Blake Ross wrote:
> Whoever has a patch for a bug is typically the owner of the bug, or should
> be.
What's the problem with the bug being assigned to the module owner (assuming
that the patch is waiting for review)? If the contributor does not have CVS
access, the module owner will have to check it in anyway.
> It is really the person with the patch's responsibility to push for
> review.
Then it's another thing that makes it harder for people to contribute. Not
every contributor is going to be a full-time mozilla hacker, and it's usually
hard enough to get the source, compile it, fix a bug, and put the patch in
bugzilla. You can do all this in one day. If you're required to push for
review, it becomes a task that lasts for several days, if not weeks. Of course
every patch author should respond to comments from the reviewer and update the
patch until it's acceptable. But it is really not at all obvious that a
contributor has to mail some people directly for review, and repeat until you
get an answer. In every other situation we tell people to operate within the
system, and the system is Bugzilla. In this case, if you file a bug in the
correct component, the owner should be an appropriate person to do the review.
If you then add the patch & review keywords, it should be a clear enough
indication that you want the owner to look at the patch. Is it really so much
better if you add additional "please review" comments? Or is emailing the owner
necessary because they usually don't pay attention to bugmail (aka spam)?
> I don't think it's fair to post a patch and wait for someone to do something
> with it.
Then this should be published more explicitly than it is today. At least the
description of the "patch" keyword and the "Attachment" section on
http://www.mozilla.org/bugs should contain a pointer to the "must-read for
contributors" document. There could be more places.
> http://mozilla.org/hacking/reviewers.html clearly states the process for
> getting review, and it includes e-mailing the appropriate people and cc'ing
> [EMAIL PROTECTED] Reading through these documents are really
> considered prerequisites for creating patches and/or getting CVS accounts.
That's mainly about the process for getting super-review, and a quick look at
recent posts in the .reviewers newsgroup will reveal that there are even
netscape employees who do not follow all the points mentioned there (e.g. the
recommendation about the subject lines). How can you require first-time
contributors to go through all this?
Anyway, super-review is the less critical part. The thing is that bug owners
should either review patches that appear in their bugs or point contributors to
more appropriate persons and helpful documents.
Andreas
