Path: utzoo!utgpu!watserv1!watmath!att!linac!uwm.edu!bionet!agate!ucbvax!bloom-beacon!eru!hagbard!sunic!mcsun!hp4nl!charon!timo
From: timo@cwi.nl (Timo Krijnen)
Newsgroups: comp.sources.d
Subject: Re: remove() and Unix (was Re: problems building abc)
Message-ID: <2776@charon.cwi.nl>
Date: 14 Jan 91 09:28:51 GMT
References: <53567@eerie.acsu.Buffalo.EDU> <1991Jan08.140043.12986@ctr@italy.eu.net> <50078@prls.UUCP>
Sender: news@cwi.nl
Organization: The ABC Team, CWI, Amsterdam
Lines: 30

In article <50078@prls.UUCP> gordon@prls.UUCP (Gordon Vickers) writes:
>
>     From a recent article in this group that addressed this problem
>     [the recently posted program 'mbase' (AKA dox) calls remove() ],
>     it seems that the remove() call can be replaced by unlink() .
>
> -Gordon

Except in ABC, because this implements the removal of an entry from
a list (a datastructure in ABC), e.g. to implement

	REMOVE x FROM list.of.x's

We have come accross many other examples of names we used for our own
internal routines, which later showed up in standard libraries.
Just look at the list of redefines at the start of bhdrs/bobj.h.

The most notorious example is "compound()" used in the parser to
parse an ABC compound (like a structure in Pascal), which is in the 
libraries on our Sparc+'s to implement compound interest, sic.

It seems you should not use any common name, since eventually you will
run into some system where it is an internal:-).

Groetjes, timo.
-- 
Groetjes,
 Timo Krijnen
 CWI (Center for Math. & Comp. Science), Amsterdam
 timo@cwi.nl