I would like to push the fixes for Windows into 0.11, as they are mostly small bug fixes. There's already an umbrella JIRA for these, https://issues.apache.org/jira/browse/PIG-2793
Alan. On Oct 15, 2012, at 1:26 PM, Rohini Palaniswamy wrote: > Me and Cheolsoo are kicking off a new run for the tests to check the > current failures. Will update the jira with those. Interested folks can > pick up the sub-task jiras. > > Should we create another umbrella ticket just for the windows test > failures? HW and Microsoft folks are putting in effort to fix them. Not > sure if the plan is to get them into 0.11 or make it part of trunk. Either > way another umbrella ticket for that would be good. > > Regards, > Rohini > > On Mon, Oct 15, 2012 at 12:17 PM, Cheolsoo Park <cheol...@cloudera.com>wrote: > >> Now we have the umbrella ticket for test failures (unit/e2e): >> https://issues.apache.org/jira/browse/PIG-2972 >> >> Please add new jiras as sub-tasks to it. >> >> Thanks, >> Cheolsoo >> >> On Mon, Oct 15, 2012 at 11:49 AM, Cheolsoo Park <cheol...@cloudera.com >>> wrote: >> >>>>> Thinking of creating a umbrella ticket to track all current unit >>> test failures and get them fixed. >>> >>> Once we open a umbrella ticket for test failures, everyone should link >>> their jiras that are for failing tests to it. I know that there are >> already >>> many open jiras for failing tests. I have several jiras myself. >>> >>> Please let me know if you want me to open the umbrella ticket and link >>> existing tickets to it. >>> >>> Thanks, >>> Cheolsoo >>> >>> >>> On Mon, Oct 15, 2012 at 11:36 AM, Cheolsoo Park <cheol...@cloudera.com >>> wrote: >>> >>>> This will be great! Please feel free to assign me bug fixes. >>>> >>>> >>>> On Mon, Oct 15, 2012 at 11:32 AM, Jonathan Coveney <jcove...@gmail.com >>> wrote: >>>> >>>>> I think this is an excellent idea, Rohini. We have resources to fix >> these >>>>> failures, so pointing them out should get them fixed. The error we've >> run >>>>> into is errors that are transient or related to a specific build on a >>>>> specific machine (and thus hard to replicate) vs. more serious errors. >>>>> Both >>>>> should be fixed, but it's hard to do the former when you have no >> control >>>>> over the CI. Throw in all the false fails from clover issues or >>>>> what-have-you and we're in a bad state. Cleaning that up would be >>>>> awesome. >>>>> >>>>> 2012/10/15 Julien Le Dem <jul...@twitter.com> >>>>> >>>>>> I think this is a great idea. >>>>>> I don't think we have access to the build right now. >>>>>> Daniel seemed to know more about that. >>>>>> Julien >>>>>> >>>>>> On Mon, Oct 15, 2012 at 11:15 AM, Rohini Palaniswamy < >>>>>> rohini.adi...@gmail.com> wrote: >>>>>> >>>>>>> Hi all, >>>>>>> Quite a few unit tests are broken in trunk/0.11. Now that 0.11 >> is >>>>>>> branched it is more important to have them fixed and pass for >>>>> stability. >>>>>>> Also we need to figure out some way to have tests not broken. But >>>>> with >>>>>> the >>>>>>> time unit tests take it is difficult for everyone to run it. I >> think >>>>> the >>>>>>> first thing to do would be to have the Apache jenkins build fixed >> and >>>>>> treat >>>>>>> build failures there seriously. I can help out with that if someone >>>>> can >>>>>>> guide me with access, modifying jenkins scripts, etc. Any other >>>>> ideas? >>>>>>> Thinking of creating a umbrella ticket to track all current unit >>>>> test >>>>>>> failures and get them fixed. Or is there some other process that we >>>>>> usually >>>>>>> follow? >>>>>>> >>>>>>> Thanks, >>>>>>> Rohini >>>>>>> >>>>>> >>>>> >>>> >>>> >>> >>
