Thoughts on code reviews.

Marcus Hirt marcus.hirt at oracle.com
Thu Sep 6 22:45:39 UTC 2018


Hi all,

The JMC project does not formally require code reviews, in the OpenJDK bylaws
sense of the term. That said, we still have the notion of reviewers, and each 
check-in must have been reviewed by someone else. We only have one active 
reviewer on the project today (me), and this has proven quite limiting.

I am suggesting that we do three things:

1. We appoint Guru reviewer. Especially for all things infrastructure and 
   release engineering, a review by Guru is enough to push.
   
2. We appoint Miro author (has two sponsored commits).

3. Two authors together (with experience in the area of the code) can together 
   perform a review. At least for the foreseeable future. (Continuing the 
   summer's temporary policy.)

What do you think?

Kind regards,
Marcus




More information about the jmc-dev mailing list