On Fri, Dec 04, 2015 at 07:25:41AM +0000, James Hogarth wrote:
So with my little one keeping me up all night I figured I could at
least be
a little productive.
It turns out we have just under 1000 bugs against package review in the NEW
state.
Now if even just half of these don't have requesters that respond that's
still 500 that according to stalled review policy should block
FE-DEADREVIEW (and realistically it's more than that, especially once you
start including stuff stalled that was actually ASSIGNED too)...
The question on my mind is how useful is that when you reach that number of
things blocking it? Who in their right mind is going to check one of the
thousand or so packages marked blocking that as a place to start packaging?
That's not even pondering what the bugzilla limits for this likely are
before it goes poof...
Real dead reviews are not the problem: the list is easily searchable and
anyone who tries to pick up packaging of some piece of software can
check the list and recycle previous work and review efforts.
The real problem is reviews where the packager still wants to proceed,
but progress is blocked by sponsorship requirements or lack of review.
But only a subset of this 1000 bugs on the list is in this state.
Unfortunately it is not obvious which ones.
We should be more aggresive in reassigning bugs to FE-DEADREVIEW
to make the "live" ones stick out more.
Zbyszek