* Re: 1.1.8: Too large entry in termcap file

[David A. Cobb]

At 6/18/01 06:12 AM (Monday), Alois Steindl wrote:
>On Sat, 16 Jun 2001 23:29:36 -0400,
>Christopher Faylor <cgf at redhat dot com> wrote:
>
>>Looking at the entry that is in termcap for linux, I don't see any way
>>around this.  I did compare it against the entry from Red Hat and I see
>>that they just squeak in under 1024.
>
>
>I get 1042 for linux and 1034 for cygwin
>
>
>>I compared the two and obviously the Cygwin version does have more "stuff"
>>but I don't think that any of it is obviously wrong.  So, the trivial
>>fix for this is to increase the size of your buffer.  I suspect that this
>>is what most applications who use termcap had already done years ago.
>the problem is, that the length 1024 is cited in the man page. Violating this
>
>constrained _is_ a bug and not "my alleged cygwin problem", as you stated 
>in your email. As I wrote in my first message, the problem disappears if I 
>increase this buffer or avoid termcap at all.
>Increasing the limit silently (quite likely accidently) can break a lot of 
>existing programs - like e.g. fweb - , even if it were documented in the 
>man page. Buffer overflow is a major source of programming problems. Let's 
>hope that this kind of errors is not growing in the Red Hat programs, 
>since I use Linux Red Hat much more frequently than cygwin.

[cgf:] To say nothing of security breaches.  I've had 3 BugTraq notices in 
2 days about buffer overrun exploits in code that we include with Cygwin.


>>Sorry, no.  The whole point of mailing lists like cygwin, gdb, gcc,
>>linux-kernel, etc.  is for people to communicate with each other about
>>problems.  I work at Red Hat.  We're an open source company.  I work
>>with *a lot* of free software developers.  I am not aware of any of them
>>who desires personal email regarding their project when there is an
>>existing project mailing list.  In fact, it is company policy that
>>discussions should be carried out in public.
>
>
>There is a simple solution to avoid unwanted email: Set up a different 
>mail account which just responds without insults (even multiple tries) and 
>redirects or returns (maybe dismisses) the email and cite that account in 
>the ChangeLog file.
>I use mainly Fortran and especially the very helpful people from Digital 
>Fortran (nowadays Compaq) respond to every answer very quickly, providing 
>solutions and frankly admit their bugs.

I suppose everyone is going to jump in and say this, so I'm trying to get 
in first ;-) :
The kind folk at Digital/Compaq are getting /paid/ to respond to every 
[question] very quickly!  Chris has other things his masters at RedHat 
expect him to do, and everyone here is on their own time!

>I have _never_ got any response like yours. Maybe the impression, that 
>there might be something wrong with termcap will start to sink in some 
>time. By correcting it you wont help me much (I solved the problem locally 
>before "waisting your patience") but maybe a lot of other people who use
>cygwin to compile and run programs which they find on the net.
>
>Alois
>
>
>
>
>
>
>--
>
>___________________________________________________________________________
>
>Alois Steindl,                  Tel.: +43 (1) 58801 / 32558
>Inst. for Mechanics II,         Fax.: +43 (1) 58801 32598
>Vienna University of Technology,
>A-1040 Wiedner Hauptstr. 8-10   Email: Alois.Steindl+E325@tuwien.ac.at
>___________________________________________________________________________
>
>
>
>--
>Want to unsubscribe from this list?
>Check out: http://cygwin.com/ml/#unsubscribe-simple

David A. Cobb, Software Engineer, Public Access Advocate, All around nice guy.
Get my PGP key at
:< http://pgpkeys.mit.edu:11371/pks/lookup?op=get&search=superbiskit >
Fingerprint=0x{6E3E_DB8C_2E8C_4248_62B2_FE29_08EE_CF0A_3629_E954}
"By God's Grace I am a Christian man, by my actions a great sinner."
--The Way of a Pilgrim, R. M. French [tr.]
<---.----!----.----!----.----!----.----!----.----!----.----!----.---->


--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple