Re: [pool] 1.5-RC1 available for review

Phil Steitz Sat, 30 May 2009 04:43:28 -0700

Luc Maisonobe wrote:

Phil Steitz a écrit :

sebb wrote:

On 29/05/2009, Phil Steitz <phil.ste...@gmail.com> wrote:

The files are here:
 http://people.apache.org/~psteitz/commons-pool-1.5-RC1/

findbugs filter and license-header.txt are missing from source archives.

This is intended.  The associated reports are not enabled in the release
pom, so there is no reason to distribute these files with the release.

RELEASE-NOTES.txt is missing from the SVN tag.


I feel umcomfortable with this choice. Users may want to recompile from
source themselves and modify the code to suit specific needs sometimes.
It would be nice if in this case they could get the same level of
confidence we have by enabling findbugs. If there are some known issues
that we filter out (often false positive), they may encounter them too
and wonder if their changes are to be blamed for this.

Interesting perspective. I will put the development reports back in.Thanks for the feedback.


Phil

Luc

This is a generated file.  I guess I could check in before making the
tag, but then it would not be generated from the tag itself.

NOTICE.txt has 2008.

Good catch.  Thanks!

I'm not sure why, but the $Id: SVN tags in some of the XML files
differ between the source file and a checkout of the SVN tag. This
should not happen if the tag is used to create the build.

The release was generated from the tag.  Which files?

It would be nice if the sources jar manifest had Spec. and Impl. entries.

Looks like you fixed this.  Thanks!

Ant test works OK on Java 1.3.1

Mvn test (Java 1.6.0) reports

[WARNING] Using platform encoding (Cp1252 actually) to copy filtered
resources, i.e. build is platform dependent!

Also the test fails for me:

Failed tests:
testBorrowObjectFairness(org.apache.commons.pool.impl.TestGenericObjectPool)

 Time elapsed: 2.468 sec  <<< FAILURE!
junit.framework.AssertionFailedError
    at junit.framework.Assert.fail(Assert.java:47)
    at junit.framework.Assert.fail(Assert.java:53)
    at
org.apache.commons.pool.impl.TestGenericObjectPool.testBorrowObjectFairness(TestGenericObjectPool.java:1434)


The Test case ought to be a bit more explicit as to why the test failed.

I suspect it may be an error in the unit-test code - the TestThread
class has various fields that are accessed across threads without any
synchronisation.

I don't think the TestThread instance fields are accessed across
threads.  The test case is timing sensitive, as indicated in the
comment.  That said, this is a new test case, testing new code, so we
need to find out exactly what is going on.

IIUC what the test is trying to do (Mark can correct me if I am wrong),
it is launching 500 threads, numbered by their array indices,  with 10
ms delays between starts and they confirming that they get served in
start order.  The way that it confirms order is by leveraging the fact
that the test factory produces numbered objects, 0, 1, 2, ....,
maxActive - 1.  So thread i should get object i mod maxActive.  That is
what is being tested in the line you have narrowed the failure to.

 The tag is here:
http://svn.apache.org/repos/asf/commons/proper/pool/tags/POOL_1_5_RC1/

NOTICE.txt says:

Copyright 1999-2008 The Apache Software Foundation

I think this does need to be fixed.

DOAP file does not have an AL header (not a blocker, I'll fix trunk)

Thanks!

Thanks for reviewing and for your help with the issues.

Phil

 Have at it!

 Thanks!

 Phil

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

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

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



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



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

Re: [pool] 1.5-RC1 available for review

Reply via email to