Gerrit code review velocity
ed.tanous at intel.com
Thu Nov 16 08:50:58 AEDT 2017
Is there anywhere that the SLAs and formats for how community code reviews are conducted?
I'm having some concerns around the code merge velocity on both sides of the spectrum, and I thought I'd see what everyone's thoughts were on the matter.
A great example of these concerns is below. First and foremost, I'm not picking on this specific committer here, this seems to be a greater issue.
This patchset was pushed last night, approved by a single developer, and merged by the submitter <24 hours later. Considering this is a "phosphor" reference component, changes an architecture level change (supporting multiple HWmon clients) and doesn't seem to have any supporting unit tests, I'm concerned at the velocity that it was moved through the progression, and into master.
Considering that we had some discussions on this specific change before, I would've liked to have input, and I'm sure others in the community would've as well, but were not given the opportunity. Do we have a process for how long code review->test->merge should take, and what process it should follow?
On the other end of the spectrum, I see code commits where the last activity was several months ago, and do not see anyone picking it up, nor any forward progress on merging. Is there a process defined for abandoning commits?
If the answer to all of these questions is, "no" or "nothing written down" I'm happy to put down some of my thoughts on integration rules down in a spec-like form, and see if we can come up with something that everyone is happy with.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the openbmc