Date: Tue, 10 Feb 1998 13:04:35 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: andrewc AT rosemail DOT rose DOT hp DOT com
cc: robert DOT hoehne AT gmx DOT net, dj AT delorie DOT com, djgpp-workers AT delorie DOT com
Subject: Re: iostream concern
In-Reply-To: <199802091822.AA114578565@typhoon.rose.hp.com>
Message-ID: <Pine.SUN.3.91.980210130412.12622C@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Precedence: bulk


On Mon, 9 Feb 1998, Andrew Crabtree wrote:

> The new gxx handles things directly.  I believe that 
> with the old gxx, if you pass it a .c file, it will invoke cc1 on it
> (since gcc decides).  With the new gxx, I think that it will 
> invoke cc1plus on .c files as well as .cc files.

I'd rather imagine that g++ will invoke cc1 on .c files, unless it
uses different rules for computing the default source language.

But this is exactly the kind of subtle changes I was afraid of.
Without a lot of testing, you cannot be sure the programs behave the
same, and I don't think we can ever invest the resources for such a
massive testing.  If somebody assumed certain features or misfeatures
of the old version to exist, the new one will break their work.