This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Question about #pragma removal
- From: Bo Thorsen <bo at sonofthor dot dk>
- To: dewar at gnat dot com (Robert Dewar), brett dot cook at intel dot com,gcc at gcc dot gnu dot org
- Date: Fri, 15 Mar 2002 10:47:39 +0100
- Subject: Re: Question about #pragma removal
- Organization: SuSE Labs
- References: <20020314165244.D8B64F28C5@nile.gnat.com>
On Thursday 14 March 2002 17:52, Robert Dewar wrote:
> <Basically, all this does is create further extensions to the C language
> that are non-ansi. Seems like Microsoft does the same sort of thing all
> the time (take JAVA for example).
>
>
> I think people should look at the Microsoft Java extensions carefully
> before casually using this as an example of unwarranted extensions. In fact
> they made perfectly reasonable sense to me. There is nothing inherently bad
> about something being an extension. GNU C has lots of valuable non-standard
> extensions (nested functions being one of the most clear examples).
Perhaps you should follow your own advice first.
Microsoft changed the semantics of two lowlevel methods and changed the
arguments to another. And their extensions were not just language extensions,
they needed vm changes too.
They broke source code compatibility (granted, that would only trigger for
0.1% or less of the code out there) and they broke binary compatibility.
I'd say that qualifies as stupidity beyond reason.
And I do not think there's anything in the MS extensions you couldn't do with
inner classes.
Bo.
--
Bo Thorsen | Praestevejen 4
Free software developer | 5290 Marslev
SuSE Labs | Denmark