mo DeJong Red Hat Inc From spolk@redhat.com Wed Dec 6 20:21:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 06 Dec 2000 20:21:00 -0000 Subject: Running SN from Cygwin's bash shell. References: Message-ID: <3A2F1115.125C3810@redhat.com> Hmm. I agree. Could you also ping the gnu-win32-local group? Mo DeJong wrote: > > I have noticed some strange problems running Source-Navigator > from Cygwin's bash shell on NT. The CPU usage of SN goes to > 100% and it crashes on exit. SN works just fine if your > launch it from the explorer of from a cmd.exe command > line. At this point, I think we need to just suggest that > folks not run Source-Navigator directly from bash until > we can get this sorted out. > > cheers > Mo DeJong > Red Hat Inc From mailing@devspace.com Thu Dec 7 10:32:00 2000 From: mailing@devspace.com (Christian Gross) Date: Thu, 07 Dec 2000 10:32:00 -0000 Subject: Deciphering VC++ output Message-ID: Hi I have been trying to decipher VC++ output so that the build environment can jump for me to the file. Here is an example output buffer C:\projects\tredix\db\dbCDObjects.cpp(42) : error C2065: 'storedProcedure' : undeclared identifier In the sn_cmp_g.pat file I added the following pattern # c:\dir\filename.c(789): some error ([^c:\\ ]+)\(([0-9]+)\): And when I run SN and build something I can click on the build output, but it does not jump to the file. Could some please help me? Thanks Christian From kwythers@forestry.umn.edu Fri Dec 8 09:22:00 2000 From: kwythers@forestry.umn.edu (Kirk R. Wythers) Date: Fri, 08 Dec 2000 09:22:00 -0000 Subject: help with codefusion install Message-ID: <3A311AD9.C196D16D@forestry.umn.edu> I need to set the LD_PRELOAD and LD_LIBRARY_PATH environment variables on a RH7.0 machine. Installed into the directory /opt/cygnus/... I want to make sure that I do this right... So could some kind sole clue me in as to the correct procedure for setting these environment variables? Thanks, Kirk -- Kirk R. Wythers University of Minnesota Email: kwythers@forestry.umn.edu Department of Forest Resources Tel: 612.625.2261 1530 Cleveland Ave. N. Fax: 612 625.5212 Saint Paul, MN 55108 From echristo@cygnus.com Fri Dec 8 10:24:00 2000 From: echristo@cygnus.com (Eric Christopher) Date: Fri, 08 Dec 2000 10:24:00 -0000 Subject: help with codefusion install References: <3A311AD9.C196D16D@forestry.umn.edu> Message-ID: <3A311C36.C77921D3@cygnus.com> "Kirk R. Wythers" wrote: > > I need to set the LD_PRELOAD and LD_LIBRARY_PATH environment variables > on a RH7.0 machine. Installed into the directory /opt/cygnus/... You'll only need to set LD_PRELOAD if you are mixing versions of glibc, in particular, 2.0 and 2.1. Keep in mind that we never tested this on RH 7.0. > > I want to make sure that I do this right... So could some kind sole clue > me in as to the correct procedure for setting these environment > variables? Depends on the shell. Using a csh (or similar): setenv LD_PRELOAD using an sh (or similar): export LD_PRELOAD= There is also a support mailing list for this: codefusion-support@cygnus.com -eric From DoveJ@intgame.com Wed Dec 13 02:00:00 2000 From: DoveJ@intgame.com (Dove.John) Date: Wed, 13 Dec 2000 02:00:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK background ? Message-ID: <867F40AC7ACBD4119AC100105A13C50306C847@manbwbapp1.intgame.com> I would like to use SN4.5.2, but I can't see the cursor with the colour scheme I'm using. My preferred colour scheme is BLACK background with mostly green text. I can't find any mechanism for changing the colour of the cursor which is always BLACK, when it needs to be green or white etc. Any solutions would be greatly appreciated. -- JD From mdejong@cygnus.com Wed Dec 13 02:11:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 13 Dec 2000 02:11:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK b ackground ? References: <867F40AC7ACBD4119AC100105A13C50306C847@manbwbapp1.intgame.com> Message-ID: On Wed, 13 Dec 2000, Dove.John wrote: > I would like to use SN4.5.2, but I can't see the cursor with the colour > scheme I'm using. My preferred colour scheme is BLACK background with mostly > green text. > > I can't find any mechanism for changing the colour of the cursor which is > always BLACK, when it needs to be green or white etc. > > Any solutions would be greatly appreciated. > -- > JD Does this Tk code do what you want? Try running it in the wish shell installed in the sn bin dir. text .t pack .t .t configure -bg black -fg green .t configure -insertbackground green If that is what you wanted, then you will need to either hack that "$text configure -insertbackground green" bit into the source code or set the proper X resource for it. There is no preferences option for the text insert cursor color, we really should add one. Anyone feel like taking a stab at adding a preferences option for the text insertion cursor? Mo DeJong Red Hat Inc From DoveJ@intgame.com Wed Dec 13 02:28:00 2000 From: DoveJ@intgame.com (Dove.John) Date: Wed, 13 Dec 2000 02:28:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK b ackground ? Message-ID: <867F40AC7ACBD4119AC100105A13C50306C848@manbwbapp1.intgame.com> My apologise for forgetting to mention that I'm using SN4.5.2 on an NT workstation. -----Original Message----- From: Mo DeJong [ mailto:mdejong@cygnus.com ] Sent: 13 December 2000 10:11 To: sourcenav@sources.redhat.com Subject: Re: How do you change the colour of the cursor to WHITE for a BLACK b ackground ? On Wed, 13 Dec 2000, Dove.John wrote: > I would like to use SN4.5.2, but I can't see the cursor with the colour > scheme I'm using. My preferred colour scheme is BLACK background with mostly > green text. > > I can't find any mechanism for changing the colour of the cursor which is > always BLACK, when it needs to be green or white etc. > > Any solutions would be greatly appreciated. > -- > JD Does this Tk code do what you want? Try running it in the wish shell installed in the sn bin dir. text .t pack .t .t configure -bg black -fg green .t configure -insertbackground green If that is what you wanted, then you will need to either hack that "$text configure -insertbackground green" bit into the source code or set the proper X resource for it. There is no preferences option for the text insert cursor color, we really should add one. Anyone feel like taking a stab at adding a preferences option for the text insertion cursor? Mo DeJong Red Hat Inc From marta.stojanovic@nrc.ca Wed Dec 13 07:09:00 2000 From: marta.stojanovic@nrc.ca (Marta Stojanovic) Date: Wed, 13 Dec 2000 07:09:00 -0000 Subject: SNsdk on WinNT ? Message-ID: <3A3791D3.B4F3D2CD@nrc.ca> Hi all! I saw a (unfortunatelly unanswered) message from Andrew Brackley on SN mailing list (dated July 27 2000). He had the exact same problem as I do now : I'm not able to run a Tcl script using SNsdk on WinNt. The error message is : Error in startup: CreateProcess failed: The system cannot find the file specified. This message appears when trying to run SNsdk from MSDOS prompt, both alone and with my Tcl file as an argument. I assumed that it has something to do with environment, so I set TCL_LIBRARY as I did on Linux, I also put SN452/bin and SN452/share/Tcl8.1 in PATH, but it doesn't work. I tried to run it from the console opened by hyper, but it doesn't work either. I would really like to avoid using Cygwin - as Andrew proposed -(to make it easier to use. From the SN doc it seems that it should work without it. Do you have any idea ? Thanks a lot in advance. Kind regards, Marta. P.S. My script works fine on Linux (I set TCL_LIBRARY and left the first three lines used to start wish : "exec hyper ..."). For NT I left out these lines. From jlmlists@linuxfan.com Wed Dec 13 07:12:00 2000 From: jlmlists@linuxfan.com (Jackie Manning) Date: Wed, 13 Dec 2000 07:12:00 -0000 Subject: Search default settings Message-ID: <00121309115600.01413@main.manningj.com> Is there a way to set the default settings for the search dialog. I would like to, by default, turn off the ignore case setting From spolk@redhat.com Wed Dec 13 09:22:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 13 Dec 2000 09:22:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK background ? References: <867F40AC7ACBD4119AC100105A13C50306C847@manbwbapp1.intgame.com> Message-ID: <3A37B053.F08AF225@redhat.com> "Dove.John" wrote: > > I would like to use SN4.5.2, but I can't see the cursor with the colour > scheme I'm using. My preferred colour scheme is BLACK background with mostly > green text. > > I can't find any mechanism for changing the colour of the cursor which is > always BLACK, when it needs to be green or white etc. > > Any solutions would be greatly appreciated. > -- > JD You can change it using the .Xresource option database. I don't remember what the entry for the cursor is, however. From spolk@redhat.com Wed Dec 13 09:23:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 13 Dec 2000 09:23:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK b ackground ? References: <867F40AC7ACBD4119AC100105A13C50306C848@manbwbapp1.intgame.com> Message-ID: <3A37B091.8F426B0F@redhat.com> But your other applications must have problems with black cursors, too. We have no control of the cursor color on NT. "Dove.John" wrote: > > My apologise for forgetting to mention that I'm using SN4.5.2 on an NT > workstation. > > -----Original Message----- > From: Mo DeJong [ mailto:mdejong@cygnus.com ] > Sent: 13 December 2000 10:11 > To: sourcenav@sources.redhat.com > Subject: Re: How do you change the colour of the cursor to WHITE for a > BLACK b ackground ? > > On Wed, 13 Dec 2000, Dove.John wrote: > > > I would like to use SN4.5.2, but I can't see the cursor with the colour > > scheme I'm using. My preferred colour scheme is BLACK background with > mostly > > green text. > > > > I can't find any mechanism for changing the colour of the cursor which is > > always BLACK, when it needs to be green or white etc. > > > > Any solutions would be greatly appreciated. > > -- > > JD > > Does this Tk code do what you want? Try running it > in the wish shell installed in the sn bin dir. > > text .t > pack .t > .t configure -bg black -fg green > .t configure -insertbackground green > > If that is what you wanted, then you will need > to either hack that "$text configure -insertbackground green" > bit into the source code or set the proper X resource > for it. > > There is no preferences option for the text insert > cursor color, we really should add one. Anyone > feel like taking a stab at adding a preferences > option for the text insertion cursor? > > Mo DeJong > Red Hat Inc From spolk@redhat.com Wed Dec 13 09:46:00 2000 From: spolk@redhat.com (Syd Polk) Date: Wed, 13 Dec 2000 09:46:00 -0000 Subject: How do you change the colour of the cursor to WHITE for a BLACK b ackground ? References: <867F40AC7ACBD4119AC100105A13C50306C848@manbwbapp1.intgame.com> <3A37B091.8F426B0F@redhat.com> Message-ID: <3A37B5EE.16AE05A9@redhat.com> Are you running the NT version of SN, or are you running a UNIX version on another machine and using an NT XServer? Syd Polk wrote: > > But your other applications must have problems with black cursors, too. > We have no control of the cursor color on NT. > > "Dove.John" wrote: > > > > My apologise for forgetting to mention that I'm using SN4.5.2 on an NT > > workstation. > > > > -----Original Message----- > > From: Mo DeJong [ mailto:mdejong@cygnus.com ] > > Sent: 13 December 2000 10:11 > > To: sourcenav@sources.redhat.com > > Subject: Re: How do you change the colour of the cursor to WHITE for a > > BLACK b ackground ? > > > > On Wed, 13 Dec 2000, Dove.John wrote: > > > > > I would like to use SN4.5.2, but I can't see the cursor with the colour > > > scheme I'm using. My preferred colour scheme is BLACK background with > > mostly > > > green text. > > > > > > I can't find any mechanism for changing the colour of the cursor which is > > > always BLACK, when it needs to be green or white etc. > > > > > > Any solutions would be greatly appreciated. > > > -- > > > JD > > > > Does this Tk code do what you want? Try running it > > in the wish shell installed in the sn bin dir. > > > > text .t > > pack .t > > .t configure -bg black -fg green > > .t configure -insertbackground green > > > > If that is what you wanted, then you will need > > to either hack that "$text configure -insertbackground green" > > bit into the source code or set the proper X resource > > for it. > > > > There is no preferences option for the text insert > > cursor color, we really should add one. Anyone > > feel like taking a stab at adding a preferences > > option for the text insertion cursor? > > > > Mo DeJong > > Red Hat Inc From mdejong@cygnus.com Wed Dec 13 12:34:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 13 Dec 2000 12:34:00 -0000 Subject: SNsdk on WinNT ? References: <3A3791D3.B4F3D2CD@nrc.ca> Message-ID: On Wed, 13 Dec 2000, Marta Stojanovic wrote: > Hi all! > > I saw a (unfortunatelly unanswered) message from Andrew Brackley on SN > mailing list (dated July 27 2000). He had the exact same problem as I do > > now : I'm not able to run a Tcl script using SNsdk on WinNt. The error > message is : > > Error in startup: CreateProcess failed: The system cannot find the > file specified. > > This message appears when trying to run SNsdk from MSDOS prompt, both > alone and with my Tcl file as an argument. > > I assumed that it has something to do with environment, so I set > TCL_LIBRARY as I did on Linux, I also put SN452/bin and > SN452/share/Tcl8.1 in PATH, but it doesn't work. > > I tried to run it from the console opened by hyper, but it doesn't work > either. > > I would really like to avoid using Cygwin - as Andrew proposed -(to make > > it easier to use. From the SN doc it seems that it should work without > it. CreateProcess is an NT thing. It is like a fork+exec, it must mean that the system can not find a program it wants to run. This has nothing to do with TCL_LIBRARY or anything like that. You are going to have to debug this one yourself, I would say double check that the fully qualified SN bin dir is on your PATH. You should just be able to run SN by double clicking on the SN program from the Windows Explorer. Good luck. Mo DeJong Red Hat Inc From Olivier.Lefevre@genedata.com Wed Dec 13 13:01:00 2000 From: Olivier.Lefevre@genedata.com (Olivier Lefevre) Date: Wed, 13 Dec 2000 13:01:00 -0000 Subject: SN bugs Message-ID: <037001c06547$b02ac410$daada19d@ch.genedata.com> Sorry if this is dumb but where does one send bug reports? I've looked all over the SN web site but did not find anything. Regards, Olivier Lefevre From mdejong@cygnus.com Wed Dec 13 13:09:00 2000 From: mdejong@cygnus.com (Mo DeJong) Date: Wed, 13 Dec 2000 13:09:00 -0000 Subject: SN bugs References: <037001c06547$b02ac410$daada19d@ch.genedata.com> Message-ID: On Wed, 13 Dec 2000, Olivier Lefevre wrote: > Sorry if this is dumb but where does one send bug reports? I've looked > all over the SN web site but did not find anything. > > Regards, > > Olivier Lefevre Well, you can always send patches for bugs that you have fixed to this list. You can also send bug reports, but we like patches a lot more. Mo DeJong Red Hat Inc From bje@redhat.com Wed Dec 13 14:06:00 2000 From: bje@redhat.com (Ben Elliston) Date: Wed, 13 Dec 2000 14:06:00 -0000 Subject: Search default settings References: <00121309115600.01413@main.manningj.com> Message-ID: Is there a way to set the default settings for the search dialog. I would like to, by default, turn off the ignore case setting You could modify the source to change these settings, if that's what you want to do. Ben From marta.stojanovic@nrc.ca Wed Dec 13 14:14:00 2000 From: marta.stojanovic@nrc.ca (Marta Stojanovic) Date: Wed, 13 Dec 2000 14:14:00 -0000 Subject: SNsdk on WinNT ? Message-ID: <3A37F55A.C9D4A3A9@nrc.ca> >CreateProcess is an NT thing. It is like a fork+exec, it must >mean that the system can not find a program it wants to run. >This has nothing to do with TCL_LIBRARY or anything like >that. You are going to have to debug this one yourself, >I would say double check that the fully qualified SN bin dir >is on your PATH. You should just be able to run SN by double >clicking on the SN program from the Windows Explorer. I just wanted to clarify : Source Navigator works very well on NT, it produces all the database files. But I would like to use your Database API, which I do get running my script with : SNsdk