[EMAIL PROTECTED]">I'm not sure where the breakdown is here. if you're checking into the tree, you're looking at tbox (WAP/vxml tbox access isn't enough), and at the top of tbox is the text:
"Checkin comment should include the reviewer (e.g. [EMAIL PROTECTED]), the super-reviewer (e.g. [EMAIL PROTECTED]), a clear explanation of the fix , and the bug number. Reviewers and approvers are considered to be 'on the hook'! "
A couple comments:
Looking at the top of tinderbox, I see "All Mozilla contributors: blah blah blah blah blah. Add new files to the Mac/Win/Unix packages! Netscape employees: blah blah blah blah blah." That is, the text is there, but it's not presented in such a way that it will be read. It's very hard not to gloss over. Of course, I hardly see that as an excuse for not following the rules, but it's something to think about. I'll just say that presenting information in such a way that you're certain it's been both read, understood, and remembered, is hard.
You're also making an assumption. It's plain to see that some poor souls aren't even looking at tinderbox when they check in, let alone waiting to see that their checkins have cleared.
Perhaps a part of getting CVS access should be a test of the rules.
The rules aren't too hard to find, but the point is that the contributor-to-be
has found them at least once and is using part of their brain to process
and transcribe them into the test...
[EMAIL PROTECTED]">I see two options: revoke certain folks' CVS access, and issue periodic knuckle-rappings. I don't see reviewing checkin comments as a viable option, since checkins may often be immediately necessary cleanups/back-outs, etc. Okay, so I don't really think we should revoke anybody's CVS access either (and yes, this is the part where I run and hide). Truthfully, I unfortunately agree with Jud, that the only thing to do is to attack the problem on a case-by-case basis...[EMAIL PROTECTED]">How are we to encourage good checkin comments?
Dan
[EMAIL PROTECTED]">
