You are viewing a plain text version of this content. The canonical link for it is here.
Posted to derby-dev@db.apache.org by Dy...@Sun.COM on 2008/04/11 16:21:15 UTC
[VOTE CANCELLED] 10.4.1.1 release candidate
The fact that a fix appears to be available for
https://issues.apache.org/jira/browse/DERBY-3347
(and maybe DERBY-3611)
which is a serious data corruption, AND that the investigation has shown
that this is a regression in 10.3 (caused by DERBY-801), has convinced
me that the right thing to do is to spin another RC.
I plan to do this as soon as the fix for DERBY-3347 has been merged to
the 10.4 branch, and then call for a vote on the new RC.
Apologies to those who have spent time testing and voting on the
current release candidate. :(
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Kathey Marsden <km...@sbcglobal.net> writes:
> Dyre.Tjeldvoll@Sun.COM wrote:
>> I plan to do this as soon as the fix for DERBY-3347 has been merged to
>> the 10.4 branch, and then call for a vote on the new RC.
>>
>>
> It would be great also to see the alternate fix for DERBY-2892 go in
> and then it will eliminate the necessity for the release note and
> incompatibilities in the 10.4 release.
Thanks for bringing this up. These are the chekins that have been made
to 10.4 after 10.4.1.1 was spun:
------------------------------------------------------------------------
r645438 | kristwaa | 2008-04-07 12:03:01 +0200 (Mon, 07 Apr 2008) | 4 lines
DERBY-3571: LOB locators are not released if the LOB columns are not accessed by the client.
Merge from trunk revision 643819 (derby-3571-2a-simple_release.diff).
The patch added the client side LOB state tracking, used to release LOB locators in a timely manner.
------------------------------------------------------------------------
r645706 | mamta | 2008-04-07 23:45:48 +0200 (Mon, 07 Apr 2008) | 10 lines
migrating revision 645665 from trunk into 10.4 codeline (DERBY-3320).
The commit comments were as follows
Fix for the test that was added for DERBY-3320. The test was using DriverManager.getConnection
but DriverManager is not available in JSR 169 environment. This caused the test to fail
when run in JSR 169 environment. I have fixed the test to use testUtil.getConnection
which is smart enough to use DriverManager or DataSource depending on what VM the test
is getting run in.
------------------------------------------------------------------------
> Kristian said:
>
> Committed to trunk with revsion 646255.
> I plan to backport the fix to 10.4 when the last patch has been
> committed, and also investigate the possibility of backporting to
> 10.3.
> [ Show » <https://issues.apache.org/jira/browse/DERBY-2892> ]
> Kristian Waagan
> <https://issues.apache.org/jira/secure/ViewProfile.jspa?name=kristwaa>
> -
> 09/Apr/08 03:21 AM 'derby-2892-1b-alternative_fix_partial.diff' is
> slightly changed from 1a; reordered keywords (abstract public ->
> public abstract) and removed an added blank line. Committed to trunk
> with revsion 646255. I plan to backport the fix to 10.4 when the last
> patch has been committed, and also investigate the possibility of
> backporting to 10.3.
>
>
> I'm not sure of the timeframe. I'd be happy to help in anyway I can.
I believe Kristian is at ApacheCon Europe, but I'm willing to
merge these fixes to 10.4.
What about the last patch 'derby-2892-2a-alternative_fix_cleanup.diff'?
Should that be committed to trunk and merged as well, or is that less
urgent?
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Kathey Marsden <km...@sbcglobal.net> writes:
> Dyre.Tjeldvoll@Sun.COM wrote:
>> I plan to do this as soon as the fix for DERBY-3347 has been merged to
>> the 10.4 branch, and then call for a vote on the new RC.
>>
>>
> It would be great also to see the alternate fix for DERBY-2892 go in
> and then it will eliminate the necessity for the release note and
> incompatibilities in the 10.4 release.
>
> Kristian said:
>
> Committed to trunk with revsion 646255.
I've merged this to 10.4
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Myrna van Lunteren <m....@gmail.com> writes:
> Also, I'd like to see if 1 test bug can be backported that I ran into
> when I ran with weme6.1 (i.e. using IBM's JSR169 implementation):
> DERBY-3548 (fixed in 10.5) I just may not have time to look at it
> further today.
I've merged it to 10.4.
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Myrna van Lunteren <m....@gmail.com>.
On 4/11/08, Martin Zaun <Ma...@sun.com> wrote:
>
> Myrna van Lunteren wrote:
> > Now that there is a plan for a new release candidate, I think maybe
> > the javadoc should be fixed?
> > There are the following errors:
> >
> > [javadoc]
> C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> > warning - @param argument "address" is not a parameter name.
> > [javadoc]
> C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> > warning - Tag @see: reference not found: NetworkServerControl
> >
> ...
> > [javadoc]
> C:\svn\10.4\java\engine\org\apache\derby\authentication\SystemPrincipal.java:47:
> > warning - Tag @see: can't find name in java.security.Principal
> >
>
> I noticed that at least these three javadoc issues have been fixed
> already in the trunk by Mike:
> 636875 mikem
>
> Guess it's best for a committer to propagate these javadoc fixes
> to the 10.4 branch.
>
> Martin
>
Thanks for finding those revisions, Martin!
I understand the pressure Dyre, I've been there...
I can spend some time on Saturday to look at backporting & test &
commit the javadoc fixes & I'll also backport & commit the test fix I
mentioned if it's minor.
Myrna
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Myrna van Lunteren <m....@gmail.com> writes:
>> ------------------------------------------------------------------------
>>
>> There are still 7 javadoc warnings left on the 10.4 branch AFAICT, but
>> none on trunk. Not quite sure what to do about those. I guess we could
>> fix the javadoc only on the branch, but that could lead to conficts if
>> the checkins that fixed them on trunk are later merged to the branch...
>>
>> --
>> dt
>>
> Thanks for doing those merges, Dyre! I hunted around and I think the
> final 7 were fixed on trunk by mike with revision 636114; I looked at
> the diffs and they looked innocent but fixed the javadoc. I just
> merged that to 10.4 without any problems and committed with revision
> 647502.
Ah, I missed that one. Thanks for doing this Myrna.
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Myrna van Lunteren <m....@gmail.com>.
On 4/12/08, Dyre.Tjeldvoll@sun.com <Dy...@sun.com> wrote:
> Martin Zaun <Ma...@Sun.COM> writes:
>
> > Myrna van Lunteren wrote:
> >> Now that there is a plan for a new release candidate, I think maybe
> >> the javadoc should be fixed?
> >> There are the following errors:
> >>
> >> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> >> warning - @param argument "address" is not a parameter name.
> >> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> >> warning - Tag @see: reference not found: NetworkServerControl
> > ...
> >> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\authentication\SystemPrincipal.java:47:
> >> warning - Tag @see: can't find name in java.security.Principal
> >
> > I noticed that at least these three javadoc issues have been fixed
> > already in the trunk by Mike:
> > 636875 mikem
> >
> > Guess it's best for a committer to propagate these javadoc fixes
> > to the 10.4 branch.
>
> I have merged this and two other javadoc fixes to 10.4:
>
> ------------------------------------------------------------------------
> r647468 | dyre | 2008-04-12 18:35:24 +0200 (Sat, 12 Apr 2008) | 4 lines
>
> Javadoc fixes (no jira).
> Merged to 10.4 with svn merge -r 636874:636875 ../derby-scratch
> No conflicts.
>
> ------------------------------------------------------------------------
> r647470 | dyre | 2008-04-12 18:50:16 +0200 (Sat, 12 Apr 2008) | 4 lines
>
> Javadoc fixes (no jira).
> Merged to 10.4 with svn merge -r 637199:637200 ../derby-scratch
> No conflicts.
>
> ------------------------------------------------------------------------
> r647472 | dyre | 2008-04-12 19:02:25 +0200 (Sat, 12 Apr 2008) | 4 lines
>
> Javadoc fixes (no jira).
> Merged to 10.4 with svn merge -r 639014:639015 ../derby-scratch
> No conflicts.
>
> ------------------------------------------------------------------------
>
> There are still 7 javadoc warnings left on the 10.4 branch AFAICT, but
> none on trunk. Not quite sure what to do about those. I guess we could
> fix the javadoc only on the branch, but that could lead to conficts if
> the checkins that fixed them on trunk are later merged to the branch...
>
> --
> dt
>
Thanks for doing those merges, Dyre! I hunted around and I think the
final 7 were fixed on trunk by mike with revision 636114; I looked at
the diffs and they looked innocent but fixed the javadoc. I just
merged that to 10.4 without any problems and committed with revision
647502.
Myrna
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Martin Zaun <Ma...@Sun.COM> writes:
> Myrna van Lunteren wrote:
>> Now that there is a plan for a new release candidate, I think maybe
>> the javadoc should be fixed?
>> There are the following errors:
>>
>> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
>> warning - @param argument "address" is not a parameter name.
>> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
>> warning - Tag @see: reference not found: NetworkServerControl
> ...
>> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\authentication\SystemPrincipal.java:47:
>> warning - Tag @see: can't find name in java.security.Principal
>
> I noticed that at least these three javadoc issues have been fixed
> already in the trunk by Mike:
> 636875 mikem
>
> Guess it's best for a committer to propagate these javadoc fixes
> to the 10.4 branch.
I have merged this and two other javadoc fixes to 10.4:
------------------------------------------------------------------------
r647468 | dyre | 2008-04-12 18:35:24 +0200 (Sat, 12 Apr 2008) | 4 lines
Javadoc fixes (no jira).
Merged to 10.4 with svn merge -r 636874:636875 ../derby-scratch
No conflicts.
------------------------------------------------------------------------
r647470 | dyre | 2008-04-12 18:50:16 +0200 (Sat, 12 Apr 2008) | 4 lines
Javadoc fixes (no jira).
Merged to 10.4 with svn merge -r 637199:637200 ../derby-scratch
No conflicts.
------------------------------------------------------------------------
r647472 | dyre | 2008-04-12 19:02:25 +0200 (Sat, 12 Apr 2008) | 4 lines
Javadoc fixes (no jira).
Merged to 10.4 with svn merge -r 639014:639015 ../derby-scratch
No conflicts.
------------------------------------------------------------------------
There are still 7 javadoc warnings left on the 10.4 branch AFAICT, but
none on trunk. Not quite sure what to do about those. I guess we could
fix the javadoc only on the branch, but that could lead to conficts if
the checkins that fixed them on trunk are later merged to the branch...
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Martin Zaun <Ma...@Sun.COM>.
Myrna van Lunteren wrote:
> Now that there is a plan for a new release candidate, I think maybe
> the javadoc should be fixed?
> There are the following errors:
>
> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> warning - @param argument "address" is not a parameter name.
> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
> warning - Tag @see: reference not found: NetworkServerControl
...
> [javadoc] C:\svn\10.4\java\engine\org\apache\derby\authentication\SystemPrincipal.java:47:
> warning - Tag @see: can't find name in java.security.Principal
I noticed that at least these three javadoc issues have been fixed
already in the trunk by Mike:
636875 mikem
Guess it's best for a committer to propagate these javadoc fixes
to the 10.4 branch.
Martin
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dyre Tjeldvoll <Dy...@Sun.COM>.
Myrna van Lunteren wrote:
> On 4/11/08, Dyre.Tjeldvoll@sun.com <Dy...@sun.com> wrote:
[snip]
>>
> I too had understood a Release Candidate to become a Release once approved.
>
> Now that there is a plan for a new release candidate, I think maybe
> the javadoc should be fixed?
> There are the following errors:
>
> Also, I'd like to see if 1 test bug can be backported that I ran into
> when I ran with weme6.1 (i.e. using IBM's JSR169 implementation):
> DERBY-3548 (fixed in 10.5) I just may not have time to look at it further today.
>
Ok, maybe I shouldn't be rushing this. Originally I was planning to spin
the new RC over the weekend, but maybe I rather should spend that time
merging stuff over and fixing the javadoc. Perhaps a new candidate can
be spun on Monday or something.
I do feel that there is a certain element of urgency here, for a number
of reasons:
- A new feature release is overdue (release early, release often)
- We have a number of new features that we want to put into user's hands
- Delaying long would be unfair to those who got their stuff pushed
out to 10.5 before the feature freeze
- For personal reasons, I cannot continue as release manager indefinitely
Thanks,
Dyre
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Myrna van Lunteren <m....@gmail.com>.
On 4/11/08, Dyre.Tjeldvoll@sun.com <Dy...@sun.com> wrote:
> Bryan Pendleton <bp...@amberpoint.com> writes:
>
> >>> I plan to do this as soon as the fix for DERBY-3347 has been merged to
> >>> the 10.4 branch, and then call for a vote on the new RC.
> >
> > I'm not sure I understand why the availability of additional fixes
> > required the cancellation of the previous vote.
>
> It probably doesn't. I just chose that subject because the 10.3 Feature
> Release Summary on http://wiki.apache.org/db-derby/DerbyTenThreeRelease,
> says "Sanity Testing and Voting (Cancelled - blocker DERBY-1963)".
>
> > It seems to me that there are always additional fixes on the horizon.
> >
> > I don't think we were voting on whether 10.4.1.1 was to be the
> > actual *release*;
>
> Ummm, I thought we were. Had it not been for DERBY-3347 I was going to
> put it on the mirrors and so on (provided it got PMC approval).
>
> > we were simply voting on whether it was an
> > acceptable release *candidate* for making available for advanced
> > users interested in helping us test the software prior to release.
>
> So you you considered 10.4.1.1 to be a BETA release candidate that, once
> approved through a vote would become a BETA release that users could try
> out? And that based on their feedback there would be a proper release
> candidate that, if approved, would turn into a proper release?
>
> That was not my plan. Sorry if I gave that impression. The current
> description of the release process only mentions one vote, and my
> understanding is that it determines if a release candidate can
> become a release.
>
> --
> dt
>
I too had understood a Release Candidate to become a Release once approved.
Now that there is a plan for a new release candidate, I think maybe
the javadoc should be fixed?
There are the following errors:
[javadoc] Loading source file
C:\svn\10.4\java\engine\org\apache\derby\iapi\error\SQLWarningFactory.java...
[javadoc] Loading source file
C:\svn\10.4\java\client\org\apache\derby\client\am\SqlWarning.java...
[javadoc] Loading source file
C:\svn\10.4\java\tools\org\apache\derby\impl\tools\ij\ijWarningResult.java...
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\error\SQLWarningFactory.java:46:
warning - Tag @link: reference not found: SQLState
org.apache.derby.shared.common.reference.SQLState
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\error\SQLWarningFactory.java:63:
warning - Tag @link: reference not found: SQLState
org.apache.derby.shared.common.reference.SQLState
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
warning - @param argument "address" is not a parameter name.
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\jdbc\DRDAServerStarter.java:107:
warning - Tag @see: reference not found: NetworkServerControl
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\store\access\sort\ExternalSortFactory.java:129:
warning - @returns is an unknown tag.
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\store\access\sort\UniqueWithDuplicateNullsExternalSortFactory.java:37:
warning - Tag @see: reference not found:
MethodFactory#primaryImplementationType
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\store\access\sort\UniqueWithDuplicateNullsExternalSortFactory.java:44:
warning - Tag @see: reference not found:
MethodFactory#supportsImplementation
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\store\replication\master\MasterController.java:290:
warning - Tag @see: reference not found:
org.apache.derby.iapi.services.replication.master.MasterFactory#startFailover()
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\authentication\SystemPrincipal.java:47:
warning - Tag @see: can't find name in java.security.Principal
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\iapi\sql\dictionary\IndexRowGenerator.java:275:
warning - Tag @see: can't find isisUniqueWithDuplicateNulls in
org.apache.derby.catalog.IndexDescriptor
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\sql\execute\CreateIndexConstantAction.java:140:
warning - @param argument "isUniqueWithDuplicateNulls" is not a
parameter name.
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\sql\execute\UniqueWithDuplicateNullsIndexSortObserver.java:56:
warning - @param argument "distinct" is not a parameter name.
[javadoc] C:\svn\10.4\java\engine\org\apache\derby\impl\sql\execute\UniqueWithDuplicateNullsIndexSortObserver.java:74:
warning - @returns is an unknown tag.
[javadoc] 13 warnings
[C:/svn/10.4]
Also, I'd like to see if 1 test bug can be backported that I ran into
when I ran with weme6.1 (i.e. using IBM's JSR169 implementation):
DERBY-3548 (fixed in 10.5) I just may not have time to look at it further today.
Myrna
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Dy...@Sun.COM.
Bryan Pendleton <bp...@amberpoint.com> writes:
>>> I plan to do this as soon as the fix for DERBY-3347 has been merged to
>>> the 10.4 branch, and then call for a vote on the new RC.
>
> I'm not sure I understand why the availability of additional fixes
> required the cancellation of the previous vote.
It probably doesn't. I just chose that subject because the 10.3 Feature
Release Summary on http://wiki.apache.org/db-derby/DerbyTenThreeRelease,
says "Sanity Testing and Voting (Cancelled - blocker DERBY-1963)".
> It seems to me that there are always additional fixes on the horizon.
>
> I don't think we were voting on whether 10.4.1.1 was to be the
> actual *release*;
Ummm, I thought we were. Had it not been for DERBY-3347 I was going to
put it on the mirrors and so on (provided it got PMC approval).
> we were simply voting on whether it was an
> acceptable release *candidate* for making available for advanced
> users interested in helping us test the software prior to release.
So you you considered 10.4.1.1 to be a BETA release candidate that, once
approved through a vote would become a BETA release that users could try
out? And that based on their feedback there would be a proper release
candidate that, if approved, would turn into a proper release?
That was not my plan. Sorry if I gave that impression. The current
description of the release process only mentions one vote, and my
understanding is that it determines if a release candidate can
become a release.
--
dt
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Bryan Pendleton <bp...@amberpoint.com>.
>> I plan to do this as soon as the fix for DERBY-3347 has been merged to
>> the 10.4 branch, and then call for a vote on the new RC.
I'm not sure I understand why the availability of additional fixes
required the cancellation of the previous vote.
It seems to me that there are always additional fixes on the horizon.
I don't think we were voting on whether 10.4.1.1 was to be the
actual *release*; we were simply voting on whether it was an
acceptable release *candidate* for making available for advanced
users interested in helping us test the software prior to release.
thanks,
bryan
Re: [VOTE CANCELLED] 10.4.1.1 release candidate
Posted by Kathey Marsden <km...@sbcglobal.net>.
Dyre.Tjeldvoll@Sun.COM wrote:
> I plan to do this as soon as the fix for DERBY-3347 has been merged to
> the 10.4 branch, and then call for a vote on the new RC.
>
>
It would be great also to see the alternate fix for DERBY-2892 go in and
then it will eliminate the necessity for the release note and
incompatibilities in the 10.4 release.
Kristian said:
Committed to trunk with revsion 646255.
I plan to backport the fix to 10.4 when the last patch has been
committed, and also investigate the possibility of backporting to 10.3.
[ Show » <https://issues.apache.org/jira/browse/DERBY-2892> ]
Kristian Waagan
<https://issues.apache.org/jira/secure/ViewProfile.jspa?name=kristwaa> -
09/Apr/08 03:21 AM 'derby-2892-1b-alternative_fix_partial.diff' is
slightly changed from 1a; reordered keywords (abstract public -> public
abstract) and removed an added blank line. Committed to trunk with
revsion 646255. I plan to backport the fix to 10.4 when the last patch
has been committed, and also investigate the possibility of backporting
to 10.3.
I'm not sure of the timeframe. I'd be happy to help in anyway I can.
Kathey