Hello, Please review the updated fix: http://cr.openjdk.java.net/~pchelko/9/8041990/webrev.02/
After offline discussion with Sergey we've decided that an Accessor should not contain any logic, so it was moved to a separate method in InputMethodEvent. Thank you. With best regards. Petr. On 06 июня 2014 г., at 12:25, Alexander Scherbatiy <[email protected]> wrote: > > The fix looks good for me. > > Thanks, > Alexandr. > > On 5/28/2014 12:52 PM, Petr Pchelko wrote: >> Hello, >> >> Please review an updated fix: >> http://cr.openjdk.java.net/~pchelko/9/8041990/webrev.01/ >> >> I've removed null-checks in EventQueueAccessor, because these could mask >> bugs. >> All the rest is the same. >> >> With best regards. Petr. >> >> On 28 мая 2014 г., at 1:21, Phil Race <[email protected]> wrote: >> >>> Is there not some (bad) implication of returning 0 for the time ? >>> >>> -phil. >>> >>> On 5/27/14 8:45 AM, Petr Pchelko wrote: >>>> Hello, AWT Team. >>>> >>>> Please review a next AppContext issue: >>>> https://bugs.openjdk.java.net/browse/JDK-8041990 >>>> The fix is available at: >>>> http://cr.openjdk.java.net/~pchelko/9/8041990/webrev/ >>>> >>>> When non-latin characters a typed AWT tries to check if InputMethods are >>>> going to handle the key event. >>>> To do it we create an InputMethodEvent on Appkit thread which fails with >>>> NPE due to lack of AppContext. >>>> >>>> Thank you. >>>> With best regards. Petr. >
