This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Making GNATS fixes a release criterion


>>>>> "Martin" == Martin v Loewis <martin@loewis.home.cs.tu-berlin.de> writes:

    Martin> Hi Mark,

    Martin> Given that GCC bug reports are collected in GNATS now, I'd
    Martin> like to propose that bug classification is used to
    Martin> determine release of GCC as well.

This is a good suggestion.  I've added something along these lines to
the release criteria document. (Well, I have done so locally;
gcc.gnu.org is down, so I can't check in the change yet.)

I think we'll certainly need to peruse the open bugs, both in GNATS
and elsewhere, to make sure that critical bugs have been fixed.

I think it would be great if when we decide a bug is critical, it went
it GNATS and was flagged as high-priority.  So I proprose that there
be, as of now, no high-priority bugs in GNATS.  Then, we can make bugs
high-priority if we decide they're important enough for the release.

To start, let's say that maintainers make those decisions.  (So, for
example, Jeff Law could decide that an HPPA bug is critical, and make
it a high-priority bug.)  Or, Jason Merill could decide that a C++ bug
is critical, and do likewise.

To summarize:

  - I'd like all the presently high-priority bugs to be moved to 
    a lower priority.

  - I'd like you to prepare a web page that explains how to enter a
    new bug in GNATS and how to make it have high priority.

  - I'd like maintainers to start following Martin's instructions, and
    to make bugs have high-priority.  

    Note that doing so means that you are saying the release cannot
    ship until that bug is fixed, so taking this step should be done
    sparingly, and in proportion to your willingness to make sure the
    bugs gets fixed, even if a volunteer to fix cannot be found.

  - In my capacity as a C++ co-maintainer, I'd like to delegate to you
    the authority to decide critical C++ bugs, in addition to Jason 
    and myself.

If anyone has objections to any of these points, feel free to shout.
(I'm not exactly sure how much authority to make these kinds of calls
I'm supposed to have; if I overstep the boundaries, I trust the SC
will ask me to back off.)

Martin, I hope you're willing to tackle the above chores.  Your
assistance would be much appreciated.

Thanks,

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]