Re: Current and upcoming toolchain changes for jessie
On Tue, May 07, 2013 at 10:03:58AM -0600, Adam Conrad wrote:
> On Tue, May 07, 2013 at 05:48:52PM +0200, Aurelien Jarno wrote:
> >
> > Even if I have to admit I currently don't have a lot of time, it would
> > have been nice to keep the other people in the team in the loop about
> > such an upload.
>
> You'd been fairly inactive of late, and I felt I'd take some initiative
> here instead of asking you about every upload. Apologies, no offense
> was meant. I'd been told that you're fairly busy with real life, so
> was trying to pull up some slack, that's all.
That's true that I have been inactive, and I don't think you should ask
for permission about every upload. However for major toolchain changes,
it seemed obvious to me that you should coordinate with other people in
the team, as well as the release team.
> > > The other unfortunate issue with the glibc bump to 2.17 is that it
> > > has left kfreebsd behind with some pthread mutex changes that need
> > > some love from BSD porters to untangle. We had hoped that leaving
> > > it FTBFS in experimental for several months and gently pinging would
> > > have been enough of a hint, but we can't wait forever either. For a
> > > porter who knows FreeBSD and pthreads inside out, the work to fix
> > > glibc 2.17 on kfreebsd should be trivial. Patches welcome.
> >
> > I haven't been able to find mails about that on the debian-bsd mailing
> > list. Could you please point me to them?
>
> It was discussed on IRC with both 2.16 and 2.17 with you, among other
> people. I gave a stab at fixing some of it, but my BSD/pthreads know-
> how wasn't up to the task. You're right that I didn't mail the list,
> however.
The main glibc BSD developer is not on IRC, but is actively reading the
mailing list.
> > > For the Debian glibc maintainers, Adam Conrad
> >
> > Now I understand why I was not in the loop, I am not member of the team
> > anymore. Glad to learn that in such an email.
>
> You're clearly still a member of the team. That was just how doko
> signed the email from both of us, I imagine he also didn't ask all the
> GCC uploaders/committers for their permission on his uploads either.
He is the only uploader, so there is no really point to ask others...
> Again, no personal offense was meant here, and the beginning of a new
> release seemed like the optimal time to make major toolchain changes,
> rather than waiting several months. The bugs had been filed (and many
> fixed) long ago. Other than the kfreebsd situation, this isn't a
> particularly nasty transition, per se.
I love your way of thinking. eglibc and thousand of packages (the memcpy
change makes every package using memcpy depends on libc6 >= 2.14) are
blocked in sid until eglibc is fixed on kfreebsd. Yeah, this isn't a
particulary nasty transition.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
Reply to: