Hi folks,

Jan tells me that he has a number of changes he's checked into the
STABLE branch that make some significant changes, in order to deal
with both the 8.1 issues and a number of subscription performance
issues.

I understand the importance of this, but I'm starting to (!) get
extremely nervous about how "stable" releases are getting released.

So, I want to ask the group for either consensus or not: should we be
making serious changes, without significant release candidates &c.,
to the so-called stable branch.  Nobody would suggest that the 1.1
release of Slony has been as robust as the 1.0 release, I think. 
That said, do we want to continue making significant changes in this
branch, or just admit that this one didn't work as planned, and work
on HEAD and getting 1.2 out the door (and recommend to people that
maybe 1.1 isn't their best bet)?

I'm extremely worried about release practices these days.  I think
we've done a not terribly good job recently, and I'd like for us to
discuss that in a separate thread.  This is merely a question of
whether we ought to back these changes out of the 1.1 branch now, and
just admit we made a mistake; or not.  It's a community decision, but
I'd like it to _be_ such a decision.  Note that Jan is in fact
testing this weekend, on the principle that it's already in there
anyway.  But I'm nervous of proceeding without at least a prima facie
indication that the community is comfortable with this direction.

A

-- 
Andrew Sullivan  | ajs at crankycanuck.ca
Information security isn't a technological problem.  It's an economics
problem.
		--Bruce Schneier