home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!swrinde!mips!darwin.sura.net!uvaarpa!cv3.cv.nrao.edu!laphroaig!cflatter
- From: cflatter@nrao.edu (Chris Flatters)
- Newsgroups: comp.lang.fortran
- Subject: Re: Scientists as Programmers (was Re: Smal
- Message-ID: <1992Aug27.215104.2917@nrao.edu>
- Date: 27 Aug 92 21:51:04 GMT
- References: <9224105.4585@mulga.cs.mu.OZ.AU>
- Sender: news@nrao.edu
- Reply-To: cflatter@nrao.edu
- Organization: NRAO
- Lines: 28
-
-
- In article 4585@mulga.cs.mu.OZ.AU, fjh@munta.cs.mu.OZ.AU (Fergus James HENDERSON) writes:
- >minch@lotka.stanford.edu (Eric Minch) writes:
- >
- >>... in my experience FORTRAN-writing scientists
- >>can whip out working algorithms just fine. The problem comes when they
- >>have to deal with data structures more complex than arrays, programs
- >>longer than about a thousand lines, and mainly when turning algorithms
- >>into maintainable production code. I've sometimes wished that
- >>scientists were taught in a language with no I/O statements.
- >
- >Maybe you should try teaching Haskell to upcoming scientists.
- >Haskell is a pure functional programming language (with lazy evaluation)
- >that has no statements that cause side-effects. I/O is programmed by
- >considering the program as a function from input to output, or more
- >generally a program is a function from one operating system state to
- >another (roughly speaking).
- >
- >Personally I think that a clear separation of functions which have side-effects
- >from those that don't is a very good thing, but there is still a need for
- >those in the former class. ML is probably a better alternative as a
- >fuctional language if efficiency is much of a concern.
-
- How about Sisal. Its I/O facilities are primitive (to say the least)
- but it is demonstrably at least as fast as FORTRAN.
-
- Chris Flatters
- cflatter@nrao.edu
-