On Tue, May 31, 2011 at 10:35 AM, Neil Kandalgaonkar
<neilk(a)wikimedia.org> wrote:
> Are we all in deadlock or something? Are the users who can push waiting
> from some proposals/work from the rest of the community?
We had a hallway conversation about this just now (Neil, Trevor, Brion
and I, and then just Brion and I), which I think was pretty useful.
Here's where we went with it:
1. We rehashed the pre-commit review proposal that Neil suggested a
few months ago, and agreed that pre-commit would be helpful in keeping
the backlog down
2. Given our current tools/process, we agreed that insisting on
pre-commit review would be a pain in the butt.
3. Brion and I further discussed review process, trying to come up
with a system that give us the benefits of pre-commit review, without
actually switching to pre-commit review
Here's where things I think got interesting. Brion pointed out that
in ye olden days, he was much more aggressive about reverting things
he didn't understand. I pointed out that, as we broaden the pool of
committers, "I don't understand"-based reversions lead to a lot of
ugliness, since very few people claim to have a broad understanding of
the system and therefore an expectation of understanding every change.
Most reviewers, faced with a commit they don't understand, will leave
it for others to comment on. There's been a lot of unnecessary drama
and churn over reversions because of misunderstandings about what a
reversion means.
So, there's a number of possible solutions to this problem. These are
independent suggestions, but any of these might help:
1. We say that a commit has some fixed window (e.g. 72 hours) to get
reviewed, or else it is subject to automatic reversion. This will
motivate committers to make sure they have a reviewer lined up, and
make it clear that, if their code gets reverted, it's nothing
personal...it's just our process.
2. We encourage committers to identify who will be reviewing their
code as part of their commit comment. That way, we have an identified
person who has license to revert if they don't understand the code.
I coulda swore there are other ideas that came out of that
conversation, but alas, I wasn't taking notes. Anyway, I'm sure
they'll come up in this thread.
Rob