On Thu, 2010-03-25 at 08:24 -0400, James Laska wrote:
> Just to clarify, the motivation for us (Bugzappers) starting
this thread
> is that it's our belief maintainers aren't really using the Target
> trackers for this purpose lately. We haven't seen anything to indicate
> those bugs are receiving any more attention than any other bugs.
Unintentional thread hijacking :) But seriously, for my own
understanding, my responses are focused on better understanding how
Target is used to make more informed suggestions on how we might improve
it's use in light of revised release criteria and critpath packages.
Sure, that's an alternative.
My understanding of the Target and Blocker process is they are a
mechanism to draw attention to what Fedora has decided [1] are the
important issues.
Well. We have a process for blocker bugs, and they are used for this
purpose. We really don't have a process for Target, besides haphazardly
dropping proposed Blocker bugs to it. We don't review Target bugs, we
don't poke anyone to work on them, we don't summarize them anywhere.
There isn't really a 'Target process'.
In order to understand if there is a place for Target
bugs, I think we need to answer some questions:
1. What is the scope of bugs we want to consider? Critpath or all
components?
2. How restrictive do we want to be when it comes to the scope and
milestones (Alpha, Beta and Final)? Do we want ...
A. fixes *only* for reviewed Blocker bugs?
B. fixes *only* for reviewed Blocker bugs and reviewed
'nice to have' bugs?
C. fixes for reviewed Blocker bugs and anything else
maintainers feel are necessary
D. fixes for anything a maintainer feels is necessary
Have we answered these questions?
I'm not quite sure they're tightly enough phrased. The key question is,
'when?' When are you talking about? Certainly, so far, throughout Alpha
and Beta, we're closer to D than anything else. Maintainers still can
push (and are pushing) any type of build they like into F13, up to now -
late Beta stage. They require a +1 from a proventester for critpath
packages, but none of the proventesters have been -1ing updates because
they're not 'the kinds of fixes we want', only if testing actually shows
them to be broken.
Closer to Final, I expect we'll progress to A, as we have in the past. I
guess the question here is whether we want to have a time-defined B
stage, and whether we can use a tracker bug to help with that.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net