On Tue, 5 Sep 2006, Matt Benson <[EMAIL PROTECTED]> wrote: > Antunit depends on HEAD. So theoretically antunit > can't be used in the wild until 1.7 is released (let's > assume "in the wild" means released components only). > Do we want to play by these rules ourselves and wait > to integrate antunit tests into Ant core until after > 1.7's release, or can we go ahead and start > adding/migrating au tests?
1.7beta would be enough for me, but we'd also need an AntUnit 1.0beta IMHO. I'm going to post a release plan for it soon, but am still waiting to get the .NET Antlib out of the door first (the vote on the actual ZIP to release seems to be failing and I'll need a fresh start, though). > Secondly, whenever we do integrate antunit tests into > core, a simplistic approach would be to add a new > directory, e.g. src/au-testcases whose structure might > mirror src/etc/testcases, and invoke the antunit task > from the core build in one way or another (unimportant > ATM IMO). For the antlibs I renamed src/testcases to src/tests/junit and added a src/tests/antunit directory. The common build file autodetects whether either directory is present and runs <junit> and/or <antunit> if the corresponding directories exists. I didn't think of the *-support.xml case, that would need to be added. > Finally, I think it will be best for antunit if as > many as possible of our existing junit testcases are > (eventually) ported to antunit. This should expose > any missing generic functionality IMHO. Agreed. Stefan --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
