Am 2022-11-15 um 14:32 schrieb Oleg Kalnichevski:
We have an implicit commit-them-review policy ever since the inception
of the project in the year of 2005. We all are free to commit what we
deem appropriate but no commit can be considered safe until it has been
voted upon and tagged with a release tag.
If an objection has been raised about any commit in the development
branch it can be reverted and should be resubmitted through a PR and put
to a review. This basically applies to all committers and any commit.
Any committer can ask for a commit to be reverted and the change-set put
to a review.
If any commiter no longer feels comfortable with this approach I
personally will have no problem switching to the pre-commit review
policy whereby every change must go through a PR first. Naturally that
would call for a format vote.
You guys might want to read dev@maven.a.o. We had this discussion
recently. Coincedence.
After the Maven 3.7.0 fiasco we have moved from CTR to RTC which -- yes,
it slows down the process -- *but* drastically improved the quality of
the code/new changes /before/ they land and reverts are basically down
to zero. Additionally, people tend to do better self reviews. I did a
lot of self reviews on recently Doxia changes and noticed myself that my
PRs were incomplete. I would never go back to CTR, personally. (except
for typo fixes or alike)
Of course, CTR means that you have qualified people who can review in a
decent timeframe (needs to be defined). It is never wrong to have
another pair of eyes on a solution.
Michael
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@hc.apache.org
For additional commands, e-mail: dev-h...@hc.apache.org
