commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Phil Steitz <>
Subject Re: [VOTE] Release Commons Lang 2.5 based on RC1
Date Sun, 21 Feb 2010 15:12:42 GMT
Phil Steitz wrote:
> Niall Pemberton wrote:
>> The last release of Lang was in March 2008 and there have been quite a
>> few bug fixes and enhancements since then. Many of the changes in the
>> current trunk which are compatible have been backported to the Lang
>> 2.x branch to create this release.
>> [ ] +1 Yes go ahead an release based on RC1
>> [ ] -1 No, because...
>> Lang 2.5 RC1 is available for review here:
>> Maven artifacts are here:
>> Details of changes since 2.4 are in the release notes:
>> I have tested this with JDK 1.3 & 1.4 using ant and JDK 1.5 & 1.6 using maven2.
>> (note JVMRandom tests failed for JDK 1.3 & 1.4 - but I re-ran a couple
>> of times and both passed)
>> The tag is here:
>> Site:
>> (note some *relative* links are broken and the 2.5 directories are
>> not yet created - these will be OK once the site is deployed)
>> Clirr Report (compared to 2.4):
>> RAT Report:
>> tia
>> Niall
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> For additional commands, e-mail:
> Sigs, hashes are good
> Clirr is good
> Tag matches src
> Tested build on Apple JDKs 1.4.2_18, 1.5.0_19, 1.6.0_17
> Sun JDKs 1.4.2_19, 1.5.0_22, 1.6.0_17 (using Ant for 1.4 JDKs).
> I did see one RandomUtilsTest failure (Apple JDK 1.6).  These tests
> are failing at too high a frequency, which I am investigating; but I
> do not see this as a showstopper, so
> +1
> Phil

I hate to do this; but I would like to change my vote to +0.  The
random data tests are failing at a high enough frequency to be
annoying / alarming to users.

I investigated the high incidence of test failures and found nothing
wrong with what the tests are doing and nothing to indicate
systematic bias in the data being generated; but the addition of 40+
chi-square tests in the test methods added in r907159 makes the
probability of failure in a given run > 1/25.  This is why there is
a high incidence of test failures.

I verified that failures appear to be evenly distributed (too many,
too few even/edd, too many, too few above/below range midpoints) and
that the chisquare statistics are being computed correctly, with the
right critical values applied.

If you do cut another RC, I would recommend one of the following:

1) Grab / copy and extend [math]'s RetryTestCase (will cut incidence
of failure in half)

2) Disable the stochastic test cases for the release

3) Reduce sensitivity of the chi-square test (change to e.g., .0005
level of significance)

4) Reduce the number of tests

My recommendation is 2) - leave in the source but comment out.  The
tests are valuable as they would fail regularly and miserably if
there were systematic bias (as there used to be on odd/even); but
without reducing significantly the number of tests or the
sensitivity (or limiting to a single "successful" PRNG sequence),
there is no way to leave them all in without generating an
annoyingly high rate of random failures.


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message