non-MOTU Hopeful contributions

Emmet Hikory emmet.hikory at gmail.com
Fri Oct 19 10:02:12 BST 2007


On 10/19/07, Neil Wilson <neil at aldur.co.uk> wrote:
> The process isn't at all complicated, but it is difficult. Any process
> is useless unless there is the manpower in the target review team to
> execute it. That is where the bottleneck lies.

    Do you really see a bottleneck here?  I can't speak for main, but
I believe we've been doing a pretty good job of keeping the universe
sponsors queue down to 10-20 items awaiting feedback at any given
time, with an average response time for each issue of no more than a
day or so.  Further during more active times, I've seen the queue
being processed every two hours or less, which doesn't indicate a lack
of sponsors to me.

> If I was a MOTU I wouldn't want to spend my entire day dealing with an
> essentially clerical process. I would have upload privileges and
> things I wanted to get done of my own. And I can certainly understand
> it if other MOTUs feel that way. However as the process stands there
> is no-one to delegate to, so MOTUs have to do it all.

    Actually, there are some MOTUs who do like to spend significant
time each day reviewing packages, and helping people get the patches
working.  In some cases, it's an easy way to meet personal QA goals,
as reviewing a patch is often simpler than fixing something in the
first place; in other cases, the MOTUs derive satisfaction from
providing guidance and or review to others; and sometimes, when
another issue is being addressed, fixes from the sponsor queue are
merged in the new upload.

> The problem is not the process, it is the level of people you are
> asking to operate the process and the general lack of manpower. To get
> around that you have to reduce the work they are asked to do, or get
> people other than MOTUs to review packages.
>
> For example is there any reason why MOTU wannabes couldn't do the
> package reviews and +1 them when they are happy? Then MOTUs only have
> to look at packages that have already passed through an acolyte. That
> helps train the acolytes and reduces load on MOTUs.

    There's absolutely no reason why non-MOTUs couldn't look through
the patch queue, test things, and advocate them if they are happy.
This certainly increases the chances that the MOTU will feel
comfortable uploading something which otherwise may be undecided
(especially in cases where an issue relies on special hardware or
configuration).  I don't see any value to restricting sponsoring to
things that have previously been determined as good by another
reviewer: this seems to require that there is another layer of active
reviewers, and may reduce our ability to keep the queue small.
Rather, I would invite those interested in review (and not MOTU) to
take a look at candidates: the majority are in good shape, but some
may benefit from constructive suggestions for improvement.

-- 
Emmet HIKORY



More information about the Ubuntu-motu mailing list