This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: How we train numerical programmers
- To: Eleftherios Gkioulekas <lf at amath dot washington dot edu>
- Subject: Re: How we train numerical programmers
- From: "Gregory R. Warnes" <warnes at biostat dot washington dot edu>
- Date: Mon, 17 Jul 2000 13:09:10 +0000 (GMT)
- cc: edwin at netwood dot net, gsl-discuss at sourceware dot cygnus dot com
On Sun, 16 Jul 2000, Eleftherios Gkioulekas wrote:
EG>> The problem is that desperately needed innovations in software
EG>> do not count as "research" in our community. ...
This is also true in my discipline. Anyone who wants to work on software
tools has to either flee academia to one of a few research labs or do
software 'on the side' without getting any credit. Those who can't or
wont do one of those two things end up moving out of the field.
This is particulary unfortunate because my field is data driven and the
available software *dictates* what methods are used by the majority of
practitioner. Worthy methods that don't make it into software never get
used.