Re: Windows OS Compatibility and Jenkins Postcommit

2017-06-09 Thread Lukasz Cwik
I think we will have to ask infra to install python+virtualenv on the windows machines or figure out how to install python/virtualenv using maven or something. On Fri, Jun 9, 2017 at 4:22 PM, Ahmet Altay wrote: > Thank you Luke, this is great! I hope that we will see a

Re: Windows OS Compatibility and Jenkins Postcommit

2017-06-09 Thread Ahmet Altay
Thank you Luke, this is great! I hope that we will see a mac version of this at some point. It looks like there is no python binary and python tests are failing for that. Ahmet On Thu, Jun 8, 2017 at 3:55 PM, Flavio Fiszman wrote: > Thanks Luke! I'm working on

Re: Java Cross-JDK Test Available on Jenkins Postcommit!

2017-06-09 Thread Mark Liu
Good catch! Actually beam_PreCommit_Java_MavenInstall and beam_PostCommit_Java_MavenInstall is running on JDK 1.8. I think it's reasonable to

Re: Java Cross-JDK Test Available on Jenkins Postcommit!

2017-06-09 Thread Ted Yu
This is nice. Nit: I don't see JDK 1.8 (Oracle) variant. Is this intentional ? Thanks On Fri, Jun 9, 2017 at 3:49 PM, Mark Liu wrote: > Hi all, > > I worked on BEAM-1544 > for > a while, which is to bring Java

Re: [DISCUSS] GitBox

2017-06-09 Thread Kenneth Knowles
On Fri, Jun 9, 2017 at 9:55 AM, Davor Bonaci wrote: > A few of us have been looking into this for a while, and, if I understand > correctly, a concrete proposal with several variants is coming up shortly > ;-) > > On a high level, there are a few pros and cons: a lack of submit

[DISCUSS] GitBox

2017-06-09 Thread Kenneth Knowles
Hi all, There have been a couple threads on other Apache dev lists [1] [2] discussing ASF GitBox [3]. I don't think we've discussed it here, for Beam. I think labeling and assigning PRs would be very helpful. Thoughts? Kenn [1]

Re: Dynamic file-based sinks

2017-06-09 Thread Reuven Lax
I'm putting together a proof-of-concept PR for option 1 to see how it looks. On Thu, Jun 8, 2017 at 4:07 PM, Reuven Lax wrote: > After looking at everyone's comments, I think option 1 is the better > approach - map destinations to a FilenamePolicy. It is a good parallel to >