Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!elroy.jpl.nasa.gov!swrinde!mips!dimacs.rutgers.edu!rutgers!cmcl2!lanl!cochiti.lanl.gov!jlg
From: jlg@cochiti.lanl.gov (Jim Giles)
Newsgroups: comp.lang.c
Subject: Re: low level optimization
Message-ID: <22027@lanl.gov>
Date: 19 Apr 91 23:09:13 GMT
References: <21660@lanl.gov> <1991Apr17.225944.15261@zoo.toronto.edu> <1991Apr18.190403.29049@Think.COM> <21846@lanl.gov> <1991Apr18.233807.19552@zoo.toronto.edu> <21868@lanl.gov> <21227@paperboy.OSF.ORG>
Sender: news@lanl.gov
Organization: Los Alamos National Laboratory
Lines: 25



> This has never been the case.  Using pre-ANSI syntax:
> B:
> 	some_routine(123);
>
> A:
> 	some_routine(arg)	{...
>
> A after editing:
>	some_routine(flag, arg)

Irrelevant to this discussion since it was always clearly understood
(and clearly stated at the beginning of the discussion) that the valid
means of communication between separately translated units were external
functions and external variables.  If you introduce a change which makes
these incompatible, you have stepped outside the domain of this discussion.

The main purpose of separate compilation is two-fold: to save compilation
time when only part of the code has been modified and to allow a vendor
to alter his libraries for revisions without having to distribute the
_whole_ library again or distribute source.  Presumably, incompatible
revisions require more effort than the more common compatible ones.

J. Giles