<AWT Dev> [9] Review request for 8069361 SunGraphics2D.getDefaultTransform() does not include scale factor

Sergey Bylokhov Sergey.Bylokhov at oracle.com
Tue Apr 21 13:43:31 UTC 2015


Hi, Alexander.
The fix looks good.

On 20.04.15 14:48, Alexander Scherbatiy wrote:
>
> Could you review the updated fix:
>    http://cr.openjdk.java.net/~alexsch/8069361/webrev.01/
>
>   - CGraphicsConfig.getDefaultTransform() is updated to use 
> AffineTransform.getScaleInstance()
>   - SG2D.getDefaultTransform() is updated to not check 
> GraphicsConfiguration.getDefaultTransform() on null
>   - the test is updated to compare SG2D transform with 
> GraphicsConfiguration transform on all graphics configurations
>
>  Thanks,
>  Alexandr.
>
>
> On 4/17/2015 4:28 PM, Sergey Bylokhov wrote:
>> Hi, Alexander.
>> I assume that the code in SG2D.getTransform/setTransform is the same 
>> as was before the fix of 8000629.
>> Code in SG2D.getDefaultTransform can be simplified, id do not think 
>> that GraphicsConfiguration.getDefaultTransform. and 
>> SG2D.getDeviceConfiguration can return null for NullSurfaceData only.
>> The test should check that default transform of graphics 
>> configuration(all screens should be checked) is the same as a 
>> transform of the Graphics.
>> The method CGraphicsConfig.getDefaultTransform() can use 
>> getScaleInstance but it is up to you.
>>
>> On 14.04.15 17:51, Alexander Scherbatiy wrote:
>>> Hello,
>>>
>>> Could you review the fix:
>>>   bug: https://bugs.openjdk.java.net/browse/JDK-8069361
>>>   webrev: http://cr.openjdk.java.net/~alexsch/8069361/webrev.00
>>>
>>>   SunGraphics2D.getDefaultTransform() now returns transform which 
>>> includes GraphicsConfiguration default transform.
>>>
>>>   This can break a compatibility for applications which assumes that 
>>> default Graphics2D transform is always identity
>>>   and restore it as sg2d.setTransform(new AffineTransform()). 
>>> However, this is not now true for HiDPI displays.
>>>
>>> Thanks,
>>> Alexandr.
>>>
>>
>>
>


-- 
Best regards, Sergey.



More information about the awt-dev mailing list