This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [wwwdocs] Initial criteria.html (was Re: Where's the currentlist of release criteria?)
- To: Phil Edwards <pedwards at disaster dot jaj dot com>
- Subject: Re: [wwwdocs] Initial criteria.html (was Re: Where's the currentlist of release criteria?)
- From: "Joseph S. Myers" <jsm28 at cam dot ac dot uk>
- Date: Tue, 30 Oct 2001 18:23:03 +0000 (GMT)
- cc: <gcc at gcc dot gnu dot org>
On Tue, 30 Oct 2001, Phil Edwards wrote:
> I saw your new directory in the -cvs list, and I'm checking in a new
> criteria.html now. The process was:
>
> - add a warning at the top
> - change the version number
> - remove the list of criteria WRT specific sections of the compiler; the
> section names remain (e.g., "Preprocessor"), but the contents are empty
> - add more applications for C++ testing
Other things that need to change:
* There is now an Ada front end, which needs to be mentioned in the
appropriate sections. Do we expect to have an Ada testsuite in-tree
before the branch?
* "Java Front-End Garbage Collection" is no longer an appropriate item.
"Installation Documentation" and "GCC Support Library" may well no longer
be. Given that the development plan is based on dates, and the date for
major new features has passed, it might be better to start that section
with "Open Bugs" only and then consider whether there are any features
that need to be done.
* Chill should probably not be mentioned at all (unless our volunteer does
the necessary work in time).
* Regression testing is presumably based primarily on regressions from
3.0.x. Do we try to avoid regressions from 2.95.x as well?
* At least two of the listed open issues should go away: -fstrict-aliasing
and setjmp/longjmp exception-handling.
I think we should primarily try to fix bugs, improve documentation and
ensure that application tests work (given volunteers, application tests
should include building complete BSD systems, including packages/ports,
and GNU/Linux distributions); secondly, deal with performance issues; only
thirdly consider new features as criteria.
--
Joseph S. Myers
jsm28@cam.ac.uk