Re: Baby's First AldyVec/AltiVec patch

[Matt Austern <austern at apple dot com>]

On Dec 23, 2003, at 10:34 AM, Mark Mitchell wrote:

> > I hope people are willing to reconsider, at least to the extent that
> > this
> > will be accepted in a non-default mode.
>
> With my CodeSourcery hat on:
>
> - Our customers would like to have the old syntax restored.  However,
> they have adapted to the new syntax.
>
> With my FSF hat on:
>
> - I'm not at all eager to have yet another language extension,
> especially when there is a perfectly reasonable way of expressing the
> idea in our current vector syntax.  At the very least, I think that we
> need to have a document that explains how the Motorola syntax interacts
> with ISO C and C++, if at all.  The parenthesized form is much more
> likely to be in conflict with *something*.

I think documenting the interaction is a reasonable request.  What I can
say: empirically, Apple, Metrowerks, and IBM have all managed to 
introduce
the Motorola syntax so that it has essentially no conflicts with any
standard syntax.  (Literally the only conflict I know of was one parsing
bug, in a fairly obscure corner case, that we fixed about six months 
ago.)

And remember the downside if we don't introduce the Motorola syntax:
Altivec code won't be portable between the various compilers that all
claim to support Altivec.  That's a very serious problem, and it's 
exactly
what standards are supposed to prevent.

			--Matt