I am not sure that is necesary for the point of the test, I assume it
was a generic
problem with switching from any L&F to any other.
The test does assume at least 2 installed L&Fs ... but since we have at
least
4 on each platform right now that's probably OK.
So +1
-phil.
On 11/29/18, 2:09 AM, Prasanta Sadhukhan wrote:
Hi Sergey,
I see that only the 1st 2 L&Fs UIManager.getInstalledLookAndFeels()[0]
and UIManager.getInstalledLookAndFeels()[1]
are being tested? Why is it so? Shouldn;t we iterate over all L&Fs to
see if preview panel works when L&F changes from one to other and not
just from 1st to 2nd L&F?
Regards
Prasanta
On 27-Nov-18 11:28 AM, Krishna Addepalli wrote:
Hi Sergey,
The changes look fine.
Krishna
On 27-Nov-2018, at 9:59 AM, Sergey Bylokhov<sergey.bylok...@oracle.com> wrote:
Hi, Krishna.
Thank you for review, the new version of the fix:
http://cr.openjdk.java.net/~serb/8198398/webrev.01
On 13/11/2018 22:25, Krishna Addepalli wrote:
Hi Sergey,
I think you need to wrap the code in run() method inside a try/finally block,
since if setLookAndFeel function throws exception, the frame will not be
disposed.
Other than that, it looks fine.
Thanks,
Krishna
-----Original Message-----
From: Sergey Bylokhov
Sent: Saturday, November 10, 2018 5:35 AM
To:swing-dev@openjdk.java.net
Subject:<Swing Dev> [12] Review Request: 8198398 Test
javax/swing/JColorChooser/Test6199676.java fails in mach5
Hello.
Please review the fix for jdk 12.
Bug:https://bugs.openjdk.java.net/browse/JDK-8198398
Webrev:http://cr.openjdk.java.net/~serb/8198398/webrev.00
Exception reported in this bug is not reproducible since we run our tests in othervm mode, but we
have an opposite bug, the jtreg may report "PASS" status for the tests even before the
"Runnable" will start to execute.
Change description:
- bug6735286.java: invokeLater() was replaced by the invokeAndWait(), so the
jtreg will wait till completion of the test.
- Test6199676.java: the test used a sequence of invokeLater(), each executes
part
of test case. It was replaced by one step inside invokeAndWait().
I also confirm that the updated tests still can be used for verification of:
"@bug 6199676" and "@bug 6735286"
--
Best regards, Sergey.
--
Best regards, Sergey.
