Hi Semyon, The fix looks good to me. I have only a minor remark regarding the test. It might be reasonable to use Util.testPixelColor() from regtesthelpers package instead of Robot.getPixelColor(). However it is up to you and you may keep it as is.
Thanks, Dmitry > On 31 Oct 2017, at 15:05, Semyon Sadetsky <semyon.sadet...@oracle.com> wrote: > > Thank you, Alexander. I have corrected the webrev. > > --Semyon > > On 10/31/2017 02:25 AM, Alexander Zvegintsev wrote: >> Hi Semyon, >> >> the fix looks good to me, but the test link is broken in this webrev. >> Thanks, >> Alexander. >> On 30/10/2017 20:37, Semyon Sadetsky wrote: >>> >>> Hello, >>> >>> Please review fix for JDK10: >>> >>> bug: https://bugs.openjdk.java.net/browse/JDK-8190230 >>> <https://bugs.openjdk.java.net/browse/JDK-8190230> >>> >>> webrev: http://cr.openjdk.java.net/~ssadetsky/8190230/webrev.00/ >>> <http://cr.openjdk.java.net/%7Essadetsky/8190230/webrev.00/> >>> >>> After the 8169589 fix siblings of the same parent are overlapped in the >>> order they appear in the array of child windows. But this order do not >>> preserve the previous overlapping order in which the sibling windows have >>> been shown before the order update. That is the cause of unexpected >>> reorder. >>> >>> In the fix each platform window is attributed with a timestamp of the >>> moment when it was show in front last time. Then this timestamp is used to >>> sort the sibling windows array before putting them into an order to each >>> other. >>> >>> --Semyon >>> >> >
