[17:01:43] <asdfasdf_>	 hi, can someone from WMF/mobile please see https://en.wikipedia.org/wiki/Wikipedia:Help_desk#Help_with_Android_app
[17:02:17] <asdfasdf_>	 says they wrote to you 3 weeks ago, no reply. Anyway. That's all, cheers
[17:03:45] <asdfasdf_>	 maybe I should say dbrant  too. Worth a try. ^
[17:11:08] <niedzielski>	 asdfasdf_: o/ I'm not sure who's listening to the email account referenced on that page so I've updated it to the latest.
[17:20:15] <dbrant>	 (I answered on Help Desk)
[17:45:57] <mdholloway>	 dbrant: niedzielski: it looks like tapping the thumb is still (improperly) starting the action mode as of https://gerrit.wikimedia.org/r/#/c/342057/ though i'm not quite sure how that could be
[17:47:04] <niedzielski>	 mdholloway: it seems to be working ok for me?
[17:47:26] <niedzielski>	 i long press any item to start action mode and tapping anywhere within that item changes the selection state
[17:47:27] <mdholloway>	 let me rebuild and try again, maybe i'm using an older build than i think i am
[17:48:14] <niedzielski>	 mdholloway: i'm running on the topmost patch i guess, ceee00e
[17:48:16] <mdholloway>	 niedzielski: dbrant: never mind, it's working correctly
[17:48:32] <mdholloway>	 i was still in the action mode with nothing selected
[17:48:36] <niedzielski>	 👍
[17:48:53] <mdholloway>	 it might be that we want to end the action mode when the last selected item is un-selected, but that would be a different patch
[17:49:01] <niedzielski>	 mdholloway: so you don't see that android studio gradle issue on the other patch? what version are you running?
[17:49:01] <dbrant>	 yep, the final patch makes it so that unselecting the last item ends the actionmode.
[17:49:17] <mdholloway>	 nice :)
[17:49:46] <mdholloway>	 niedzielski: Android Studio 2.3, Build #AI-162.3764568, built on February 24, 2017
[17:50:29] <niedzielski>	 mdholloway: ok, i'm on 2.2.3. i tried to do the upgrade several times yesterday and today but it keeps failing in an odd way. i guess i'll have to try a fresh install
[17:55:10] <mdholloway>	 niedzielski: also, while i'm thinking of it, having the gradle plugin handling SDK installation probably means we can remove the install-android-sdk step from our testing jobs in integration/config, right?
[17:55:24] <mdholloway>	 at least for our routine CI tests, not sure about the periodic job since that takes an emulator
[17:55:57] <mdholloway>	 though that job too is now getting the gradle-plugin-installed SDK...
[17:56:01] <niedzielski>	 mdholloway: yeah i'm not sure about periodic but i really hope for the other jobs!
[17:56:52] <niedzielski>	 mdholloway: if we have trouble, it's just unsetting that ANDROID_HOME variable i think and allowing the jenkins plugin use it's own downloader
[17:57:02] <mdholloway>	 i'll put in a patch in a bit, after giving the periodic job a chance to run
[17:57:06] <mdholloway>	 (if it hasn't already)
[17:58:24] <niedzielski>	 mdholloway: you should be able to trigger builds of the periodic job by logging in
[17:58:31] <niedzielski>	 integration.wikimedia.org/ci/job/apps-android-wikipedia-periodic-test/
[18:07:19] <mdholloway>	 niedzielski: looks like the periodic job ignores $ANDROID_HOME and does its own thing (i.e., still uses the SDK at /srv/jenkins-workspace/tools/android-sdk)
[18:07:29] <mdholloway>	 https://integration.wikimedia.org/ci/job/apps-android-wikipedia-periodic-test/1475/console
[18:07:42] <mdholloway>	 i think we can safely remove the install-android-sdk step from the others though
[18:09:14] <niedzielski>	 mdholloway: hm, sounds good. it might be worth trying to remove the install sdk step on the periodic build too if you haven't already
[18:10:08] <niedzielski>	 mdholloway: there are many broken periodic tests on master btw
[20:15:37] <mdholloway>	 interesting.  jake archibald has an offline wikipedia project: https://github.com/jakearchibald/offline-wikipedia
[20:15:44] <mdholloway>	 i haven't really looked at it yet
[20:44:09] <bearND>	 mdholloway: Nice. He's also got a Udacity course about Service Workers
[20:44:45] <mdholloway>	 bearND: oh, i'll check it out!
[20:45:09] <mdholloway>	 bearND: i left a question for you on https://gerrit.wikimedia.org/r/#/c/331824/ btw, not sure if you're still monitoring android patches
[20:45:49] <bearND>	 mdholloway: ah, I'm not monitoring those anymore. Thanks for the ping
[21:35:03] <niedzielski>	 mdholloway dbrant: not a problem to solve today, but Android Studio doesn't seem to be able to run a number of plain JUnit tests on master such as ZeroConfigClientTest (they work fine from command line). This is probably an issue introduced with the recent Robolectric or Gradle plugin changes. Just a heads up!
[21:35:18] <dbrant>	 :/
[21:36:12] <hashar>	 mdholloway: niedzielski:  Hello! There a couple jenkins jobs change, I am wondering whether you have deployed them
[21:36:18] <hashar>	 if so, I guess I can just +2 ? :}
[21:36:47] <hashar>	 eg https://gerrit.wikimedia.org/r/#/c/342263/
[21:37:04] <niedzielski>	 hashar: I haven't deployed anything today :]
[21:37:23] <hashar>	 ahhh ;}
[21:37:38] <mdholloway>	 hashar: i think those are ready to +2 -- the changes they depend on have been made
[21:37:48] <mdholloway>	 nothing's been deployed per se though
[21:37:50] <hashar>	 guess I will do them on monday
[21:38:13] <hashar>	 refresh the jobs and rerun them then write some status on the task
[21:38:28] <hashar>	 looks like you would get some more freedom by having gradle to handle the SDK installation
[21:39:23] <mdholloway>	 yeah, that way seems both more reliable and easier to maintain
[21:40:31] <mdholloway>	 hashar: i just added you as a reviewer to another change (in puppet) specifically to set the environment variable for the periodic job
[21:40:44] <mdholloway>	 setting it in the repo scripts as we do for the others doesn't work for that oen
[21:40:45] <mdholloway>	 one
[21:41:30] <mdholloway>	 (the jenkins plugin needs the sdk location before that script runs)
[21:42:51] <hashar>	 sounds good
[21:43:16] <hashar>	 will review / test all of that monday.  Now is not a good time for me to babysit or think about all of that
[21:43:29] <mdholloway>	 ok, no rush :)
[21:43:37] <hashar>	 thanks for taking care of those jobs!
[21:44:08] <mdholloway>	 hashar: you're welcome, thanks for checking in!
[21:55:47] <hashar>	 \O/
[22:25:04] <dbrant>	 mdholloway: is it in fact the same crash? (can you paste the stack trace in the PS?)
[22:25:25] <mdholloway>	 dbrant: it is -- i'll paste both in