[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bits from the ftpmasters

On Sun, Feb 20, 2005 at 12:55:37PM -0800, Steve Langasek wrote:
> On Sun, Feb 20, 2005 at 12:38:07PM -0700, Joel Aelwyn wrote:
> > One of my 'real problems' is a completely stalled NEW queue; secondhand
> > comments indicate that vorlon may have implied that it won't be processed
> > for a while (maybe until after Sarge releases) because "Do we really want
> > more packages in unstable while trying to freeze?" (this could be flat
> > wrong, of course, and simply misreported).
> No, what I said was that people shouldn't make a *fuss* about the lack of
> NEW processing, because there are 108 open RC bugs right now loudly
> declaring that there are more useful things they could be doing for the
> project with their time before criticising how the ftpmasters are spending
> theirs.  The release team has not asked the ftpmasters to stall NEW
> processing; I've merely made an observation about the negative impact NEW
> processing (or any code churn of a significant nature, really) has on the RC
> bug count in testing.

Then call it misrepresented or misunderstood. I would still ask, however,
whether any of those 108 bugs are the direct consequence of a NEW
processing run, or simply the consequence of "Debian has a fricking huge
number of packages". The latter may be an issue, but the answer to it isn't
"stop processing NEW".

If the answer is "Testing does not help because code churn in unstable
is still making it through and breaking things", then maybe we have no
choice but to go back to using frozen, because testing isn't doing what
it was designed to do. Or maybe the answer is that post-Sarge, we review
and adopt one of the other release methodologies. I've been leaving that
discussion until post-Sarge, for the obvious reason (not that it doesn't
help release Sarge, but that trying to change Sarge this close would be
insane, so there's no use in discussing it for Sarge anyway).

Note that I would, in fact, be OK (true, I wouldn't *like* it, but I would
understand it) if it really came down to the Release team saying "OK, this
isn't working, we need to block <X>". That's the job we've asked you to do.
I'd probably try to make sure we figured out how to avoid the problem in
the future, sure, but that isn't the same.

None of my packages have RC bugs. I regularly skim the open list and see if
there's anything I can meaningfully contribute to, though many of them are
beyond what I can manage to accomplish, often because they require intimate
knowledege of the package in question. So I don't particularly feel like
this discussion is somehow delaying Sarge; I'm already doing what can be
done on those 108. But I'll go look again.
Joel Aelwyn <fenton@debian.org>                                       ,''`.
                                                                     : :' :
                                                                     `. `'

Attachment: signature.asc
Description: Digital signature

Reply to: