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: Beyond GCC 3.0: Summing Up



> I'd prefer to delay such an action until a release is imminent, on the
> grounds that a failing test is a useful piece of information we shouldn't
> lose (of course this requires that we keep track of why it's failing).

If a release is imminent, and we have known testsuite failures, we are in
trouble.  That means we have to try to fix those bugs, and all the other
bugs we will inevitably find when testing for the release.  This is part
of how we get to having such a long release cycle.

We have to remember that we are not a normal development organization.
There is *no mechanism* for making anyone fix changes, ever.  We
cannot simultaneously meet the goal of (even semi-) regular releases,
and high quality releases, without making someone responsible for
fixing newly introduced bugs.

In any case, I came to a very basic conclusion during the 3.0
release cycle.  If we allow buggy checkins (again note earlier
compromise regarding "latent" bugs), then the job of Release
Manager is not a management position at all.  It is the job
of Cat Herder.  And there are a lot of cats.

We need to compromise between the bazaar and the bizarre, here,
folks.


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