Re: Feature Freeze
On Fri, Oct 06, 2006 at 02:19:02PM +1000, Drew Parsons wrote:
> David Nusinow wrote:
> > > I thought you had made a simple mistake in calling the freeze before
> > > they had been processed, so I thought I was providing a service in
> > > helping get them done as quick as possible.
> >
> > It was a simple mistake, you were 100% right in that.
>
> So it's like this. You unilaterally announce a policy without
> consulting with other members of the team. The policy has missed
> something, so a team member takes the initiative to fill in the bits
> which are missing. You respond by pontificating about "team playing",
> using the vilest possible threats.
>
> You don't see a problem in this picture?
>
> Hint: an alternate response might look like "oh damn, I'd forgotten
> about the input drivers. Thanks for getting on to them. Is there
> anything else we missed before ratifying the freeze?"
So what am I supposed to do if someone does something that is, in fact, a
very bad idea and violates the freeze? What do I do if you decided that,
the new i810 driver is a good idea too? And the new ati driver? Or RandR
1.2? The list goes on and on, and it's my responsibility to make sure we
stop somewhere. Branden assigned me this responsibility when letting me
work on xorg in his place, and it's something I take very seriously. Just
because I've freely invited every interested party to work directly on the
archive does not mean that I have given up this responsibility.
With that responsibility comes the right, as well as responsibility, for
unilaterally deciding when to freeze, the same way the Debian RM's and the
Xorg RM's and every other RM in every other project has the right and
responsibility to do this. This is a standard working model for most
projects, and as a result, no, I don't see a problem with this.
Furthermore, I wouldn't have responded the way I did if you had just
committed to svn. But you uploaded to the archive too. Since you have
upload privledges, I *can't* prevent you from doing these things as release
manager, effectively limiting me in my ability to do my job. And in case
you haven't noticed, it's *me* that people in Debian hold directly
responsible for the quality of these packages right now. So it's me that
has to answer to everyone for such problems. So when I come down hard on
you for disregarding the *one* real restriction I have as RM, that I be
allowed to freeze at release times, try to keep it in perspective. We could
always make it a pull-only system, but neither of us want that.
So we have a few options. Either I kick everyone out of uploaders but me
and consider any freeze-breaking upload to be a hijack, or we cooperate.
Those are the only ways that I can see where I can effectively do my job.
If you don't think we should have an RM for the team, or that I'm doing a
bad job of it, then let's make that the issue. But as it stands, I'm still
the RM, and I need to have the ability to do that job. Now I'd prefer not
to go the route where I'm the only uploader, but I am willing to do that.
After all, that's how it was for 6.8 and 6.9, and that's how it was for
most all of 7.0. I'd rather cooperate though, but for that to work you need
to let me do my job.
- David Nusinow
Reply to: