Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!ames!hc!lanl!jlg
From: jlg@lanl.gov (Jim Giles)
Newsgroups: comp.lang.fortran
Subject: Re: Exception handling (was Fortran vs C for computations)
Message-ID: <3632@lanl.gov>
Date: 15 Sep 88 17:38:12 GMT
References: <405@quintus.UUCP>
Organization: Los Alamos National Laboratory
Lines: 30

From article <405@quintus.UUCP>, by ok@quintus.uucp (Richard A. O'Keefe):
> [...]
> (2) Mr Giles now reveals that his condemnation of my suggestion is not
>     so much for its present half-baked form as for what some hypothetical
>     standards committee *might* do to it.  Well, of course, they could
>     make it arbitrarily bad.  I can't think of any defence against that
>     method of judgement.

You have half my objection.  You proposal requires two things which I
think are completely unnecessary: 1) cooperation from the standards
committees; 2) modification of the compiler for each language to support
a feature which _should_ be common to all of them.

My proposal is that a single standard be devised (so only one committee
is involved - not one for each language), that the standard be for a
set of callable routines (procedure calls are already sufficiently
similar in most procedural languages to make a common interface possible),
and that these routines should be provided in a single common library
for all languages in a given environment which are call-compatible.
This proposal is _very_ different from yours.
> 
> (3) Correcting <3576@anl.gov> wouldn't get us anywhere.  I repeat what
>     I said in my previous message:  it is time that we had someone else
>     in this discussion, or dropped it entirely.

Fine, go ahead and drop the discussion.  I only responded _this_ time
because you again misrepresented my argument.

J. Giles
Los Alamos