<AWT Dev> [9] Review request for 8176528: Progress state for window is not displayed in taskbar

[Sergey Bylokhov]

Looks fine.

> 15 марта 2017 г., в 20:49, Philip Race <philip.race at oracle.com> написал(а):
> 
> +1 from me.
> 
> -phil.
> 
> On 3/15/17, 10:47 AM, Alexander Zvegintsev wrote:
>> 
>> How about that version?
>> 
>> http://cr.openjdk.java.net/~azvegint/jdk/9/8176528/04/ <http://cr.openjdk.java.net/%7Eazvegint/jdk/9/8176528/04/>
>> Thanks,
>> Alexander.
>> On 15/03/2017 20:18, Philip Race wrote:
>>> 
>>> Even if all we can add (to what is already proposed) is
>>> "whether an icon for a window is displayable in the task bar is
>>> dependent on all of window type, platform, and implementation"
>>> then I think that clearer than the current words which could be
>>> misinterpreted as meaning something like addressing that
>>> if you call hide() then of course it disappears ..
>>> 
>>> -phil.
>>> 
>>> On 3/15/17, 10:02 AM, Sergey Bylokhov wrote:
>>>> 
>>>>>>>> It is not necessary, if the frame has utility type.
>>>>>>> 
>>>>>>> So can we say this with that qualification ?
>>>>>> 
>>>>>> This is also implementation detail, that such windows are not visible in taskbar. This is not specified anywhere and can be changed in the future.
>>>>> 
>>>>> If this is consistent in our current implementation on all platforms then
>>>>> there is no problem making this part of the spec and changing the spec in
>>>>> the future if that changes.
>>>> 
>>>> It was not implemented on systems other than windows, so if we change the signature we will not be able implement/change this on linux/osx.
>>>> 
>>>>> 
>>>>> We have to say something. I can't agree the current words are enough.
>>>> 
>>>> What is wrong in the current wording? If the element is not visible in taskbar we cannot show something on top of it. We never specified when something is visible/non-visible in taskbar.
>>>> We can update the words, but it will be good to use current methods signatures.
>>>> 
>>>> 
>>>>> -phil.
>>>>> 
>>>>>> 
>>>>>>> 
>>>>>>>> 
>>>>>>>>> - Dialogs are always visible if they have no owner
>>>>>>>> 
>>>>>>>> This is implementation details on windows. I am not sure why it was implemented this way, is it really intended behavior or not.
>>>>>>> 
>>>>>>> So we'd have to say its platform and implementation dependent
>>>>>>> and add an implNote for audiences other than JCK.
>>>>>> 
>>>>>> Not sure that we can describe all the cases when it works.
>>>>>> 
>>>>>> To me the description like «Has no effect if this window is not visible in the task area.» is enough because it also covered the case when the window is invisible.
>>>>>> 
>>>>>>>> 
>>>>>>>>> - All other cases (eg Window) are platform dependent ?
>>>>>>>>>  Or can we say for the latter, other descendants of Window (including
>>>>>>>>>   Window itself) are never visible ?
>>>>>>>> 
>>>>>>>> 
>>>>>>> What about the last point ?
>>>>>> 
>>>>>> It depends from the implementation.
>>>>>> 
>>>>> 
>>>> 
>> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/awt-dev/attachments/20170315/a16891d8/attachment.html>