Status and outlook for weekly I-build
and this final M6 week ...
Thanks to the ever vigilant Carl, and
the never resting Tran :) I'd like to respin one more "I-build"
today, before promoting it for others to build upon ... the reason being
that a late-discovered version intolerance problem for one of our prereqs
from DTP (bug 304892). I'm not sure how long its been there, but
this final week everyone will be trying to "build on all the latest"
and they will be blocked from doing so if we didn't fix this. That's
worse than merely having a bad bug in the declared build ... as we know
we do ... which we can still fix before the milestone build, without impacting
others forward progress (well, at least, not as much impact).
I actually did set aside our last I-build
candidate, I-3.2.0-20100305211931, in case something goes wrong with this
past last minute re-spin, or in case anyone objects to this plan. And
I did confirm that the expected jpt change was only change in the map files.
Those teams that had some blockers to
verify these weekend, can still use I-3.2.0-20100305211931 and change
your red-X's to green once confirmed.
In the mean time, for others that had
fixes or final things for the milestone build, feel free to release at
will. It will be in a subsequent build (by Sunday). We may end up with
even other, informal "declared builds" before our M6 build, as
we adjust our other prereqs. I've forgotten how "hot and heavy"
M6 is! .. let's just be sure to land on our feet by Thursday, for our Friday
M6 deliverable. This means you should be careful as always during this
"rampdown" week ... and you should do a little testing each day,
with the "build of the day" to make sure we stay whole.
To be explicit, I know other prereqs
have changed (beside DTP) but in my development environment, none of these
appeared to cause us any version range problems so I thought I'd update
our other prereqs after this last candidate I-build (to change as little
as possible for it).
I know there is a little risk in this
plan (who knows .... DTP changes could completely break JPT) but it seemed
a favorable trade-off with the known problem of no one being able to build
on our final I-build before the milestone.
Let me know if anyone disagrees or has
other suggestions or sees other problems.
If something goes wrong, or if anyone
objects, we could always declare I-3.2.0-20100305211931 and promote
something with even more fixes on Monday or Tuesday, so be sure to say
if anyone wants another plan.