Thanks Yuri.
Here is the updated webrev with suggested change.
http://cr.openjdk.java.net/~aghaisas/8031423/webrev.01/
Regards,
Ajit
-Original Message-
From: Yuri Nesterenko
Sent: Thursday, April 14, 2016 6:48 PM
To: Ajit Ghaisas; Sergey Bylokhov; awt-dev@openjdk.java.net
Subject: Re:
I think it makes sense to remove the typedef of Pixel and directly use
unsigned long since the sole reference to Pixel is in order to prepare
values to pass to XSetForeground which expects "unsigned long".
Updated webrev : http://cr.openjdk.java.net/~prr/8154269.1/
Boolean is used in about 22
On 4/14/2016 9:25 PM, Reto Merz wrote:
This is a small part of our focus/validation framework.
I just try to explain it as simple as possible.
javax.swing.InputVerifier gives only access to the source component.
We need to know the target component.
Would it be possible to overload verify(..)
https://bugs.openjdk.java.net/browse/JDK-8154269
http://cr.openjdk.java.net/~prr/8154269/
https://bugs.openjdk.java.net/browse/JDK-8047931
listed a number of X11 and medialib files that are unused in the build.
In looking at the VDrawingArea ones it lead to realising
we still include Intrinsic.h
Thank for explanation.
It seems to me such functionality belongs to the input validation domain
and should not be mixed up with the Swing focus transfer internals.
Swing has javax.swing.InputVerifier which you could extend and set to
each component of your framework. I suppose it would be
Hi,
Please review fix for bug 8129569 in JDK 9
BUG ID: https://bugs.openjdk.java.net/browse/JDK-8129569
Webrev: http://cr.openjdk.java.net/~arapte/ramesh/8129569/webrev.01/
Issue:
While checking for Frame State when Maximum bounds is set in Mac and Ubuntu
Operating system sets
Hi Ajit,
it's good you didn't change java.awt.Robot!
As to the test, you have created that instance of Robot:
now, pass it to Util.waitForIdle() and not recreate every time.
Thanks,
-yan
On 04/14/2016 02:59 PM, Ajit Ghaisas wrote:
Hi,
Adding delay in test code in main thread fixes this
Okay... You said the you need to intercept the focus request in the
place where the target is set.
From my point of view opening
Component#requestFocusInWindow(CausedFocusEvent.Cause cause) resolves that.
But it looks like you need extra. What is the final purpose? Maybe you
want to reject
Thanks for your investigation Semyon.
Make Component#requestFocusInWindow(CausedFocusEvent.Cause cause)
public/protected will not help us.
I can explain it more detailed if you want.
What would help is to introduce a "protected boolean
Component#isRequestFocusAccepted()"
which is invoked by
Hi,
Adding delay in test code in main thread fixes this test timeout.
I considered adding some delay in Robot.waitForIdle() method so that the
calling thread yields, but realized that, it will have undesirable effect of
slowing down test runs as many tests use that method.
This
Hello Semyon,
We need to know the target component: "which component will be focused".
Target component means the component which should be "really" focused,
so this can only be known after all the FocusCycle / FocusTraversalPolicy logic
has processed.
DefaultKeyboardFocusManager
Hi Ramesh,
Please check the test fails on 64 bit Ubuntu 14.04 LTS.
Thanks,
Ambarish
From: Ramesh Gangadhar
Sent: Thursday, April 14, 2016 12:04 PM
To: awt-dev@openjdk.java.net
Subject: Review Request For 8129569 : Test
Hi,
Please review fix for bug 8129569 in JDK 9
BUG ID: https://bugs.openjdk.java.net/browse/JDK-8129569
Webrev: http://cr.openjdk.java.net/~jdv/Ramesh/8129569/webrev.00/
Issue:
While checking for Frame State when Maximum bounds is set Mac Operating system
sets state =
Hi Reto,
It is unlikely that AWTAccessor API may be opened. It gives access to
internal undocumented methods not to even mention that those methods
have private and package accesses.
As I understand you need the way to intercept focus transfer initiated
by a traversal key. Why to subclass the
14 matches
Mail list logo
