Robert Godfrey wrote:
It sounds like consensus is building around commit-then-review with
Jira, so let's go for that. I'll start a formal vote later today.
One thing I would like to see is some discussion about whether we will
all be following a common process or whether the components in each
language will be doing their own thing.
The process and meetings up until now have focused mainly on the Java
side of things; and we have been quite keen on enforcing things like
"no commit without an accompanying JIRA"...
I think if we want to have a formal commit-then-review process then
linking each commit to a JIRA will be necessary so that we can
document the review.
I already informally review quite a lot of commits. I focus more on
those that touch areas of the code I am familiar with and the depth of
review tends to vary with the amount of time I have and the nature of
the revision.
When I make changes that I feel may particularly warrant review for
whatever reason, I will often send out a patch before committing it
either to the list or to someone who I know to be familiar with the area
in question.
I fully accept that the code doesn't get as much review as we would
like, and areas that are primarily worked on by a single individual tend
to be worse in that regard. However, personally I don't feel that a
formal process is needed.
I'm very willing to have anyone point out concerns they may have with
any of my commits or even just request more detailed explanations. I
feel entitled to do the same for any commits that concern me.
I'm willing to review specific patches or revisions on request (subject
to time constraints) and I will continue to request comments or review
for specific changes I make that I feel merit the extra scrutiny.
