Community
Participate
Working Groups
According to the wiki [1] repository-specific duplicate detector should have kind attribute specified. However duplicate detector org.eclipse.mylyn.internal.bugzilla.ide.StackTraceDuplicateDetector which is using BugzillaRepositoryQuery internally is missing this attribute. [1] http://wiki.eclipse.org/Mylyn_Integrator_Reference#Duplicate_Detectors
Rob: I'm not sure that Meghan is doing any work on duplicate detectors, so this should happen after your other duplicate detector clean-up.
Since bug 197181 is more involved (and it doesn't seem like it having P1 or even P2 priority), I would recommend to fix this one first and then refactor bugzilla detector once connector-neutral search facility is in place.
Fixed.
Created attachment 74896 [details] mylyn/context/zip
Rob, just add kind to plugin.xml isn't enough. You need to filter detectors by connector type in bugzilla task editors, otherwise they would show other detectors.
(In reply to comment #5) > Rob, just add kind to plugin.xml isn't enough. You need to filter detectors by > connector type in bugzilla task editors, otherwise they would show other > detectors. Thanks. Done.
Still doesn't seem right. I think it should be: detector.getKind() == null || detector.getKind().equals(getConnector().getConnectorKind())
Thanks again. I had the wrong semantics in mind for getKind().
You are welcome. :-) I am just interpreting Mik's description from the wiki [1], even so that description isn't really accurate. See my comment on the wiki. [1] http://wiki.eclipse.org/Mylyn_Integrator_Reference#Duplicate_Detectors