Draft procedure for processing the Universe Sponsor Queue

[Emmet Hikory]

Le samedi 26 mai, Adrien Cunin a écrit :
> You say self-assigning when beginning a review, but I guess the sponsor
> will stay assigned once the bug is closed. See below why I don't agree
> with that. IMO adding a comment is enough, if you think you will need
> some time to review.

   With regard to the use of a comment rather than the Status tag, I
firmly believe that the use of the Status for the bug is preferable to
a comment, as the standard strings can easily be translated, ensuring
that the procedure will not need modification to support a
native-language interface for LaunchPad.  Please see below for
discussion of assignment.

> So far, when sponsoring, I've self-assigned the bug when it's a bug
> requesting sponsoring, for example the merge bugs.
> When it's an actual bug that have been picked by a contributor, who
> attached a debdiff, I let him assigned. Why? because the real person who
> fixed the bug is not me, I just check the sanity of the debdiff and
> upload. Letting people assigned to the bugs also makes easier for them
> to keep track of what bugs they have been working on (useful for eg.
> MOTU application :)). We keep the contributor in the Changed-By field
> when uploading, so why not keep him as bug assignee?
> Btw, but it's probably off-topic, I wonder why sync requests' Changed-By
> field is set to the developer who ACKed the sync, and not to the
> contributor who requested it.

    I understand this argument, and don't necessarily disagree.  On
the other hand, the published guidelines currently encourage
contributors to assign "Nobody" when requesting a review.  Perhaps a
solution would be for the sponsor to assign the contributor when
sponsoring?  My only worry here is that it may require more effort on
the part of the sponsor to ensure the bugs are properly closed
(especially for new contributors).

    Separately, and anecdotally, of all of my revision candidates that
were sponsored prior to joining MOTU, I do not remember a single case
where I was encouraged to remain the assignee for the bug, and in at
least two cases, I was specifically reprimanded for resetting the
assignment to myself while chasing build failures for a sponsored
upload.  As the intent of the documentation is to formalise best
practice, I would encourage feedback from other sponsors as to which
is the appropriate assignment model.

-- 
Emmet