Re: [OpenJDK 2D-Dev] RFR: 8196086: java/awt/image/DrawImage/IncorrectSourceOffset.java fails

2020-11-23 Thread Aleksey Shipilev
On Sun, 22 Nov 2020 07:36:12 GMT, Sergey Bylokhov wrote: > These tests draw some specific pattern to the VolatileImage and to the > BufferedImage, and then compare pixels. > > The test uses the getSnapshot() method to get the pixels from VolatileImage, > and this method produces some interpola

Re: [OpenJDK 2D-Dev] RFR: 8256321: Some "inactive" color profiles use the wrong profile class

2020-11-23 Thread Prasanta Sadhukhan
On Thu, 19 Nov 2020 16:51:00 GMT, Sergey Bylokhov wrote: >> It will be good to know where, in spec or any link, it is mentioned to use >> CLASS_ABSTRACT/CLASS_COLORSPACECONVERSION instead of CLASS_DISPLAY? > >> It will be good to know where, in spec or any link, it is mentioned to use >> CLASS_

Re: [OpenJDK 2D-Dev] RFR: 8196086: java/awt/image/DrawImage/IncorrectSourceOffset.java fails

2020-11-23 Thread Sergey Bylokhov
On Mon, 23 Nov 2020 09:48:07 GMT, Aleksey Shipilev wrote: > I have a drive-by question: why do some tests run with > `-Dsun.java2d.uiScale=3`, and some are not? The tests which use odd scale=3 are more tolerant to the artifacts caused by the interpolation when a bigger image is drawn to the sm