On Mon, Jan 11, 2010 at 2:09 PM, Stephen Connolly < [email protected]> wrote:
> 2010/1/11 Kristian Rosenvold <[email protected]>: > > I think SUREFIRE-555 needs to be reverted. There is a problem in the log > > output when running the tests, and the output log does not properly > reflect > > the status of the running tests. While this would've affected only a > > microscopic number of users 7 months ago, the 4.7 range has gotten > > sufficiently mainstream that this i unacceptable today. junit 4.7 is > > "default" in spring 3. > > > > The issue applies only when the new provider is used, and that is only > for > > users of junit [4.7,). I will fix the problem, but I don't think it > should > > delay the release. > > > > Any respin will delay the release. > > We either go with what we have deployed to the staging repo today, or > we call a vote again (a.k.a. take 3) > > Are you saying you're -1 on 2.5 as is? > > My personal view is that there is nothing stopping us from doing 2.5.1 > tomorrow, i'd rather get 2.5 out today unless somebody is -1. > > Also, can you file a JIRA, we can always release note the logging > output being incorrect when using 4.7 and parallel threading (or is it > 4.7 no matter what) > > It's [4.7,+) no matter what, which is really the part that makes me uncomfortable. It /may/ affect CI system build result tracking. The provider detection is based on which junit artifact is present only. I now understand I probably should have stayed with the old provider when no concurrency options were present. As bugs come I find that this is an annoyance. A bit of an embarrasing one at that, but given a fairly quick 2.5.1 I think it can pass. I have tested the plugin and I'm a sort of embarrassed +1. But it hurts my professional pride. I'll make the issue as soon as we conclude this? Kristian
