GSL for Windows (ha, ha, ha)

Robert G. Brown rgb@phy.duke.edu
Tue Mar 16 12:38:00 GMT 2004


On Tue, 16 Mar 2004, DENISON Francis wrote:

> >> Is there any particular reason its nearly impossible to compile GSL for 
> >> windows using cl??  In my opinion, this isn't very bright if you wish to 
> >> have as wide an audience as possible!!
> 
> >There was a discussion a short time ago with the link to ATLAS project
> >FAQ.

See:

  http://math-atlas.sourceforge.net/faq.html#utone

to be precise.  It points out that one way to NOT have the developers
feel great about helping you is to include the observation that they are
not very bright in the request.

> Maybe it would help if people had some positive feedback from windows users?
> I first used GSL about 2 years ago - I had NO programming experience & had
> to teach myself from scratch. I'm a windows user - I don't have much choice
> at work.... I found Dev-Cpp (windows GPL C/C++ IDE) and the pre-compiled
> windows GSL library. It took quite a while to figure out how to install/link
> etc. but I managed by searching old posts from the gsl-discuss archive so I
> never sent any "Hi I'm a clueless windows n00b and GSL sucks because I can't
> figure out how to install it" messages. Quite the contrary - GSL and Dev-Cpp
> introduced me to the GPL & GNU and has given me a VERY positive impression
> of Free Software. I can understand why developers and maintainers of GSL
> want to "raise the bar" by not providing pre-compiled windows libraries, but
> for every "clueless n00b" message you used to get there was possibly someone
> else like me who learned a lot & got a great impression of Free Software. If
> the compiled libraries hadn't been available the bar would have probably
> been too high for me... 
> So to anyone who has contributed to GSL: 
> THANK YOU!

...and this is the way it works in the free software world.  The
products are developed by the people who use them.  Period.  Not one
project in fifty has developers that are employed by somebody and paid
for the sole purpose of developing a free software package for use by
others (for obvious reasons, given the economics of it).

Most of the projects were conceived and begun by people who needed the
tools involved for OTHER aspects of their own work (or for work being
done by others at their site).  Once underway, they put the projects up
under an open/free license and anyone on the Net who wishes can both
take advantage of the fruits of their labor for free and can join up and
help if they wish OR if the product doesn't quite meet their needs and
they wish to modify or augment it so that it does.  After all, if they
have to figure it out enough to change it, why not make their
constructive changes available for still others to share?  

It is this progressive cycle of contribution and evolution by people
that actually use the tool(s) (not people trying to meet some abstract
design criteria thought up by marketing and engineering together) that
makes the open development process so powerful and yes, so sparse.
Fluff gets added only when somebody needs the fluff or has some spare
time and loves the tool enough to fluff it up.

So there are two phrases that anybody that USES free software should use
early and use often, to the extent that their nature allows.  One is
"Thank you!" -- never out of line, given that it is the only reward many
developers get (aside from the ability to use their own tools;-).  The
other is "How can I help?".

SO, the way to REALLY deal with the issue of poor Windows support in the
free software world is for each person that uses GSL under Windows to
try to contribute some of their work back to the project.  This work
need not be coding -- in fact, in many cases contributing DOCUMENTATION
is more important.  For example, if you've learned some ritual that
compiles the entire library for use under Windows, write a document
explaining how (called a HOWTO).  There are templates for doing so and
there is a website for their distribution at www.tldp.org.

Writing documentation will usually either be done semi-collaboratively
or at least in close contact with the developers (who will have
suggestions and might want to review it for correctness).  If this
interaction is positive, and it isn't too difficult, they might well be
willing to make changes to the toolset to accomodate any special
requests or changes made to headers (for example, instrumentation with
#ifdef's) to facilitate the product's operation under Windows).  What
they'd really like is colleagues and partners, not whines.

The other alternative is to pay someone to do it for you.  If you use
Windows, you're used to paying for software, so pay for a programmer
instead (you can just about afford to hire one with what you save
relative to commercial library licenses).  Get them to import and
compile the library so that it is link ready, and keep them around as
consultants to help you over the rough spots.  With the GPL, you can
even turn around and sell the resulting "product" as long as you make
the source for all your changes and packaging openly available along
with the product itself.

   rgb

-- 
Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb@phy.duke.edu





More information about the Gsl-discuss mailing list