Iterators (was: Re: GCC's statement expression extension)

["Joseph S. Myers" <jsm28 at cam dot ac dot uk>]

On Thu, 27 Jul 2000, Mark Mitchell wrote:

> This isn't really my call (which is a good thing).  I'm just finding
> out how many people agree.  Even before touching C++, I'd want Jason
> Merrill's agreement.  Before doing anything to C, we'd certainly need
> the SC to sign off -- and I don't plan on raising the issue, at least
> in C, based on these discussions.

While discussing removal or deprecation of extensions: I think the removal
of iterators needs considering again.  This was discussed two years
ago; in http://gcc.gnu.org/ml/gcc/1998-06/msg00760.html it was suggested
that something similar might appear in c9x (nothing relevant in C99) and
that it might be useful for MMX (or similar support), and that:

> The iterator scheme certainly needs some documentation is it is going
> to stay in the compiler though.

Two years later, there is no documentation for iterators, and there are no
testcases for them either.  (Iterators were discussed again in September
1999, again with no conclusion.)

Given that in two years no-one has contributed any documentation or
features for this extension, perhaps the people working on anything
vaguely related to MMX could comment on whether iterators have in fact
proved to be relevant; but in the absence of a clear use and a commitment
to document the extension I think following the intention expressed two
years ago that iterators should be removed if undocumented is long
overdue.  Various patches have been produced in the past.

Glibc does not use iterators.  Nor does the Linux kernel.  Those are the
biggest users of GCC extensions.  I'll repeat the doubt expressed in the
past that anyone is using iterators at all.

-- 
Joseph S. Myers
jsm28@cam.ac.uk