harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Lang Yang <yangl...@gmail.com>
Subject Re: [general] Windows 7 support
Date Tue, 29 Jun 2010 06:31:10 GMT
Hello guys,

I have ran tests on Windows 7. The test result shows there 2 failures and 32
errors occurred. 99.87% of 26276 test cases successfully passed.

Test environment:
Windows 7 Professional
test-jre-vm-info:
     [echo] java -version
     [echo] Apache Harmony Launcher : (c) Copyright 1991, 2010 The Apache
Software Foundation or its licensors, as applicable.
     [echo] java version "1.5.0"
     [echo] Apache Harmony (1.5.0)
     [echo] DRLVM (1.5.0-r958818)
     [echo] pre-alpha : not complete or compatible
     [echo] svn = r958818, (Jun 29 2010), Windows/ia32/msvc 1500, release
build
     [echo] http://harmony.apache.org
     [echo] hy.test.vm.name = drl

The two failures:
1. Class BasicScrollBarUITest, name testGetThumbBounds

expected: 16, actual: 0

junit.framework.AssertionFailedError: expected: 16, actual: 0
at
javax.swing.plaf.basic.BasicScrollBarUITest.checkIsCloseTo(BasicScrollBarUITest.java:267)
at
javax.swing.plaf.basic.BasicScrollBarUITest.testGetThumbBounds(BasicScrollBarUITest.java:85)
at java.lang.reflect.VMReflection.invokeMethod(VMReflection.java)
at javax.swing.BasicSwingTestCase.runBareSuper(BasicSwingTestCase.java:116)
at javax.swing.BasicSwingTestCase.runBareImpl(BasicSwingTestCase.java:121)
at javax.swing.BasicSwingTestCase$1.run(BasicSwingTestCase.java:135)
at java.lang.Thread.run(Thread.java:669)

2. Class NetworkinterfaceTest, name dtest_getInetAddresses

validate that address without permission is not returned

junit.framework.AssertionFailedError: validate that address without
permission is not returned
at
org.apache.harmony.luni.tests.java.net.NetworkInterfaceTest.test_getInetAddresses(NetworkInterfaceTest.java:163)
at java.lang.reflect.VMReflection.invokeMethod(VMReflection.java)


Errors occurred on Swing, luni and nio packages. Detailed results please
refer to  [0]

Please let me know if you need more information about this.

Regards,

Lang


[0] http://bit.ly/d5GoP5

On Mon, Jun 28, 2010 at 7:06 PM, Antony Miguel <aemiguel@gmail.com> wrote:

> Hi Mark,
>
> Thanks for the build but its giving me a "MSVCR71D.dll not found"
> error - I think thats because its a debug build and I dont have Visual
> Studio installed, any change of a release build?
>
> I'll grab the HDK once there is a release build available and run the
> tests and let you know.
>
> cheers
>
> A
>
> On 28 June 2010 23:33, Mark Hindess <mark.hindess@googlemail.com> wrote:
> >
> > In message <AANLkTilUQ6E43kGN2aJDiA8sfXWoCyoEoLRSFtIybfYN@mail.gmail.com
> >,
> > Antony Miguel writes:
> >>
> >> 32-bit please :)
> >
> > http://people.apache.org/~hindessm/win7/
> >
> > These are debug rather than release builds.  I'm happy to make release
> > builds if they are more suited to your requirements.  I've only made
> > java5 artifacts but let me know if you wanted java7.
> >
> > It would be useful if you could run the tests and report the results
> > back.  If you get the hdk build and unzip it you can run the tests with:
> >
> >  ant -f harmony-5.0-hdk-958744/build/test/build.xml
> >
> >
> >> I believe XP->Vista broke with tradition much more Vista->Windows 7 so
> >> I would imagine if Harmony works on Vista then it would very likely
> >> work for the vast majority on Windows 7.
> >>
> >> As I've said I would suggest that that "it will probably work on
> >> future versions" should be the default assumption, or at least that
> >> the version check could be a warning which could be overridden with
> >> a command line switch?  That way you could retain the forced exit if
> >> you wished but people creating products on top of Harmony would have
> >> the option to use the command line switch in the knowledge that their
> >> products would work wherever possible and wouldn't just exit because
> >> Microsoft had released a new version.
> >
> > I don't disagree.  This change was easier for now.  Assuming the test
> > results are consistent on windows 7 then accepting dwMajorVersion >=
> > 6 is probably reasonable.  However, until we know if the tests run
> > correctly on windows 7, allowing versions greater than that doesn't
> > really make sense.
> >
> > Regards,
> >  Mark.
> >
> >
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message