> > Some of them may duplicate existing QA or JUnit tests, so there are three > end points for a given jtreg test: drop it, convert it to QA, or convert it > to JUnit. In some cases "convert" may mean write a new test that fits in one > of the preferred frameworks and tests the same issues as the jtreg test. > > This is what I was aiming at, yes. This effort could also be done for many of the current QA tests, that dont really need the QA infrastructure. Those could be moved to JUnit (tests in the "id" and "config" categories seem good candidates).
> Any thoughts on how to organize this effort? A hundred separate JIRA issues > seems excessive. Maybe one JIRA and use comments to say which tests we are > working on? > One JIRA issue seems appropriate. But I believe there are some other JIRA issues open regarding jtreg already.
