Re: glib2.0 on i386
On Sat, Jan 08, 2011 at 08:13:08AM -0800, Andres Salomon wrote:
> On Sat, 8 Jan 2011 12:29:15 +0100
> Kurt Roeckx <kurt@roeckx.be> wrote:
>
> > On Sat, Jan 08, 2011 at 02:44:55AM +0100, Cyril Brulebois wrote:
> > > Andres Salomon <dilinger@queued.net> (07/01/2011):
> > > > This is kind of an important architecture, so it would be nice to
> > > > not have it lag behind so far (it just caused
> > > > http://bugs.debian.org/609277 to occur - while it's nice to catch
> > > > such bugs, the fact that the build failed is a concern to me).
> > > > Does the build just need to be requeued?
> > >
> > > No, a missing or laggy build doesn't *cause* FTBFS. Point out, or
> > > reveal an insufficient version in a build-dep, rather.
> > >
> > > Requeuing is cheap, given back. Please note that you want to talk to
> > > the buildd admins for such issues (Cc'd).
> >
> > I will never give back a package if it clearly has a RC bug in it
> > that made it fail. Fix the bug instead.
> >
> >
>
> Speaking of, does anyone have suggestions for how to deal w/ responses
> like the one in #609277 ("it's not an RC bug", with a
> downgrade). I point people to policy 4.2, but that can be interpreted
> broadly.
In: http://release.debian.org/squeeze/rc_policy.txt
4. Autobuilding
Packages must list any packages they require to build beyond those
that are "build-essential" in the appropriate Build-Depends: fields.
[...]
Packages must autobuild without failure on all architectures on
which they are supported.
The Build-Depends should contain the correct version. And if
that's still not clear to some ask the release team to add that
because that's always how it has been.
Kurt
Reply to: