first (very crude) stab at an abstract

[Mark Galassi <rosalia@lanl.gov>]

Dudes, I need to submit this abstract today.  It is very crude,
especially the opening sentence, and I will make some more passes at
it.

Meanwhile, *please* give me any kind of feedback asap.

                     The GNU Scientific Library
                  http://sourceware.cygnus.com/gsl/

The GNU Scientific Library (GSL) is a collection of high quality
scientific subroutines which cover what physicists normally consider
the standard numerical portfolio.

There are many classic FORTRAN libraries which cover similar
territory, but GSL has some important differences in its requirements,
and hence in its design and implementation.

Some specific requirements are that GSL must (in no particular order)
(a) be free software (sometimes called open source software), and in
particular it is distributed under the GNU General Public Licence
(GPL); (b) be implemented in C using modern coding conventions,
calling conventions, error handling and scoping; (c) clearly and
pedagogically documented; (d) use top quality state of the art
algorithms; (e) be portable and configurable using the standard GNU
conventions; (f) be easily used in very high level languages (Scheme,
Python, Perl, Tcl) via automatically generated wrappers.

GSL has been developed by a team of scientists and free software
hackers from several universities, laboratories and companies in the
United States and Europe.  Many important suggestions and patches have
come from GSL end users.

GSL is included in the standard Linux distributions, and is used by
many researchers and some software packages.  We expect our coverage
of the standard numerical portfolio to be complete by the end of 1999,
at which point we will prepare the 1.0 release of the software.
Meanwhile, since the software is hosted on Cygnus's "sourceware site",
snapshots, anonymous CVS access, mailing list archives and so forth
can be found at the URL mentioned above.