Subject: Re: ARM ELF ABI: consensus?
To: Todd Vierling <firstname.lastname@example.org>
From: Richard Earnshaw <email@example.com>
Date: 04/02/2002 15:10:07
> On Tue, 2 Apr 2002, Richard Earnshaw wrote:
> : Yes, I'm happy with this provided that...
> : firstname.lastname@example.org said:
> : > [*] "may", because given our past history with these types of things,
> : > we may decide just to keep "our" ABI in the future and be done with it.
> : ... this is not part of the plan.
> There's something to be said for keeping an ABI once a system release is
> out. I cringe at the thought that we're putting out a release at all with
> an ABI as screwed as the one we have now, but changing it *yet again* is
> bordering on severe annoyance.
> Hell, ATPCS isn't even `official' yet. That's reason enough to me to Pick
> an ABI, and then stick with it.
I've said this before, so I'll say it again. I don't support the long
term creation an ABI that is not compliant with the ATPCS. The
alternative to having an *interim* ELF-based ABI is to stick with a.out
until we can switch cleanly.