01:34 <travis-ci> jruby/jruby (master:9605d68 by Charles Oliver Nutter): The build was fixed. https://travis-ci.com/jruby/jruby/builds/209206555 [224 min 22 sec]

01:35 <headius[m]> 👍

15:46 <enebo[m]> kares: I plan on 61.0 and 52/60 releases today

15:46 <enebo[m]> and rdubya :)

15:48 <enebo[m]> rdubya: The 52.x and 60.x is for https://github.com/jruby/activerecord-jdbc-adapter/pull/1066

15:48 <rdubya[m]> 🙂 I threw a comment in that ticket but mssql inserts appear to be broken back to 50 if composite primary keys are used

15:48 <rdubya[m]> so was hoping that could be merged up through

15:49 <enebo[m]> rdubya: ok. In the one where I just commented

15:50 <rdubya[m]> I haven't seen a comment come through, or was it a different pr?

15:50 <enebo[m]> I don't know I am working on hearsay :)

15:51 <enebo[m]> rdubya: but there is something which needs to go back to 50.x forward that helps mssql but not anything else?

15:51 <rdubya[m]> yeah

15:51 <rdubya[m]> i pointed it at 50-stable but was wondering if i should have pointed it higher

15:52 <rdubya[m]> https://github.com/jruby/activerecord-jdbc-adapter/pull/1079

15:52 <enebo[m]> rdubya: I don't remember was 5.0 the first targetted version for mssql?

15:52 <rdubya[m]> yeah

15:52 <rdubya[m]> right now it supports 50 and 51

15:52 <rdubya[m]> maybe i'll get my act together and get 52 out someday ☹️

15:53 <enebo[m]> rdubya: ok no problem and this has no risk of affecting any other adapters

15:53 <enebo[m]> but it does require everything else get released

15:53 <enebo[m]> it is just how arjdbc works

15:53 <rdubya[m]> yeah unfortunatly

15:53 <enebo[m]> rdubya: ok 50 forward then

15:53 <rdubya[m]> thanks

15:55 <enebo[m]> rdubya: so we get no test results from this so I will take it that there is nothing else needed at this point and we can just merge and release

15:55 <rdubya[m]> yeah, it should be good

15:56 <enebo[m]> coolio...thanks

16:01 <enebo[m]> If I remember right the main reason we have this single common dep is that the java itself is a common base class for multiple adapters so we cannot just replicate that code if it contains the same package name

16:03 <enebo[m]> and because of modules we also would need to make a 2 modules if we wanted to extend the adapter specific java into the specific gem which would use it.

16:04 <enebo[m]> This last PR if mssql was shipped in the same gem as mssql would not have required these extra releases

16:04 <enebo[m]> I am not saying I want to do this work today but us arjdbc hackers should think about unwrapping at least this issue so db-specific fixes do not force all adapters to be released.

16:15 <rdubya[m]> yeah, that would be great to break that dependency

16:20 <headius[m]> oh for the love of

16:20 <headius[m]> spent probably six hours on this spawn thing and it turns out it is opening a redirect file for read

16:20 <headius[m]> so nothing writes but since errno is redirected I never see it

16:20 <headius[m]> oy vey

17:15 <headius[m]> ok I think I have just about fixed everything I can in spawn

17:28 <headius[m]> enebo: there is one marked for linux you might be able to confirm

17:28 <headius[m]> no idea what that is about

17:28 <headius[m]> probably broken due to posix_spawn not supporting chdir and us working around it, but give that a shot when you can

17:29 <enebo[m]> headius: ok

17:54 <headius[m]> enebo: spawn fixes merged to master

17:55 <headius[m]> enebo: fix printf

17:56 <headius[m]> printf missing features account for probably 60-70 tags in core

19:02 <travis-ci> jruby/jruby (purge:30de9db by Charles Oliver Nutter): The build was canceled. https://travis-ci.com/jruby/jruby/builds/209434353 [62 min 33 sec]

19:26 <enebo[m]> kares: if you are around...jrake build:adapters then jrake release:do right? After I commit new version number.

19:46 <enebo[m]> kares: disregard that question