Bug 493640 - Bug459961Test.testTransparentColorConstantReference consistently fails on CentOs - disable the test on GTK+ 3
Summary: Bug459961Test.testTransparentColorConstantReference consistently fails on Cen...
Status: VERIFIED FIXED
Alias: None
Product: Platform
Classification: Eclipse Project
Component: UI (show other bugs)
Version: 4.6   Edit
Hardware: PC CentOS
: P3 normal (vote)
Target Milestone: 4.6 RC3   Edit
Assignee: Sravan Kumar Lakkimsetti CLA
QA Contact:
URL:
Whiteboard:
Keywords: test
Depends on:
Blocks:
 
Reported: 2016-05-13 09:27 EDT by Dani Megert CLA
Modified: 2016-05-26 05:19 EDT (History)
4 users (show)

See Also:
daniel_megert: review+
Lars.Vogel: review+


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Dani Megert CLA 2016-05-13 09:27:57 EDT
4.6 RC1

Bug459961Test.testTransparentColorConstantReference consistently fails on CentOs

http://download.eclipse.org/eclipse/downloads/drops4/I20160512-1000/testresults/html/org.eclipse.e4.ui.tests.css.swt_ep46I-unit-cen64_linux.gtk.x86_64_8.0.html


expected:<255> but was:<237>

java.lang.AssertionError: expected:<255> but was:<237>
at org.eclipse.e4.ui.tests.css.swt.Bug459961Test.assertRGBAEquals(Bug459961Test.java:46)
at org.eclipse.e4.ui.tests.css.swt.Bug459961Test.testTransparentColorConstantReference(Bug459961Test.java:42)
at org.eclipse.test.EclipseTestRunner.run(EclipseTestRunner.java:743)
at org.eclipse.test.EclipseTestRunner.run(EclipseTestRunner.java:351)
at org.eclipse.test.CoreTestApplication.runTests(CoreTestApplication.java:36)
at org.eclipse.test.CoreTestApplication.run(CoreTestApplication.java:32)
at org.eclipse.equinox.internal.app.EclipseAppContainer.callMethodWithException(EclipseAppContainer.java:587)
at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:198)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:134)
at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:104)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:388)
at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:243)
at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:673)
at org.eclipse.equinox.launcher.Main.basicRun(Main.java:610)
at org.eclipse.equinox.launcher.Main.run(Main.java:1519)
at org.eclipse.equinox.launcher.Main.main(Main.java:1492)
at org.eclipse.core.launcher.Main.main(Main.java:34)
Comment 1 Dani Megert CLA 2016-05-18 13:19:44 EDT
Moving out of Neon. If you still want to fix something during RC3 then please retarget and make sure you follow the rules: https://www.eclipse.org/eclipse/development/plans/freeze_plan_4_6.php#FixPassAfterRC2
Comment 2 Dani Megert CLA 2016-05-20 02:41:40 EDT
(In reply to Dani Megert from comment #1)
> Moving out of Neon. If you still want to fix something during RC3 then
> please retarget and make sure you follow the rules:
> https://www.eclipse.org/eclipse/development/plans/freeze_plan_4_6.php#FixPassAfterRC2
> 

Let's disable this test until we have a fix.

+1 for RC3.
Comment 3 Sravan Kumar Lakkimsetti CLA 2016-05-20 03:02:59 EDT
I will check this. I am suspecting a configuration issue rather than a bug in this case. 

If it is indeed a configuration issue we may need to change the test.
Comment 4 Sravan Kumar Lakkimsetti CLA 2016-05-20 06:26:28 EDT
(In reply to Dani Megert from comment #2)
> (In reply to Dani Megert from comment #1)
> > Moving out of Neon. If you still want to fix something during RC3 then
> > please retarget and make sure you follow the rules:
> > https://www.eclipse.org/eclipse/development/plans/freeze_plan_4_6.php#FixPassAfterRC2
> > 
> 
> Let's disable this test until we have a fix.
> 
> +1 for RC3.

Here is the root cause. the Adwaita theme uses a different color than the standard color for background.

Expected color is extracted using COLOR_TRANSPARENT(SWT default color 255,255,255,0 not theme aware) 
Actual color is extracted from the background color from the theme currently in use. In this case Adwaita theme and in AdWaita the bg color is defined as 237,237,237. 

Now the test is failing while comparing RGB. There are two ways to fix

1. Change the expected color to SWT.COLOR_WIDGET_BACKGROUND(css based theme aware color)
2. Since the test is about verifying the transparency we can check only alpha value instead of RGBA.

Please suggest which way to proceed. I prefer the first approach
Comment 5 Alexander Kurtakov CLA 2016-05-20 07:13:04 EDT
(In reply to Sravan Kumar Lakkimsetti from comment #4)
> (In reply to Dani Megert from comment #2)
> > (In reply to Dani Megert from comment #1)
> > > Moving out of Neon. If you still want to fix something during RC3 then
> > > please retarget and make sure you follow the rules:
> > > https://www.eclipse.org/eclipse/development/plans/freeze_plan_4_6.php#FixPassAfterRC2
> > > 
> > 
> > Let's disable this test until we have a fix.
> > 
> > +1 for RC3.
> 
> Here is the root cause. the Adwaita theme uses a different color than the
> standard color for background.
> 
> Expected color is extracted using COLOR_TRANSPARENT(SWT default color
> 255,255,255,0 not theme aware) 
> Actual color is extracted from the background color from the theme currently
> in use. In this case Adwaita theme and in AdWaita the bg color is defined as
> 237,237,237. 
> 
> Now the test is failing while comparing RGB. There are two ways to fix
> 
> 1. Change the expected color to SWT.COLOR_WIDGET_BACKGROUND(css based theme
> aware color)
> 2. Since the test is about verifying the transparency we can check only
> alpha value instead of RGBA.
> 
> Please suggest which way to proceed. I prefer the first approach

Option 1. sounds better to me, better to compare the actual color.
Comment 6 Dani Megert CLA 2016-05-20 08:36:56 EDT
(In reply to Alexander Kurtakov from comment #5)
> Option 1. sounds better to me, better to compare the actual color.

+1
Comment 7 Eclipse Genie CLA 2016-05-22 14:03:34 EDT
New Gerrit change created: https://git.eclipse.org/r/73358
Comment 8 Sravan Kumar Lakkimsetti CLA 2016-05-24 07:47:43 EDT
Hi Alex and Lars,

Can you please review the patch so that we can push it for RC3?

Thanks
Sravan
Comment 9 Sravan Kumar Lakkimsetti CLA 2016-05-24 07:52:58 EDT
Opened SWT bug to track this in 4.7
Comment 10 Lars Vogel CLA 2016-05-24 11:29:37 EDT
+1 for RC3, change looks good
Comment 11 Lars Vogel CLA 2016-05-24 11:31:07 EDT
(In reply to Sravan Kumar Lakkimsetti from comment #9)
> Opened SWT bug to track this in 4.7

See Bug 494327
Comment 12 Sravan Kumar Lakkimsetti CLA 2016-05-24 11:53:34 EDT
Alex has commented that the code change fails on GTK 3.20. Now the option I have is to disable the test. I will upload a patch to disable the test shortly
Comment 13 Sravan Kumar Lakkimsetti CLA 2016-05-24 12:40:43 EDT
Alex, Lars and Dani

I have disabled the test for GTK3 for Neon. Can you please review the change again?

Thanks
Comment 15 Dani Megert CLA 2016-05-25 04:08:39 EDT
This is just a change in the tests and does not require 3 +1s.
Comment 16 Dani Megert CLA 2016-05-26 05:19:56 EDT
Verified in eclipse-SDK-I20160525-2000-win32-x86_64.