commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Luc Maisonobe <Luc.Maison...@free.fr>
Subject Re: [VOTE] Release math 1.2 (second attempt)
Date Tue, 19 Feb 2008 20:32:14 GMT
sebb a écrit :
> On 19/02/2008, Phil Steitz <phil.steitz@gmail.com> wrote:
>> This is a vote to release Commons Math 1.2.
>>
> 
> I assume the SVN tag is:
> http://svn.apache.org/repos/asf/commons/proper/math/tags/MATH_1_2_RC3/
> 
>> The zips / tars are here:
>> http://people.apache.org/~psteitz/math-1.2-RC3/
>>
>> The site included in the binary distro is here:
>> http://people.apache.org/~psteitz/math-1.2-RC3/docs/
>>
> 
> Could not find any details on which Java releases are supported.
> 
> The front page says:
> 
> "No external dependencies beyond Commons components and the core Java
> 2 platform."
> 
> AFAIK this merely implies Java 1.2+; however the jar manifests show
> version 1.3. It would be helpful if this was clearly stated on the
> web-site.
> 
> I've tried building the code using Java 1.3.2_20 and Ant, however the
> tests fail with:
> 
> internal-test:
>     [junit] Using loader null on class
> org.apache.tools.ant.taskdefs.optional.junit.XMLJUnitResultFormatter:
> java.lang.NoClassDefFoundError: org/w3c/dom/Node

This seems to be a problem with running ant itself, not [math] which 
does not deal with xml.

> 
> The Ant build works OK on Java 1.4.2_16, however the generated jar
> file manifest is rather minimal; it should have more info.
> 
> The Maven 1 build seems to work OK on 1.4, however the generated jar
> file contains two copies of META-INF/LICENSE.txt.
> 
> The Maven 1.0.2 build seems to work OK on Java 1.3, however the test
> suite seems to stall when running the test:
> 
> org.apache.commons.math.ode.GraggBulirschStoerIntegratorTest

I suspect the following ones are related to the known wrong behaviour of 
Math.cos prior to Java 1.4 (see 
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4345903). These tests 
also correspond to corner-case behaviour (dealing with infinity and NaN) 
where this version of JDK had issues.

I don't know if the previous case is similar or not. There is a simple 
unprotected loop in these tests cases which involves cosines and sine. 
It is in the TestProblem3.computeTheoreticalState() method. Could you 
change the loop from:

do {
   // some computation about Kepler equation
} while (Math.abs(corr) > 1.0e-12);

to

corr = 999.0;
for (int i = 0; (i < 50) && (Math.abs(corr) > 1.0e-12); ++i) {
   // some computation about Kepler equation
}

and check if the test exits ?

Luc

> 
> If this test is commented out, then all the other tests complete,
> however there are several errors.
> 
> The failures appear to be:
> 
> Testsuite: org.apache.commons.math.complex.ComplexUtilsTest
> Tests run: 69, Failures: 10, Errors: 0, Time elapsed: 0.188 sec
> 
> Testcase: testCosInf(org.apache.commons.math.complex.ComplexUtilsTest):	FAILED
> 1.543080634815244 is not NaN.
> junit.framework.AssertionFailedError: 1.543080634815244 is not NaN.
> 	at org.apache.commons.math.TestUtils.assertEquals(TestUtils.java:58)
> 	at org.apache.commons.math.TestUtils.assertEquals(TestUtils.java:48)
> 	at org.apache.commons.math.TestUtils.assertSame(TestUtils.java:70)
> 	at org.apache.commons.math.TestUtils.assertSame(TestUtils.java:78)
> 	at org.apache.commons.math.complex.ComplexUtilsTest.testCosInf(ComplexUtilsTest.java:173)
> 
> <snip/>
> 
> and
> 
> Testsuite: org.apache.commons.math.complex.ComplexTest
> Tests run: 82, Failures: 9, Errors: 0, Time elapsed: 0.187 sec
> 
> Testcase: testCosInf(org.apache.commons.math.complex.ComplexTest):	FAILED
> 1.543080634815244 is not NaN.
> junit.framework.AssertionFailedError: 1.543080634815244 is not NaN.
> 	at org.apache.commons.math.TestUtils.assertEquals(TestUtils.java:58)
> 	at org.apache.commons.math.TestUtils.assertEquals(TestUtils.java:48)
> 	at org.apache.commons.math.TestUtils.assertSame(TestUtils.java:70)
> 	at org.apache.commons.math.TestUtils.assertSame(TestUtils.java:78)
> 	at org.apache.commons.math.complex.ComplexTest.testCosInf(ComplexTest.java:379)
> 
> <snip/>
> 
> [I can provide the full test results if required ]
> 
>> Release notes:
>> http://people.apache.org/~psteitz/math-1.2-RC3/RELEASE-NOTES.txt
>>
>> The vote will remain open for 72 hours.
>>
>> Thanks in advance for your votes.
>>
>> Phil
>>
>> 8--------------------------------------------------------------------------------------->
>>
>> [  ] +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
> 
> 



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


Mime
View raw message