Hi Semyon, Sure I understand your concern.

Will update web rev on this and get back to you soon.

Thanks and regards,
Shashi

> On 07-Dec-2017, at 12:12 AM, Semyon Sadetsky <semyon.sadet...@oracle.com> 
> wrote:
> 
> Hi Shashi,
> 
> On 12/05/2017 08:13 PM, Shashidhara Veerabhadraiah wrote:
> 
>> Hi Semyon, the mouse key state is already part of the input to the function 
>> via the flags variable(if there is any) and same message is already sent to 
>> the action event where
> The flags parameter doesn't contain any modifiers because it is always equals 
> to zero.
>> they only save the message. This part is kept same as earlier.
> The action event receives modifiers from  the 3rd parameter of the 
> SendMouseEvent function which now may not contain any mouse modifiers, if I 
> correctly understand the root cause of the issue. My question is about 
> compatibility. If this code whenever have been working (for example, in 
> Windows 7) the BUTTON1_DOWN_MASK have been always set in the action event. 
> After your fix it may not be set (or even will never be set anymore). Does it 
> affect old applications which might expect a button presence in the action?
> 
> --Semyon
>> 
>> shashi
>> 
>> -----Original Message-----
>> From: Semyon Sadetsky
>> Sent: Tuesday, December 5, 2017 9:47 PM
>> To: shashidhara veerabhadraiah <shashidhara.veerabhadra...@oracle.com>
>> Cc: awt-dev@openjdk.java.net
>> Subject: Re: <AWT Dev> [10] JDK-8146537: TrayIcon Action Listener doesnt 
>> work in WIndows 10
>> 
>> Hi Shashi,
>> 
>> I meant, since before the modifier has been sent with the action it should 
>> be sent now as well for compatibility. Or this code has never worked?
>> 
>> --Semyon
>> 
>> On 12/04/2017 09:30 AM, shashidhara veerabhadraiah wrote:
>>> Hi Semyon, Thank for your review.
>>> 
>>> BUTTON1_DOWN_MASK upon the GetJavaModifiers() provide if the left mouse 
>>> button virtual key is pressed or not. Currently we are unable to capture 
>>> left button click state using the ::GetKeyState(). The event 
>>> NIN_BALLONUSERCLICK is generated only upon the left mouse click and any 
>>> other action does not raise this event. Hence there is no need to check it 
>>> once again.
>>> 
>>> SendActionEvent() takes on the GetActionModifiers() which is kept as it.
>>> 
>>> Hope I answer your question. Have a good day Semyon.
>>> 
>>> Thanks and regards,
>>> Shashi
>>> 
>>>> On 04-Dec-2017, at 9:45 PM, Semyon Sadetsky <semyon.sadet...@oracle.com> 
>>>> wrote:
>>>> 
>>>> Hi Shashi,
>>>> 
>>>> Shouldn't the BUTTON1_DOWN_MASK be added to the modifiers of the action?
>>>> 
>>>> --Semyon
>>>> 
>>>> On 11/27/2017 09:16 PM, Shashidhara Veerabhadraiah wrote:
>>>>> Hi Sergey, Here is the Webrev containing fixes for your comments.
>>>>> 
>>>>> http://cr.openjdk.java.net/~sveerabhadra/8146537/webrev.01/
>>>>> 
>>>>> Thanks and regards,
>>>>> Shashi
>>>>> 
>>>>> -----Original Message-----
>>>>> From: Sergey Bylokhov
>>>>> Sent: Wednesday, November 22, 2017 4:31 AM
>>>>> To: Shashidhara Veerabhadraiah
>>>>> <shashidhara.veerabhadra...@oracle.com>; awt-dev@openjdk.java.net
>>>>> Subject: Re: [10] JDK-8146537: TrayIcon Action Listener doesnt work
>>>>> in WIndows 10
>>>>> 
>>>>> Hi, Shashi.
>>>>> On 20/11/2017 02:10, Shashidhara Veerabhadraiah wrote:
>>>>>> Summary: The windows GetKeyState() api provides the key state of
>>>>>> the requested key as of that instant by sampling the current
>>>>>> thread's key messages from the message queue. Any misses either a
>>>>>> thread ran faster or slower it is possible to miss the key
>>>>>> messages. This is the problem that's happening under windows 10
>>>>>> wherein the key states are being null for the mouse button press.
>>>>>> This problem is mentioned in the msdn as well. On the other hand, a
>>>>>> tray icon message click produces NIN_BALLOONUSERCLICK event only
>>>>>> upon the left mouse button click and hence there is no check required 
>>>>>> for the same.
>>>>> The old code also request the state of the different keys like shift, 
>>>>> control, etc. And after the fix we lose this information, probably we 
>>>>> should apply it on top of the modifiers?
>>>>> 
>>>>> Another unclear thing is that we use
>>>>> java_awt_event_InputEvent_BUTTON1_DOWN_MASK which is a mask for 
>>>>> InputEvent as modifier for ActionEvent which does not have such mask.
>>>>> 
>>>>> --
>>>>> Best regards, Sergey.
> 

Reply via email to