In article <1pggn0$amq at overload.lbl.gov> bks at s27w007.pswfs.gov (Bradley K. Sherman) writes:
>If you guys keep thinking of programmers as lab techs we're not going
>to get anywhere.
First of all, let me apologize for my intemperate remarks and
especially to lab techs without whom I could not do my work.
Writing software for researchers is a difficult proposition.
It is very common for a single successful run of a program
to be the last use of that program! It is often the case that
a simple graph or picture is the culmination of hundreds of hours
of moving data from one form to another, filtering out
noise, and applying a special purpose mathematical transform
-- often some arcane algorithm thought up by Gauss after one
too many Schnapps on a chilly night in 1843 and never before
implemented in any form more concrete than a doctoral dissertation.
By the time the software is cast into a robust form, thoroughly
vetted and outfitted with its direct manipulation interface
for Curses, Windows , Windows NT, Mac, OS/2, OpenLook, NeXtStep
Motif, VMS, ... and then retested for window manager idiosyncrasies
the problem it solves may be of no interest.
Bradley K. Sherman, Dendrome Project
Institute of Forest Genetics, USDA Forest Service
P.O. Box 245, Berkeley, CA, 94701, USA
Phone: 510-559-6437 FAX: -6440 Internet: bks at s27w007.pswfs.gov