RFR(S) : 8240904 : Screen flashes on test failures when running tests from make

[Igor Ignatyev]

Hi Erik,

sure, I've actually replaced one long statement w/ multiple shorter ones, which made the comment section redundant -- http://cr.openjdk.java.net/~iignatyev//8240904/webrev.01 <http://cr.openjdk.java.net/~iignatyev//8240904/webrev.01>

Thanks,
-- Igor

> On Apr 16, 2020, at 6:17 AM, Erik Joelsson <erik.joelsson at oracle.com> wrote:
> 
> Looks ok to me. Would it be possible to break up the long lines a bit to improve readability? Backslash escape for newlines should work in properties files.
> 
> /Erik
> 
> On 2020-04-15 22:22, Igor Ignatyev wrote:
>> http://cr.openjdk.java.net/~iignatyev//8240904/webrev.00
>>> 35 lines changed: 26 ins; 0 del; 10 mod
>> 
>> Hi all,
>> 
>> 8233827[1] which added screenshots to so-called failure handler had an unexpected side-effect on linux, where users might observer flashes each time a screenshot is taken, which, to put it mildly, is annoying. the patch replaces gnome-screenshot app w/ calling java.awt API to make a screenshot. the patch also uses the same solution to make screenshots on windows and solaris (which previously didn't save screenshots on failures). we still use native app on mac as using java.awt API requires accessibility permissions, which might be as annoying as flashes on linux.
>> 
>> [1] https://bugs.openjdk.java.net/browse/JDK-8233827
>> 
>> JBS: https://bugs.openjdk.java.net/browse/JDK-8240904
>> webrev: http://cr.openjdk.java.net/~iignatyev//8240904/webrev.00
>> testing: verified that screenshots are successfully generated on headful systems, and headless systems ignore errors ('java.awt.AWTException: headless environment' from j.a.Robot::<init>).
>> 
>> Thanks,
>> -- Igor
>> 
>>