Re: Bit from the Release Team: Status of hppa
On 2010-10-20, Bill Allombert <Bill.Allombert@math.u-bordeaux1.fr> wrote:
> What is omitted is the rationale for doing so at this stage. Removing an
> release architecture with an active porter group is not something that
> should be done on the ground of some administrative evaluation table.
You surely read the IRC meeting logs? And no, this wasn't an easy decision. I
cannot agree with a porter group, though. There is (was?) one very active
porter, and we were sad to announce it, too. But given the still unresolved
trouble of multithreading on hppa which blocked quite a few packages, this
was a decision we needed to take at some point. We did leave the window open
quite a bit and let the porters time to resolve the issues, but it did not work
out in time. But then I guess you're closely familiar with our processes of
evaluating an architecture and are aware of the issues we faced (otherwise
you would imply that we really did use the table only, which is plain wrong).
And yeah, this stage is about the only stage where we *needed* to take a
decision.
Furthermore, -hppa was basically dead post-announcement, despite dannf trying
to drive a squeeze-hppa release. You might still be able to make that happen
by following up there. But I guess you should go into debug mode for the
threading code in glibc and the kernel, then.
Kind regards,
PhilipP Kern
Reply to: