Patch Pilot report

[Martin Pool]

What I'd actually most like to know now is how this looked to people
putting code up for review.  Was it welcoming?  Better than before?

It's my Friday evening, and I'm winding up as patch pilot.  I'm not
going to specifically _ignore_ reviews next week, but I don't want
people to think I have an explicit lock on those I did touch.

I think I had reasonably good focus on patches, especially towards the
start of the week.  It certainly wasn't just me: everybody was active
in mentoring and shepherding patches and John may have done more than
I did.  I think having one designated pilot still helps though, to
keep things moving.

There was some confusion about whether people should click the
'resubmit' button when they redo their patch or not.  'Resubmit'
essentially wipes the slate clean.  I think generally speaking it's
better just to push a new update.

I filed a few bugs against launchpad-code
<https://bugs.edge.launchpad.net/launchpad-code/+bugs> for things I
encountered in doing this.

Launchpad was in readonly mode for a while because of an upgrade that
took longer than expected.  This turned out not to be _too_ much of a
problem because the diffs were in mail, but it does show how much one
comes to rely on this.

Vincent will be taking over in utc+1 Friday or Monday.  He raises the
question of merging doc fixes with a single review, which I think
would be fine, though at our current level of activity he's likely to
get a second review within 24h anyhow.

So we now have

  6 reviewed and ready to land (up from four)
  12 waiting for review or under review (down from 27!, yay)

which is getting to a more reasonable number: one screenful on a large
monitor.  The oldest is six months though, which is not so good.  We
merged about 30 patches this week.   I didn't make large changes or
write large tests for any patch.

-- 
Martin <http://launchpad.net/~mbp/>