| Summary: | Session test ordering off on java 7 | ||
|---|---|---|---|
| Product: | [Eclipse Project] Platform | Reporter: | Paul Webster <pwebster> |
| Component: | UI | Assignee: | Paul Webster <pwebster> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | daniel_megert, ob1.eclipse, remy.suen |
| Version: | 3.8 | ||
| Target Milestone: | 4.2 M5 | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| Whiteboard: | |||
|
Description
Paul Webster
I've released a session test ordering fix to origin/pwebster/bug369526 Remy, Oleg, could you have a look in the context of 3.8 testing and make sure the tests still run? Our current failure list: http://download.eclipse.org/eclipse/downloads/drops/I20120123-1300/testresults/html/org.eclipse.ui.tests_win32.win32.x86_7.0.html#SessionTests PW Looks okay to me. The failures I get now are from the ones that haven't been restructured. Makes sense. There are two more session tests that fail om my computer if compiled with Java 1.6 and run with Java 1.7: NonRestorablePropertySheetTest (actual order on my machine: test02SecondOpening run before test01ActivateView) ArbitraryPropertiesEditorTest (actual order om my machine: testPartInstantiation, testSecondOpening, testOpenEditor) That was using Eclipse 3.8. Dean, could you please set yourself up for 3.8 and go through the org.eclipse.ui.tests.session.SessionTests. The only way to guarantee the correct ordering at runtime is to have each test class return a suite(), and spell out the test order in that suite (see org.eclipse.ui.tests.session.Bug95357Test.suite() for an example). PW |