Newsgroups: comp.lang.apl
Path: watmath!watserv1!torn!utcsri!rpi!usc!elroy.jpl.nasa.gov!jato!csi!sam
From: sam@csi.jpl.nasa.gov (Sam Sirlin)
Subject: Re: SIGNUM of teaching numerical methods
Message-ID: <1992Jul20.160438.26974@csi.jpl.nasa.gov>
Originator: sam@kalessin
Sender: usenet@csi.jpl.nasa.gov (Network Noise Transfer Service)
Nntp-Posting-Host: kalessin
Organization: Jet Propulsion Laboratory, Pasadena, CA
References: <1992Jul17.171448.13672@chpc.utexas.edu> <1079@kepler1.rentec.com> <ROCKWELL.92Jul19155653@socrates.umd.edu> <1089@kepler1.rentec.com>
Date: Mon, 20 Jul 1992 16:04:38 GMT
Lines: 27


In article <1089@kepler1.rentec.com>, andrew@rentec.com (Andrew Mullhaupt) writes:

|>... I can give a spec for a language which would answer _all_ my criticisms
|> without even coming close to needing any _new technology_. In fact the
|> ANSI Standard language Extended Pascal is 'almost there'. This is not
|> a style issue - I could make something that was very APL-like which would
|> do everything needful. The annoyance is that none of the established
|> languages really does a good job, but that it really isn't that hard to
|> do one.

I'd be interested in your spec. From your posts it seems it would include

1. LAPACK
2. sparse matrices & routines
3. a "good" interface to other (compiled) languages, specifically Fortran
  - including call by reference
4. access to "primitive level coding"

I personally am interested in all but 2. (I don't generally deal with
huge problems). So far though I've just added the linear solution part
of LAPACK to APLc.

-- 
Sam Sirlin
Jet Propulsion Laboratory         sam@kalessin.jpl.nasa.gov

