Re: Bug#300170: Consequences of udev being pulled in by Gnome
On Fri, Mar 18, 2005 at 02:58:06PM -0800, Steve Langasek wrote:
> On Fri, Mar 18, 2005 at 12:18:10PM +0100, Sven Luther wrote:
> > On Fri, Mar 18, 2005 at 02:38:35AM -0800, Steve Langasek wrote:
> > > On Fri, Mar 18, 2005 at 10:36:12AM +0100, Marco d'Itri wrote:
> > > > On Mar 18, Sven Luther <email@example.com> wrote:
> > > > > 3) Inform the user that a reboot is needed in order for it to work.
> > > > No, it's not! Months ago I did a lot of work in postinst to support
> > > > reboot-less installation of udev.
> > > > There is not enough data in this bug report to even start investigating
> > > > it but I believe that the problem is caused by #296776 and #296975,
> > > > which I fixed over a month ago but are being kept out of testing by the
> > > > broken arm buildd.
> > > It's also waiting on a new version of makedev, which is frozen and includes
> > > lots of changes not related to the udev change, and has only been in
> > > unstable for two days. Cc:ed to the maintainer (Bdale) as well, for
> > > comments on whether makedev should be pushed through.
> > > It would be good if someone could do a concrete d-i sid test with this
> > > version of udev, to confirm that it does solve the problems in question.
> > Ok, will do, i need a install on powerpc/sid using the daily builds, nothing
> > more, right ?
> Yes. Marco says that the changes have been tested, but I'm not sure if that
> includes a full d-i install, so I'd like to have one on record to be sure.
I didn't do a full install since powerpc unstable was broken yesterday (see,
exactly the reason why dropping testing for non-tier1 arches is not a good
idea and counter-productive), but i installed it by hand and it worked.
That said another guy posted on debian-boot about this thread (or maybe in the
udev bug report, not sure, and he had moved the sid versions to their local
archive and installed from there, and it solved the problem for him, so ...
> > > > Cc'ed to the release managers, who may want to force again udev in
> > > > testing to fix these RC bugs.
> > > I suspect so, but I think it's best to let makedev age a little more before
> > > pushing it in. Joey, is there any d-i deadline I need to worry about for
> > > this? I assume that if we have a known good fix, it's not a problem for it
> > > to not be in the d-i rc3 release, as long as it gets on the official CD
> > > builds for sarge?
> > I believe that this is not a problem for d-i, as udev is not in base, so not
> > part of anything d-i builds, apart from the full isos, which are or not d-i
> > related.
> > The makedev issue could be problematic though, as you say it is part of base.
> > Could it be possible to have a fixed udev built against the testing version of
> > makedev, and push it in through testing-proposed-updates ? With manual builds
> > if needed.
> Based on Bdale's feedback, I'm pushing both makedev and udev into testing