I'm feeling rather obstinate. I talk about changes on perl-qa. I post release announcements to here, p5p and module-authors for a reason. All those modules that this change broke and not a single one of them tried their module with the Test-Simple alphas and reported the problem. Because of that there's a big mess to clean up and its landed right in my lap. This makes me angry.
No, I'm not taking 0.61 off CPAN. No, I'm not rolling back to 0.60. There's too many fixes and changes and it'll cause nearly as much disruption to roll them back as to leave it there. I don't want to trade bad with slightly less bad. I leave this as an absolute last resort. I understand the scope of the damage and I don't want to spend another minute talking about how big this pile of shit is. Its big. All I care about is what's directly effected. Unless I missed something, it all radiates out from Test::Builder::Tester. So, that's what I'm not gonna do. This is what I am gonna do. I'm gonna get Test::Builder::Tester fixed. Because even if its doing things which are Evil and Wrong nobody really made a fuss over it until now, including me. I'm looking for ways to fix Test::Builder::Tester and keep existing test_diag() code working (at least for a while). I'm willing to temporarily put the formatting back the way it was. This will give some breathing room. I'm not going to be the one to write that patch as its a large amount of monkey work which I won't do again. Patches welcome. A good place to start might be to reverse the patch which changed the diagnostics. svn diff -r2401:2402 http://svn.schwern.org/svn/CPAN/Test-Simple/trunk I'm absorbing Test::Builder::Tester into the Test-Simple distribution. This kills three birds with one stone: - Anyone who updates Test::More gets a fixed Test::Builder::Tester which should solve most the current problem, unless you used test_diag() in which case you're just hosed. - It guarantees Test::Builder and Test::Builder::Tester will remain in sync. - I can use it to test Test::More whose tests are pretty damn primitive. I haven't been able to contact Mark Fowler and I haven't seen any activity from him on RT, so I'm just going to go ahead and do it. Once he resurfaces we can work out if this is going to be a permanent arrangement. I don't have a long term solution for users of test_diag(). I'm entertaining ideas. "Don't change the failure output" is not one of them. One temporary hack is to parse the test_diag() input, look for attempts to match the old Test::More diagnostics and translate it into the new format/regex/whatever. Since the old format is very regular this should be fairly straightforward. Above all, I want patches. I want fixes. I want code. I haven't seen a single patch in all this talk. -- Michael G Schwern [EMAIL PROTECTED] http://www.pobox.com/~schwern <Schwern> What we learned was if you get confused, grab someone and swing them around a few times -- Life's lessons from square dancing
