Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!lll-crg!nike!oliveb!hplabs!hao!nbires!rcd
From: rcd@nbires.UUCP (Dick Dunn)
Newsgroups: net.lang.c
Subject: Re: "If it has a bug, it isn't 'C'"
Message-ID: <608@opus.nbires.UUCP>
Date: Mon, 6-Oct-86 23:38:53 EDT
Article-I.D.: opus.608
Posted: Mon Oct  6 23:38:53 1986
Date-Received: Wed, 8-Oct-86 00:56:29 EDT
References: <1215@bunker.UUCP>
Organization: NBI,Inc, Boulder CO
Lines: 21
Summary: bugs vs features

> >Then it isn't a C compiler...
> 
> This is rapidly becoming one of my pet peeves.  It seems that it is
> with increasing frequency that I see comments to the effect that if
> a C compiler has a bug, then it isn't a C compiler.  This is pedantic;...

Gary's complaint is valid when people are talking about things which turn
out to be bugs - meaning something along the lines of "behavior which the
implementors agree is incorrect and should be fixed."

But that only accounts for about half of the circumstances that elicit the
"it isn't a C compiler" response.  The rest are improvements foisted on us
by implementors who decided they had a better idea and implemented it
without understanding the language.  An example of the latter showed up
just recently, where some implementation had made "long float" a different
type from "double".  As long as people come up with gratuitous incompatible
extensions to the language, they can expect the wrath of knowledgable
users.
-- 
Dick Dunn    {hao,ucbvax,allegra,seismo}!nbires!rcd	(303)444-5710 x3086
   ...Nothing left to do but smile, smile, smile.