commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From sebb <seb...@gmail.com>
Subject Re: [VOTE] Release Collections 3.2.1 based on RC1
Date Sat, 12 Apr 2008 15:41:51 GMT
On 12/04/2008, Niall Pemberton <niall.pemberton@gmail.com> wrote:
> On Sat, Apr 12, 2008 at 4:21 PM, sebb <sebbaz@gmail.com> wrote:
>  >
>  > On 12/04/2008, Niall Pemberton <niall.pemberton@gmail.com> wrote:
>  >  > On Sat, Apr 12, 2008 at 2:18 PM, sebb <sebbaz@gmail.com> wrote:
>  >  >  >
>  >  >  > On 12/04/2008, Niall Pemberton <niall.pemberton@gmail.com> wrote:
>  >  >  >  > On Sat, Apr 12, 2008 at 1:25 PM, sebb <sebbaz@gmail.com>
wrote:
>  >  >  >  >  > On 12/04/2008, Niall Pemberton <niall.pemberton@gmail.com>
wrote:
>  >  >  >  >  >  > The Sling project has requested an OSGi enabled release
of Commons
>  >  >  >  >  >  >  Collections and since it seems unlikely that Collections
3.3 will be
>  >  >  >  >  >  >  ready in the timescale they require, I have created
a release
>  >  >  >  >  >  >  candidate for Collections 3.2.1. This is basically
a re-packaging of
>  >  >  >  >  >  >  the Collections 3.2 release with the appropriate
OSGi meta data in the
>  >  >  >  >  >  >  Collections jar. The only changes I have made to
the java source code
>  >  >  >  >  >  >  is to modify source file license headers to conform
to current policy
>  >  >  >  >  >  >  and remove tab characters. An m2 build has been
added to the
>  >  >  >  >  >  >  Collections 3.2 branch I created and the site has
been updated to
>  >  >  >  >  >  >  reflect the 3.2.1 release.
>  >  >  >  >  >  >
>  >  >  >  >  >  >  The artifacts are here:
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/
>  >  >  >  >  >
>  >  >  >  >  >  All look OK, and source zip agrees with SVN tag.
>  >  >  >  >  >
>  >  >  >  >  >
>  >  >  >  >  >  >  SVN Tag:
>  >  >  >  >  >  >  http://svn.apache.org/viewvc/commons/proper/collections/tags/COLLECTIONS_3_2_1_RC1/
>  >  >  >  >  >  >
>  >  >  >  >  >  >  Site:
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/site/
>  >  >  >  >  >  >  (note m2 generates relative links, so some don't
work - but the site
>  >  >  >  >  >  >  is for info and not included in the release artifacts)
>  >  >  >  >  >  >
>  >  >  >  >  >  >  Release Notes:
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/RELEASE-NOTES.html
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/site/release_3_2_1.html
>  >  >  >  >  >  >
>  >  >  >  >  >  >  RAT Report:
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/site/rat-report.html
>  >  >  >  >  >  >
>  >  >  >  >  >  >  CLIRR Report:
>  >  >  >  >  >  >  Clirr failed and so is not available
>  >  >  >  >  >  >
>  >  >  >  >  >
>  >  >  >  >  >  That's very unfortunate ... seems like the problem may
be due to a
>  >  >  >  >  >  doubly-nested class?
>  >  >  >  >
>  >  >  >  >
>  >  >  >  > Yes I googled this and saw others having the same problem, but
>  >  >  >  >  couldn't find a solution.
>  >  >  >  >
>  >  >  >
>  >  >  >  Found the solution - see:
>  >  >  >  https://sourceforge.net/tracker/index.php?func=detail&aid=1022446&group_id=89627&atid=590799
>  >  >  >  last e-mail on the page.
>  >  >
>  >  >
>  >  > Its sort of a solution, but I wouldn't want to hack the jar around as
>  >  >  part of the release process.
>  >  >
>  >  >
>  >  >  >  After I dropped the DoubleOrderedMap$1$* classes from 3.2.1, clirr
ran
>  >  >  >  OK - it gave:
>  >  >
>  >  >
>  >  > Thanks for doing that :)
>  >  >
>  >  >
>  >  >  >  INFO: 7003: org.apache.commons.collections.map.AbstractHashedMap$HashIterator:
>  >  >  >  Method 'public java.lang.Object next()' has been removed, but an
>  >  >  >  inherited definition exists.
>  >  >  >  INFO: 7003: org.apache.commons.collections.map.AbstractLinkedMap$LinkIterator:
>  >  >  >  Method 'public java.lang.Object next()' has been removed, but an
>  >  >  >  inherited definition exists.
>  >  >  >  INFO: 7003: org.apache.commons.collections.map.AbstractLinkedMap$LinkIterator:
>  >  >  >  Method 'public java.lang.Object previous()' has been removed, but
an
>  >  >  >  inherited definition exists.
>  >  >  >
>  >  >  >  The 3.2 jar does not contain these $1$* class names, presumably
>  >  >  >  because an older compiler was used.
>  >  >  >
>  >  >  >
>  >  >  >  >
>  >  >  >  >  >  >  RC2 has been built with m2 -
>  >  >  >  >  >
>  >  >  >  >  >  Since the build says it is targetted at 1.2, building
with Java 1.5 is
>  >  >  >  >  >  not ideal.
>  >  >  >  >  >  Building with 1.2 would be a pain, but using 1.3 should
be easy enough.
>  >  >  >  >  >
>  >  >  >  >  >  I just tried, and Ant test is OK with Java 1.3.1 and Ant
1.6.5
>  >  >  >  >  >
>  >  >  >  >  >  However, Maven1.0.2 with Java 1.3.1 gives some test failures,
e.g.:
>  >  >  >  >  >
>  >  >  >  >  >  Testsuite: org.apache.commons.collections.TestAllPackages
>  >  >  >  >  >  Tests run: 13009, Failures: 0, Errors: 109, Time elapsed:
16.359 sec
>  >  >  >  >  >  These are mainly java.lang.NoSuchMethodError failures.
>  >  >  >  >  >
>  >  >  >  >  >  If I run TestAllPackages in Ant, it works fine.
>  >  >  >  >  >  Perhaps a bug in Maven1?
>  >  >  >  >
>  >  >  >  >
>  >  >  >  > Looks that way. I had the same problem, but as you say since
the tests
>  >  >  >  >  pass on Ant with JDK 1.3 I didn't think it was a big issue.
>  >  >  >  >
>  >  >  >
>  >  >  >  I also tested the pre-built jar using Java 1.3.1 and Ant 1.6.5 -
>  >  >  >  TestAllPackages completed OK.
>  >  >  >
>  >  >  >  So I agree, it should be OK.
>  >  >  >
>  >  >  >
>  >  >  >  >
>  >  >  >  >  >  And using Java 1.2 with Ant 1.6.5 I get:
>  >  >  >  >  >
>  >  >  >  >  >     [javac] C:\ReleaseCheck\commons-collections-3.2.1-src\src\java\org\apache\commons\collections\buffer\BoundedBuffer.java:188:
>  >  >  >  >  >  Var
>  >  >  >  >  >  iable lock in class
>  >  >  >  >  >  org.apache.commons.collections.collection.SynchronizedCollection
not
>  >  >  >  >  >  accessible from inner class org.apache.comm
>  >  >  >  >  >  ons.collections.buffer.BoundedBuffer. NotifyingIterator.
>  >  >  >  >  >
>  >  >  >  >  >     [javac]             synchronized (lock) {
>  >  >  >  >  >
>  >  >  >  >  >  Maybe it's time to drop 1.2 support?
>  >  >  >  >
>  >  >  >  >
>  >  >  >  > Probably for Collections 1.3, but I wanted to create this release
with
>  >  >  >  >  the same options as the original 3.2 release.
>  >  >  >  >
>  >  >  >  >
>  >  >  >  >  >  > but m1 and ant builds are available - details here:
>  >  >  >  >  >  >  http://people.apache.org/~niallp/collections-3.2.1-rc1/site/building.html
>  >  >  >  >  >
>  >  >  >  >  >  The Ant test runs a different set of tests from M1 and
M2 - I
>  >  >  >  >  >  calculate that it runs 12957 tests, instead of 13009.
Perhaps the Ant
>  >  >  >  >  >  script should default to TestAllPackages as well?
>  >  >  >  >  >
>  >  >  >  >  >  Two sections both headed "Ant Goals" - perhaps they could
be combined
>  >  >  >  >  >  for the actual site release? Also the page says that the
default build
>  >  >  >  >  >  is Ant - so why release using M2?
>  >  >  >  >
>  >  >  >  >
>  >  >  >  > Because the m2 build uses the bundle plugin which puts the OSGi
meta
>  >  >  >  >  data in automatically and is the whole point of this release.
>  >  >  >
>  >  >  >  Might have been better to use the minimum version of Java, i.e. 1.4
for M2
>  >  >
>  >  >
>  >  > Why if its been tested on earlier versions (and I did test on 1.3 and
>  >  >  1.4 before posting the RC)?
>  >  >
>  >
>  >  Why not?
>  >
>  >  If I have a Java 1.4 application that uses collections, I'd be happier
>  >  knowing that it was built on Java 1.4, not 1.5.
>
>
> I'll try to remember next time, but the important point is testing on
>  older versions to find real issues

Indeed, but the information that you had tested on earlier versions
was not provided initially, so all one could go on was the compiler
version in the Manifest.

> rather than remembering something
>  that may or may cause user concern - I don't ever remember a case
>  where this has been raised by a user to date.

Well, I just did - JMeter uses collections ;-)

>
>  Niall
>
>
>  >  >
>  >  >  Niall
>  >  >
>  >  >
>  >  >  >  Recently 1.5 was used to build/test Commons Net, and that missed a
>  >  >  >  problem with earlier JVM classes. Using M2 and 1.4 showed the problem.
>  >  >  >
>  >  >  >  >
>  >  >  >  >  Niall
>  >  >  >  >
>  >  >  >  >
>  >  >  >  >  >  >  Vote is open for 72 hours
>  >  >  >  >  >  >
>  >  >  >  >  >  >  Thanks in advance for your feedback/votes.
>  >  >  >  >  >  >
>  >  >  >  >  >  >  Niall
>  >  >  >  >  >  >  ------------------------------------------------------------------------------------------------------------->
>  >  >  >  >  >  >
>  >  >  >  >  >  >  [  ] +1  I support this release
>  >  >  >  >  >  >  [  ] +0  I am OK with this release
>  >  >  >  >  >  >  [  ] -0   OK, but....
>  >  >  >  >  >  >  [  ] -1   I do not support this release
>
>  ---------------------------------------------------------------------
>  To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>  For additional commands, e-mail: dev-help@commons.apache.org
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message