Daniel John Debrunner wrote:
Rick Hillegas wrote:
I would like to build a release candidate soon. I would like to do this
even though we know that the 10.2 optimizer changes are risky. It has
been pointed out that that risk plagues all optimizer modifications and
that we must balance that risk against the substantial performance boost
provided by these particular changes. I would be willing to delay the
release if I believed we had a better plan for encouraging users to
test-drive our distributions. Unfortunately, to me the proposed plan for
soliciting more participation doesn't seem that different from the
previous, ineffective plans. I therefore think it is unlikely that three
more weeks will polish 10.2 much. I understand that this decision will
disappoint people in the front-lines of tech support--they are the
unsung heroes of any engineering effort, the people who bear the brunt
of our mistakes.
I would like to build a 10.2 release candidate once we remove the
following obstacles:
1) We need to review and commit patches attached to the two remaining
urgent JIRAs: DERBY-1777 and DERBY-1522.
2) We need to compile Release Notes--I am close to publishing a first
draft.
3) We need to clip the JDBC4 disclaimer to the 10.2 user guides.
Please let me know if I have left something out. I hope that we can wrap
up this work and post a release candidate by early next week. Thanks
again for everyone's hard work and good humor.
Are you planning another "mega-merge"? Seems DERBY-1522 would be covered
by that.
Dan.
Thanks for pointing this out. Yes, I am planning a mega-merge. So we're
down to one unreviewed/uncommitted patch: DERBY-1777.
Regards,
-Rick
