Looks better but I guess frame's locationOnScreen point would be
different from Menu's centrepoint so we should not use same "point"
instance variable for both. Guess local "point" variable for
executeTest() should be retained.
Regards
Prasanta
On 19-Jul-19 1:11 PM, Tejpal Rebari wrote:
Hi All,
Added the blockTillDisplayed() method instead of 1000ms delay.
here is the latest changes.
webrev : http://cr.openjdk.java.net/~pkbalakr/Tejpal/8194944/webrev1
Thanks and regards
Tejpal Rebari
On 08-Jul-2019, at 1:43 PM, Prasanta Sadhukhan
<[email protected]
<mailto:[email protected]>> wrote:
Adding delay is indeterministic. I guess it will be better if you
use blockTillDisplayed() method which you can find in many of the
testcase(s) like javax/swing/JComboBox/8182031/ComboPopupTest.java,
javax/swing/JEditorPane/6917744/bug6917744.java
Regards
Prasanta
On 08-Jul-19 1:02 PM, Tejpal Rebari wrote:
Hi All,
Please review a test fix for the issue where the test fails on Mac
and Linux intermittently.
The test fails because of “component must be showing on the screen
to determine its location” error.
Which is because the test tries to get the location of the component
and the component is not displayed on the screen yet.
The test tries to get the location of menu in the executeTest method
so By adding some delay before the executeTest call solves the problem.
Tested in our internal build and test system and test passes on all
the three Platforms.
Bug : https://bugs.openjdk.java.net/browse/JDK-8194944
webrev: http://cr.openjdk.java.net/~pkbalakr/Tejpal/8194944/webrev0
Thanks and regards
Tejpal
