home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.edu:1452 comp.lang.fortran:3341 comp.lang.misc:2847 comp.arch:9130 sci.math:10802
- Path: sparky!uunet!mcsun!uknet!acorn!eoe!pwestlak
- From: pwestlak@eoe.co.uk (Peter Westlake)
- Newsgroups: comp.edu,comp.lang.fortran,comp.lang.misc,comp.arch,sci.math
- Subject: Re: Scientists as Programmers (was Re: Small Language Wanted)
- Message-ID: <1389@eouk20.eoe.co.uk>
- Date: 1 Sep 92 17:23:12 GMT
- References: <1992Aug31.195540.13074@ctr.columbia.edu>
- Organization: EO Europe Limited, Cambridge, UK
- Lines: 36
- X-Newsreader: Tin 1.1 PL3
-
- The general consensus of this thread seems to be that very few people are
- competent in programming and in an application area, especially that for
- which they are hired. Specialization does help to some extent, but it makes
- the chances of finding the right person even smaller. Certainly insisting that
- computing students take science or business courses does not help, because the
- courses are very simple and generalized - inevitably so, because the student
- is trying to take a degree in computing at the same time! I doubt that there
- will ever be a ready supply of people who are expert in computing and in an
- application area. The answer, I suggest, lies in the original topic of this
- thread: Small Languages. Rather than teach physicists (etc) computing or
- vice-versa, give them a set of high-level types and operations with which
- to program their applications for themselves. These will be written
- by someone who knows less about the science and more about programming.
- The "knowledge gap" between computers and applications is so great that
- several levels of abstraction may well be needed, each one hiding all the
- details of the level below. Of course, doing this effectively is a research
- goal in itself, but I suggest that it is a useful one to pursue.
-
- To anticipate possible objections:
-
- Multiple levels are inefficient: not if efficiency is a goal. The levels
- need not involve lots of code; they might be subroutine libraries or threaded
- code, or some other scheme suitable for the job in question.
-
- People writing the intermediate levels will still need to be experts in
- more than one field: yes, but the field will be smaller. They don't need
- to stretch all the way from computer hardware to advanced nuclear physics.
- Also the languages will be reusable by other people working in the field,
- the more so the lower the level.
-
- P.S. Sorry if this is posted twice - I'm trying to find out why it happens.
- ------------------------------------------------------------------------------
- Peter M Westlake EO Computer Ltd.,
- Tel: +223-843131 Abberley House, Granhams Road,
- Fax: +223-843032 Great Shelford,
- email: pwestlak@eoe.co.uk Cambridge CB2 5LQ, England
-