SRU policy for universe (was Re: Pending SRU Request; What to do?)

[Stefan Potyra]

Hi Martin,

Am Montag, 19. November 2007 12:36 schrieb Martin Pitt:
[..]
> > <pitti> but I am not going to take responsibility for this
>
> I might be overly cautious about it, indeed. It's because I review all
> the SRUs for main, where we have very strict standards and policy of
> what to accept. This naturally reflects to my judging of universe as
> well.
>
[..]
> If that is true, then the SRU is appropriate.
>
> > (3) pitti is referencing asking the *former* MOTU SRU team.
> > Referencing the SRU page on wiki.ubuntu.com [2]
> > [...]
>
> Right, it's not official policy any more, but since this is not a
> 'standard' SRU, I wanted to get some more opinions.
>
> >     * Any MOTU can decide that particular issue is good candidate for
> >     SRU and can prepare the update package for <release>-proposed
>
> I object to this bit of the current MOTU SRU policy since it toally
> lacks peer review and acknowledgement. I raised this to the TB a week
> ago, but did not get an answer yet. But redesigning the MOTU SRU
> policy is a separate discussion, of course.
>
> > Does this not mean the responsibility falls directly on the MOTU?  I
> > don't see how pitti would be taking responsibility for said problem.
>
> If my role for MOTU SRU is merely an archive monkey, then I wouldn't.
> But since we currently do not have any peer review in MOTU SRU I
> review the universe SRUs anyway to check them for sanity and errors,
> and thus I think I do have some implicit responsibility.
>
> > Also why are archive admins stepping up and doing so much
> > administrative work in reviewing all of the changes so in depth in
> > the first place?  My impression (especially after that meeting) was
> > that an Archive Admin had a much more clerical role instead.
>
> I said that with my ubuntu-sru hat on, not with ubuntu-archive. In
> principle you are right, we shouldn't need to do any in-depth review,
> but *someone* should. Changing stable releases without *any* peer
> review feels very wrong to me. At the moment I just (ab?)used my
> archive admin hat to get a more in-depth discussion about non-obvious
> SRUs like this one.

Well, the current SRU policy for universe was to have a very low entry barrier 
for -proposed (aiming that archive admins should basically only check the 
sanity of the version, due to a LP bug not being able to delete packages from 
the -proposed queues) and to have testing being performed as the main 
reviewing measure. This means that no peer review for actual updates would 
happen, unless archive admins would perform additional checks prior to the 
migration from -proposed to -updates.

The change was somewhat drastic from very picky reviews by motu-sru to no 
reviews for universe SRUs at all. The goal of this change was to encourage 
more SRUs being done (as the strict policy we had before seemed to have 
scared MOTUs away) and to improve the throughput of SRUs.

While I still believe, that the entry barrier to -proposed should be quite 
low, there are counter-arguments that people install packages from -proposed 
as if it were -updates. I guess universe is tied here quite to main, since 
the days it got enabled by default.

Right now, I'm not convinced that the new SRU policy for universe doesn't have 
any caveats, nor does it really target the goals, as I don't know anything 
about actual throughput figures. If you give me a few days, I'll prepare 
those numbers. Maybe that can help to underline or disprove my suspicion that 
the policy change didn't really move us closer to the goal of getting more 
updates out.

One major problem with the current SRU policy for universe that I see right 
now is, that we kind of lost track of what's going on in proposed/updates, 
leaving some packages rotting in the proposed queue for a very long time 
(even one SRU from me, shame on me *g*). Also we have no indication about the 
quality of updates, maybe you could shed some light here?

Cheers and thanks for bringing this topic up again,
    Stefan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : https://lists.ubuntu.com/archives/ubuntu-devel/attachments/20071119/3eb18c24/attachment.pgp