Having a couple of build failures is not a problem when Jenkins runs because it is configured to run through all tests regardless of any failures. So while there are a handful of broken tests at any given time, the other 8 or 9 thousand tests still run consistently. There is currently one real OSGi failure right now but most of the failures are just tests that need to be fixed because they are intermittently failing. Also, the PR I merged was a new JNDI related class that had its own test. No where else was that class instantiated so no other test would fail anyways.
On Mon, Oct 31, 2016 at 9:12 PM, Hadrian Zbarcea <hzbar...@gmail.com> wrote: > Started to look into old PRs, some should not be merged. I'll see if I > could sort it out with infra to get them closed without a commit, although > imho, a commit like "This closes #xxx. Intentionally not merged" is better, > as it would show that the close without merge was intentional. > > Another issue is that the build system looks like is in need of a bit of > love, some outdated and inconsistent dependencies for plugins. I fixed a > few, a few more that need a closer attentions. I saw that cshannon merged a > PR, I looked at it too, but I am not sure how it was tested as many test > fail intermittently, some just fail and jenkins didn't have a clean build > in a while. I'll look into those too. > > Cheers, > Hadrian > > > On 10/23/2016 02:11 AM, Jean-Baptiste Onofré wrote: > >> Good idea Hadrian. >> >> I will try to take a look. >> >> Regards >> JB >> >> >> >> On Oct 23, 2016, 03:48, at 03:48, Hadrian Zbarcea <hzbar...@gmail.com> >> wrote: >> >>> There are a number of old, even very old, PRs open [1]. Some I suspect >>> are not relevant anymore. It'd be great if somebody could look into >>> them >>> and close them down. I will try to find some time next week and help >>> with that as well. >>> >>> Cheers, >>> Hadrian >>> >>> [1] https://github.com/apache/activemq/pulls >>> >> >>