Thanks Tim.
Our current classlib-test adaptor in BTI infrastructure excludes test
cases from AWT/SWING module.
Those test cases are in classlib-gui-test but I did not see its results in
previous regression reports.
<exec-ant dir="${classlib-test.parameters.svn.classlib.location}"
targets="test"
failonerror="false"
failproperty="classlib-test.run.failed">
<jvmarg line="-Xmx800M" />
<sysproperty key="test.jre.home"
value="${classlib-test.tested.jre}" />
<sysproperty key="tested.jre.options" value="${jre.options}" />
<sysproperty key="exclude.interm" value="true" />
<sysproperty key="exclude.module" value="awt,swing" />
</exec-ant>
I also think that the failures and test case coverage for the AWT/SWING
modules needs more investigation.
I once did full classlib-gui-test and saw some less reproducible errors (in
different runs/platforms) there, but the basic functionality looks fine.
On Fri, Nov 7, 2008 at 8:34 AM, Tim Ellison <[EMAIL PROTECTED]> wrote:
> Sian January wrote:
> > Hi Chunrong,
> >
> > I'm not able to check this myself at the moment because I didn't have
> > an existing Linux setup and I've run into some problems on the machine
> > I have been trying to use. Do you think it's a blocking issue for M8?
> >
> > Apart from this one potential issue I think we're looking good for M8.
> > I've done some manual testing on Windows with Eclipse and Tomcat as
> > well as running the Harmony tests and everything has been fine so far.
> >
> > How is anyone else's testing going? Are we ready to start voting
> tomorrow?
>
> I see some crashes in the AWT code that are reported in JIRA already,
> but they do not appear to be regressions from M7.
>
> Regards,
> Tim
>
>
>
>
>
