Alan Burlison wrote:
Mike Kupfer wrote:
It's been suggested to me that we instead use the email address that's
registered in Auth for the person doing the push. I like that
suggestion for a couple reasons. One, it gets rid of a chunk of
complexity. Two, it associates the email with the person who did the
push, rather than the changeset author, which seems more appropriate to
me.
Am I missing anything? Are there compelling reasons to continue using
the changeset author for the From address?
The changeset author needs to either be a Sun employee, or have signed a
SCA. Any commits that don't satisfy that constraint need to be blocked.
And the person doing the push should satisfy the same constraints, if
it isn't the same person that authored the changeset.
And the pusher/changeset author distinction is only really relevant for
ON, and that's a gate that we mirror rather than manage, unlike the
other 450 that we host.
No, it can also be true of other gates as well.
For example, the pkg(5) gate requires new committers to have their
changesets pushed by a team member for the first few changesets. Once
approved, they are given commit access.
However, when the committs are done, they show up as having been done by
that new person and the notifications are sent as being from that new
person.
You'd have to ask Danek what he does to handle this case right now.
Cheers,
--
Shawn Walker
_______________________________________________
tools-discuss mailing list
[email protected]
