I wasn't trying to catch the bug, I was trying to illustrate the bits that work already.
On Tue, Mar 29, 2011 at 3:18 PM, <pjul...@gmail.com> wrote: > > > http://gwt-code-reviews.appspot.com/1395802/diff/1/user/test/com/google/gwt/event/shared/ResettableEventBusTest.java > File user/test/com/google/gwt/event/shared/ResettableEventBusTest.java > (right): > > > http://gwt-code-reviews.appspot.com/1395802/diff/1/user/test/com/google/gwt/event/shared/ResettableEventBusTest.java#newcode83 > user/test/com/google/gwt/event/shared/ResettableEventBusTest.java:83: > narrowScope.removeHandlers(); > With respect, I think these tests miss the bug completely. > ResettableEventBus has "registrations" which is a set. > > If you're calling removeHandlers, you're not triggering the bug because > this blindly clears the "registrations" set. To trigger the bug, you > need to save the HandlerRegistration object that was returned to you. > If you call removeHandler() on the HandlerRegistration object, you'll > find that nesting doesn't work properly. > > > http://gwt-code-reviews.appspot.com/1395802/ > -- http://groups.google.com/group/Google-Web-Toolkit-Contributors
