Good catch on
the project page. We do not really maintain that - we assume
that everyone starts their research of Che on our github
repository, so we spend most of our time checking those links.
I've updated the project page to have a good link to the wiki.
We hired a new
PM about 2 months ago - right at the time that the minutes
stopped appearing. The new PM is our release manager, and in
the transition of duties, it looks like weekly che meeting
minutes were not being captured. But we hvae the meeting every
week, and people from all across join. Brad - see below - we
are missing meeting minutes for the past 6 meetings.
The process can
be elaborated upon in the wiki. We only accept PRs that are
merge ready. But we still a maintainer to accept it into their
sprint. We have many PRs - including many that we have
drafted - that haven't gotten taken into the sprint. There
have been a number of commitments that we have had to make to
the market, and there were deliverables up through September
that took precedence. The maintainers that are responsible for
project types & the IDE were asked to prioritize getting
the language server protocol completed before considering the
PHP stuff.
If you want to
send some ire or frustration, you can send it to me, as I
personally set the priorities for the TLs within each sprint.
But we do document the intended milestone plan (sprint plan)
in advance with the sprint/next-sprint label, and we also keep
a complete list of the expected work items for a particular
group. The gropu that you care about is the team/ide group.
But I gave them priorities related to some long time bug fixes
and language server protocol before considering the PHP
items. I added these PHP items into the epic tracking for
phase 1 language server.
We have
accepted about 100 PRs from the community this year. We
usually prioritize them near the top of the stack, unless we
believe there is QA or other infrastructure epics that need to
be completed first before we consider the PR. For example,
there is a long running pair programming plugin PR and a long
running test runner (Junit) support PR that has been in our
backlog. But we cannot merge both of these until we had fixed
other things related for virtual file systems, maven bugs, and
the so forth. We discuss these items in the weekly meetings
and in the milestone plans.
Any time a
developer gets two major PRs merged in and they are
participating in the docs / support forums, we nominate them
for committer status. We have done that with people from SAP
& Red Hat this year. Just 2 days ago, Mario Loriedo from
RHT got nominated for committer. He is now the technical lead
for a team of 5 people from Red Hat who are working on Che. I
hope that all 5 of them are committers before the end of the
year.
But this is a
very complicated system. And we do not just merge PRs from the
community. We do want a TL to take it into the sprint, add it
into our integration tests, verify docs, communicate with the
PR submitter, and verify that we have a long term plan for
support. So even small quick fixes like the one liner - yes
those are easy. But I told the engineers to package that up
with the PHP PT and the language server because they were
co-related and should be released & shipped together.