From: Andrew Crabtree <andrewc AT typhoon DOT rose DOT hp DOT com>
Message-Id: <199801291756.AA153226561@typhoon.rose.hp.com>
Subject: Re: iostream concern
To: eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
Date: Thu, 29 Jan 1998 9:56:01 PST
Cc: djgpp-workers AT delorie DOT com
In-Reply-To: <Pine.SUN.3.91.980129113402.6822R-100000@is>; from "Eli Zaretskii" at Jan 29, 98 11:37 am
Reply-To: andrewc AT rosemail DOT rose DOT hp DOT com
Precedence: bulk

 
> FWIW, I have never understood why would GCC want to run fixincludes on 
> ANSI- and POSIX-compliant headers.
You could check the egcs mail archives to see the discussion that 
Jeff Law and I had on this (www.cygnus.com/egcs).  But basically, here
are a couple of reasons (they are not much though) ....

1) The ansi checking is getting better in newer versions of gcc so its 
possible that headers that you think are ansi compliant actually aren't
(like the c++ incompatibility with exception)

2) Some header use internal gcc things that can change between releases
(like using gnu_va_arglist instead of the regular ansi definition)