Re: Application development (entirely) in C

[shenkin : at : avogadro.barnard.columbia.edu (Peter S. Shenkin)]

> From: Leif Laaksonen <Leif.Laaksonen : at : convex.csc.fi>
 > Do you Peter know if most of the new scientific programs (QC, MD/MM ..)
 > are still written in fortran or is c slowly takin over? Do you know
 > if any quantum chemistry program package is written in c  (or any
 > other language than fortran)?
 I don't know -- but let me just remark that many people think that C++
 is likely to be more successful than C for scientific applications.
 > Do you have any benchmark results for any real problems where there
 > exist an algorithm both in c and fortran, I would be very interested
 > to see the results. So much about the tools!
 I don't, personally, but have seen many of these in comp.lang.fortran,
 and as I said before, there is no consensus.  Also, I think you mean
 "program," rather than "algorithm," in the last paragraph.
 Changing
 the algorithm can have a vastly larger effect than changing the language
 the algorithm is expressed in.
 > How about discussing applications now. What would be on top of your
 > shopping list if you would be asked? What kind of software would you
 > need most now? Are you happy with the graphics tools available or
 > do you see the numbercrunching part more important?
 No one who designs software for my application (visualization of and
 calculations on protein structure) has read my mind well enough to
 put into the program all the things I want to do.  Nor will they
 ever.  Therefore what we need is programs which both have high-
 efficiency underlying routines for the operations everyone needs,
 such as rotating bonds and updating xyz coordinates, on the numerical
 side of my application,  and  visualizing a molecule and portions of
 it, and rotating it in 3d, on the graphical side.  But instead of
 trying to supply native features for everything everyone might
 want -- a hopeless task -- I would much rather see a highly flexible
 4th generation language as a front end, which would allow the user to quickly
 implement and incorporate the things he/she dreams up.  I'm thinking
 of something with the functionality, but not the syntax, of APL.
 	-P.
 ************************f*u*cn*rd*ths*u*cn*gt*a*gd*jb*************************
 Peter S. Shenkin, Department of Chemistry, Barnard College, New York, NY 10027
 (212)854-1418    shenkin : at : avogadro.barnard.columbia.edu   shenkin : at :
 cunixc.BITNET
 ********** "I've got algorithm -- who could ask for anything more?"
 **********