[LLVMdev] RFC: Code Ownership

Jan Sjodin jan_sjodin at yahoo.com
Mon Nov 12 09:22:08 PST 2012

> More power to volunteers!  Hopefully if people talk

>with each other, help each other out, and maintain a friendly and cooperative
>atmosphere then all problems with code ownership will just come out in the wash.

+1 I think this is where we want the community to be, and the role of the code owners should
hopefully be minimal.

>However another take on the whole code owner thing is that the code owner is
>the person who has the final say on what goes in and what doesn't, so it is
>important for the code owner to be the über guru (and having other people
>override their decisions would just undermine their authority and the whole
>code owner system).  Personally I don't really buy this, but I can understand
>the concern.

I don't think this is the role of the code owner. The way I understand the developer policy
is that the code owner is there to ensure that all patches are reviewed. I would go further
and say that the code owner is a fail-safe, which means if the code review falls back on the code
owner, the regular review process has failed, and the code owner will have to take some measures
(any suggestions?) to make people review patches in a timely manner. It will be too easy for people
to stop reviewing, since they now think there is a code owner, which will do all code reviews, and the
result might be that person will sooner or later be overwhelmed and drop out. 

  It would perhaps be better to have the code owner be someone that has the knowledge, and cares
about a specific component in the context of quality, capability, design and documentation. But it should
probably not be the person who is the most frequent contributor to that component (might be difficult to find
such a person), since since the reviews may be likely to fall back on the same person, and he/she will have
to keep bugging other people about their own patches, or more likely, patches will go in without review if
the code owner has a review-after-commit status.

- Jan

More information about the llvm-dev mailing list