Re: Two GR concepts for dicussion
On Fri, Jun 01, 2007 at 03:15:12PM +0100, Simon Huggins wrote:
> On Fri, Jun 01, 2007 at 02:22:58PM +0100, Ian Jackson wrote:
> > Anthony Towns writes ("Two GR concepts for dicussion"):
> > > I think the process should involve:
> > > [...]
> > This sounds like a good idea to me.
> > One thing that would be really helpful would be an ability for a
> > Maintainer of this kind to make updates without review iff it can be
> > shown to be safe. (Where `safe' means `the Maintainer gets to screw
> > over people who run this program, but not anyone who doesn't.)
> I think if we have multiple developers recommending them they ought to
> be beyond the stage of fucking things up to the degree that this level
> of inspection of what they are uploading is necessary.
I think there are *DDs* who are not beyond the point where such inspection
is necessary, they're just beyond the point where it happens.
But then, some sponsored NMs are also beyond the point where it happens. So
I don't think the DM proposal hurts us at all in this respect.
> > I'm not sure exactly what the criteria would be but basically you'd
> > diff the previous and new packages and allow only certain kinds of
> > changes (eg, changes to existing programs in /usr/bin would be fine).
> In what ways can maintainers of packages generally screw over users of
> other packages? Don't people notice fairly soon and certainly before
> the packages are out of unstable?
> I imagine this is easier with library packages with many dependent
> packages but I can't imagine those would often be maintained by DMs.
If DMs not maintaining libraries is how you expect this problem to be
mitigated, you might want to consider making this an explicit policy.
Steve Langasek Give me a lever long enough and a Free OS
Debian Developer to set it on, and I can move the world.