Brett Porter wrote:
I'm just catching up again. I am planning to scan through all the commits and see if anything catches my eye. I also saw an issue reported on the users list that I'll drop into JIRA about the new code working with 2.0.6 - which versions of Maven have you been testing with? (I've tended to test with 2.0.4 and 2.0.6, but now 2.0.7 and 2.0.8 made sometimes funky changes that are worth running the integration tests with).
I've been pretty much exclusively testing with 2.0.7. I just ran the 2.4 integration tests with 2.0.6 and 2.0.8 with no problems.
maven-surefire-plugin lists 2.0.6 as a prerequisite, as of revision 598160. "mkleint: plexus-utils version 1.4.x means the plugins requires 2.0.6 at least". Not entirely certain why that is, but there you have it.
WDYT about continuing with the 2.3.1 release immediately, and giving 2.4 a short time more to simmer? If for any reason 2.4 does break certain builds, there is still a better-than-2.3 fallback. You can also use the release ann. to re-encourage testing of 2.4.
+0 on letting 2.4 simmer; -0 on releasing 2.3.1 now. I think I wouldn't want to let 2.4 simmer for more than two weeks, at the maximum.
What concerns me about 2.3.1 is that it doesn't have integration tests like 2.4, so it's hard to know for sure whether it really *is* better than 2.3 or whether new bugs have creeped in. (Surefire has a lot of corner cases and gotchas.)
On the other hand, I feel pretty good about 2.4-SNAPSHOT because it has been getting a lot of incidental testing through users who want/need TestNG support. TestNG recommends using 2.4-SNAPSHOT on their website; pretty much everyone using TestNG and Maven has been using 2.4-SNAPSHOT all along.
If Surefire didn't automatically grab the latest release, I'd cheerfully agree to release both and let users sort out what works best for them, but as it stands I'd prefer to destabilize Surefire infrequently and put all of our testing eggs in one basket.
With that said, if there's a need for 2.3.1 [Mauro suggested he'd like it last week :-)] then I certainly won't stand in its way. I'd strongly suggest that whoever wanted to try this should run the 2.4 integration tests against 2.3 and then again against 2.3.1 to identify regressions. Seems like a lot of trouble for not much benefit, IMO.
-Dan
