Re: [14] Review Request: 8231438 [macos] The dark mode is not supported yet

[Sergey Bylokhov]

I remember that we have some related issues, which caused by the ignoring 
options in the plist.
But as far as I understand we do not have some generic bug/rfe to check the 
possibility of
different layout.

On 10/10/19 11:16 am, Alan Snyder wrote:

Is there a bug report for that? :-)


On Oct 10, 2019, at 10:36 AM, Sergey Bylokhov  
wrote:

On 10/10/19 10:16 am, Alan Snyder wrote:

The root problem is that the Info.plist for an executable is baked into the 
executable, and because all Java programs run from the command line use the 
same executable, they get the same Info.plist (as well as the same 
linked-against SDK version number).
Any other issues that involve Info.plist parameters or the SDK version are also 
impacted by this problem. I call this problem the launcher problem, because it 
affects all programs (not just Java programs) that are launched by a common 
executable.


I did not check, but as far as I understand it will start to work if we 
"reimplement"
the JDK layout to be like an app layout on top of the "java" command-line app,
something like jdk.app which will have own plist.info and java as a default 
launcher.






--
Best regards, Sergey.

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Sergey Bylokhov]

Hi, Dmitry.

I suggest to make this block more generic, and describe the  "### Client UI - 
Tests".
It would be useful to mention that the tests in this category use different key 
combinations,
which might be registered as a system shortcuts, and it could cause a test 
failure. It is suggested
to disable these shortcuts:
 - macOS config location.
 - Linux config location.
 - On window some other application may have global shortcuts

As an example of such issue you can provide CTRL+F1 which is used


On 10/25/19 3:27 am, Dmitry Markov wrote:

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880 

webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/ 


Some Client UI tests use the following key sequence to show/hide tooltip message: 
“CTRL” + “F1". However this key combination is reserved by operating system on 
OS X platform. As a results the tests fail.
Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry




--
Best regards, Sergey.

Re: RFC: JDK-8231991: Mouse wheel change focus on awt/swing windows

[Sergey Bylokhov]

Hi, Mario.

The code change looks fine, but could you confirm the new behavior align with 
the native applications?

On 10/25/19 6:52 am, Mario Torre wrote:

On Tue, 2019-10-08 at 11:42 +0200, Mario Torre wrote:

Ping?

Cheers,
Mario


Hi all,

I filed a bug related to how mouse wheel interact with focus events:

https://bugs.openjdk.java.net/browse/JDK-8231991

The issue seems to be that we still consider mouse button
press/release the wheel event. This is a bit tricky since X11 doesn't
really have the concept of mouse wheel but those events are indeed
synthesized from button press and converted into the scoll-up/down
logic, furthermore, it's possible to remap the buttons from the
de-facto standard of button 4 and 5 to anything else, however I'm not
sure how to detect which button is configured for the mouse wheel.

In this patch I introduced a couple of new constants in
XConstants.java for now hard coded to button 4 and 5, but we can
refactor later if needed the way those constants are filled (or if
anyone has an idea I'll be happy to fix this right away):

http://cr.openjdk.java.net/~neugens/JDK-8231991/webrev.00/

I also added a test case. The fix is essentially to skip the focus
logic if this is a mouse wheel event. I tried with a few different
configurations and all seems to work correctly, but I appreciate
feedback here since this areas is rather tricky.

I aim at backporting this fix - if approved - all the way down to 8,
I'm not sure if the addition of those two constants is accepted
however in the backports, perhaps I should just hard code the values
for these instead?

Thanks in advance for the feedback.

Cheers,
Mario





--
Best regards, Sergey.

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Alexey Ivanov]

Got it!
Then I'd just drop the quotes from the key names:
- `“CTRL” + “F1”`
+ `Ctrl + F1`

On 25/10/2019 16:52, Erik Joelsson wrote:

The HTML is generated from the markup and should not be edited by hand.

/Erik 

--
Regards,
Alexey

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Erik Joelsson]

On 2019-10-25 08:42, Alexey Ivanov wrote:

Hi Dmitry,

Apple uses macOS to refer to its operating system, so should we also 
use it instead of former OS X?


For HTML version, I would suggest using the following markup 
Ctrl +F1 for the keyboard shortcut. [1]



The HTML is generated from the markup and should not be edited by hand.

/Erik

I think using  elements for instructions on how to disable 
the shortcut is incorrect. In fact, it does not require any special 
markup.



What about this text?
Some client UI tests devoted to tooltips may fail on macOS 
platform. The tests use Ctrl +F1 key 
sequence to show/hide a tooltip. However, this key combination is 
reserved by the operating system. To run these tests correctly, the 
default global shortcut should be disabled: choose Apple menu > 
System Preferences, click Keyboard, then click Shortcuts; deselect 
"Turn keyboard access on/of" property.


I copied the way Apple uses in its documentation to refer to commands.

On 25/10/2019 11:27, Dmitry Markov wrote:

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880
webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/

Some Client UI tests use the following key sequence to show/hide 
tooltip message: “CTRL” + “F1". However this key combination is 
reserved by operating system on OS X platform. As a results the tests 
fail.

Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry



[1] https://www.w3.org/TR/html52/textlevel-semantics.html#the-kbd-element

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Alexey Ivanov]

Hi Dmitry,

Apple uses macOS to refer to its operating system, so should we also use 
it instead of former OS X?


For HTML version, I would suggest using the following markup 
Ctrl +F1 for the keyboard shortcut. [1]


I think using  elements for instructions on how to disable 
the shortcut is incorrect. In fact, it does not require any special markup.



What about this text?
Some client UI tests devoted to tooltips may fail on macOS platform. 
The tests use Ctrl +F1 key sequence to 
show/hide a tooltip. However, this key combination is reserved by the 
operating system. To run these tests correctly, the default global 
shortcut should be disabled: choose Apple menu > System Preferences, 
click Keyboard, then click Shortcuts; deselect "Turn keyboard access 
on/of" property.


I copied the way Apple uses in its documentation to refer to commands.

On 25/10/2019 11:27, Dmitry Markov wrote:

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880
webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/

Some Client UI tests use the following key sequence to show/hide 
tooltip message: “CTRL” + “F1". However this key combination is 
reserved by operating system on OS X platform. As a results the tests 
fail.

Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry



[1] https://www.w3.org/TR/html52/textlevel-semantics.html#the-kbd-element

--
Regards,
Alexey

Re: RFC: JDK-8231991: Mouse wheel change focus on awt/swing windows

[Mario Torre]

On Tue, 2019-10-08 at 11:42 +0200, Mario Torre wrote:

Ping?

Cheers,
Mario

> Hi all,
> 
> I filed a bug related to how mouse wheel interact with focus events:
> 
> https://bugs.openjdk.java.net/browse/JDK-8231991
> 
> The issue seems to be that we still consider mouse button
> press/release the wheel event. This is a bit tricky since X11 doesn't
> really have the concept of mouse wheel but those events are indeed
> synthesized from button press and converted into the scoll-up/down
> logic, furthermore, it's possible to remap the buttons from the
> de-facto standard of button 4 and 5 to anything else, however I'm not
> sure how to detect which button is configured for the mouse wheel.
> 
> In this patch I introduced a couple of new constants in
> XConstants.java for now hard coded to button 4 and 5, but we can
> refactor later if needed the way those constants are filled (or if
> anyone has an idea I'll be happy to fix this right away):
> 
> http://cr.openjdk.java.net/~neugens/JDK-8231991/webrev.00/
> 
> I also added a test case. The fix is essentially to skip the focus
> logic if this is a mouse wheel event. I tried with a few different
> configurations and all seems to work correctly, but I appreciate
> feedback here since this areas is rather tricky.
> 
> I aim at backporting this fix - if approved - all the way down to 8,
> I'm not sure if the addition of those two constants is accepted
> however in the backports, perhaps I should just hard code the values
> for these instead?
> 
> Thanks in advance for the feedback.
> 
> Cheers,
> Mario

-- 
Mario Torre
Associate Manager, Software Engineering
Red Hat GmbH 
9704 A60C B4BE A8B8 0F30  9205 5D7E 4952 3F65 7898

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Erik Joelsson]

No need for new webrev.

/Erik

On 2019-10-25 06:45, Dmitry Markov wrote:

Thank you for the review, Erik!
I will update the fix as you suggested before push if you do not mind.

Thanks,
Dmitry


On 25 Oct 2019, at 14:00, Erik Joelsson  wrote:

Looks good. I would put in a "the" before "operating system".

/Erik

On 2019-10-25 03:27, Dmitry Markov wrote:

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880 

webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/ 


Some Client UI tests use the following key sequence to show/hide tooltip message: 
“CTRL” + “F1". However this key combination is reserved by operating system on 
OS X platform. As a results the tests fail.
Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Dmitry Markov]

Thank you for the review, Erik!
I will update the fix as you suggested before push if you do not mind.

Thanks,
Dmitry

> On 25 Oct 2019, at 14:00, Erik Joelsson  wrote:
> 
> Looks good. I would put in a "the" before "operating system".
> 
> /Erik
> 
> On 2019-10-25 03:27, Dmitry Markov wrote:
>> Hello,
>> 
>> Could you review the fix for jdk14, please?
>> 
>>  bug: https://bugs.openjdk.java.net/browse/JDK-8232880 
>> 
>>  webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/ 
>> 
>> 
>> Some Client UI tests use the following key sequence to show/hide tooltip 
>> message: “CTRL” + “F1". However this key combination is reserved by 
>> operating system on OS X platform. As a results the tests fail.
>> Test documentation should be updated with some notes regarding this.
>> 
>> Thanks,
>> Dmitry
>> 

Re: RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Erik Joelsson]

Looks good. I would put in a "the" before "operating system".

/Erik

On 2019-10-25 03:27, Dmitry Markov wrote:

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880 

webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/ 


Some Client UI tests use the following key sequence to show/hide tooltip message: 
“CTRL” + “F1". However this key combination is reserved by operating system on 
OS X platform. As a results the tests fail.
Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry

RFR 8232880: Update test documentation with additional settings for client UI tooltip tests

[Dmitry Markov]

Hello,

Could you review the fix for jdk14, please?

bug: https://bugs.openjdk.java.net/browse/JDK-8232880 

webrev: http://cr.openjdk.java.net/~dmarkov/8232880/webrev.00/ 


Some Client UI tests use the following key sequence to show/hide tooltip 
message: “CTRL” + “F1". However this key combination is reserved by operating 
system on OS X platform. As a results the tests fail.
Test documentation should be updated with some notes regarding this.

Thanks,
Dmitry