[00:00:35] <K4-713>	 awight: haaa. ah. Ow.
[00:01:54] <atgo>	 hey again ejegg
[00:02:18] <ejegg>	 hi atgo
[00:03:14] <atgo>	 i was going to ask you about the status 800's not getting into civi task
[00:03:15] <awight>	 AndyRussG: that's odd.  cos Krinkle was mentioning to me that we should move towards including all the logic to run tests, in the repo under test.  But this grunt-karma-qunit thing has gone really far in the opposite direction...
[00:03:25] <atgo>	 but now i read it a bit more thoroughly and it looks like there are still open questions.
[00:03:47] <Krinkle>	 awight: It looks that way but it isn't.
[00:03:56] <Krinkle>	 awight: It's all in mediawiki-core/Grunt, not in Jenkins
[00:04:11] <atgo>	 so... basically nevermind. let me know if it feels like a sand trap though and we can reconsider if we want to keep digging at it
[00:04:15] <Krinkle>	 Just like phpunit has been for years, mediawiki-core.
[00:04:39] <awight>	 Krinkle: ok, I'm looking for the new howto, for running qunit from the CLI... thanks!
[00:04:41] <AndyRussG>	 Krinkle: hi!
[00:05:22] <Krinkle>	 Of course, for testing code that does not depend on mediawiki, the logic can just be in the repo. But for mediawiki extensions specifically, no logic has to be duplicated in each extension repo. You install the extension and run mediawiki's qunit test.
[00:05:40] <Krinkle>	 So for qunit the "repo" is mediawiki-core. Not the extension repo. That's all :)
[00:06:06] <ejegg>	 atgo: they're definitely worth investigating!
[00:06:17] <Krinkle>	 Similar to how one registers phpunit test suites and then run phpunit.php in mediawiki core tests/phpunit
[00:06:24] <awight>	 Krinkle: also I am compelled to mention that mediawiki-core/phpunit is not a success story!  That thing has been broken for so long...
[00:06:29] <Krinkle>	 Let me know if you need any help registering the qunit tests. Although I think you've got that already.
[00:06:31] <atgo>	 yeah sounds good. I just saw your comment about the one that you solved and then actually read the rest and saw it wasn't totally solved :)
[00:07:00] <Krinkle>	 awight: Broken in what way? (btw, I'd be first to say it needs re-doing. It was this way when I got here, but it's on the list.)
[00:07:12] <awight>	 Krinkle: Your narrative sounds good, though :)  It definitely makes sense to reuse the testing logic.
[00:07:17] <ejegg>	 Yeah, when I try to replicate that failure at 600 status locally, everything goes boom
[00:07:32] <awight>	 Krinkle: PHPUnit is a huge nightmare on most fresh installs, in my experience.
[00:07:37] <ejegg>	 So i'm not even sure how prod kept going and logged anything after that
[00:07:42] <awight>	 With most projects, you stroll to the top level and "phpunit".
[00:08:08] <Krinkle>	 awight: If your code is independent and does not need mediawiki core classes or anything. Then you can totally run phpunit within the repo directly.  See https://github.com/wikimedia/cdb and https://github.com/cssjanus/php-cssjanus for examples.
[00:08:11] <awight>	 For MediaWiki core, it's this arcane commandline, and it doesn't work out of the box.  Requires root privs, then 99% of the time it fails cos of that stupid PHAR thing.
[00:08:35] <ejegg>	 wait, what about phpunit requires root privs?
[00:08:45] <awight>	 Krinkle: well, yeah.  But IMO it should do the right thing and run the entire test suite from the top level, including any enabled extensions.
[00:08:46] <Krinkle>	 And same for qunit. See oojs, and oojs-ui and visualeditor repos for a local Grunt qunit/karma example. pure 'npm install && npm test'
[00:09:03] <awight>	 ejegg: it wants to install phpunit... in a way that is not compatible with what it expects to execute.
[00:09:22] <ejegg>	 oh man, I must have erased that part from my memory
[00:09:31] <Krinkle>	 awight: phpunit itself being the nightmare or our use of it, or a specific distributions' package? Not sure I understand what you mean.
[00:09:37] <ejegg>	 running phpunit via make is also a little funky
[00:09:43] <awight>	 Krinkle: sorry, I'm not *trying* to start a flame war about the test harnesses, but... they really don't work well for me.
[00:09:45] <Krinkle>	 Run it via composer :)
[00:09:49] <ejegg>	 heh
[00:09:50] <AndyRussG>	 Krinkle: just getting back to this finally, BTW... I feel like the quesitons I'm gonna ask are probably be dumb and discoverable in the JJB config, but... so u said before, "Use npm install && grunt qunit from the MediaWiki core directory to run QUnit tests from the command-line in you local Chrome browser." ... tried that but it didn't work....
[00:10:30] <Krinkle>	 AndyRussG: OK. Let's debug. What doesn't work?
[00:10:40] <awight>	 Krinkle: let me back out further... I'd like to help fix the phpunit situation, some day :D
[00:10:42] <AndyRussG>	 awight: root privs needed to run phpunit? huh don't recall that
[00:10:45] <Krinkle>	 It's important to me that this is documented well enough to not need my intervention.
[00:11:03] <awight>	 AndyRussG: no, this is for the phpunit.php script to install PHPUnit on a fresh deployment.
[00:11:06] <Krinkle>	 So by all means ask anything you like :)
[00:11:21] <awight>	 Krinkle: yes, the immediate issue is that we'd like to run qunit from the CLI
[00:11:31] <AndyRussG>	 $ grunt qunit
[00:11:31] <AndyRussG>	 Running "karma:main" (karma) task
[00:11:31] <AndyRussG>	 Warning: Parameter 'url' must be a string, not number Use --force to continue.
[00:11:31] <AndyRussG>	 Aborted due to warnings
[00:11:49] <grrrit-wm>	 (CR) Ejegg: "The code looks good! whitespace is inconsistent though." (1 comment) [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189136 (https://phabricator.wikimedia.org/T86251) (owner: Awight)
[00:12:03] <Krinkle>	 This script: https://github.com/wikimedia/mediawiki/blob/master/tests/phpunit/install-phpunit.sh is pretty horrible yeah. It won't be neccecary once we fetch phpunit via composer-install as devDependency.
[00:12:06] <AndyRussG>	 awight: ah hmmm yeah I think I did have to install qunit somehow
[00:12:28] <AndyRussG>	 awight: sorry I meant phpunit
[00:12:35] <Krinkle>	 AndyRussG: Latest mediawiki-core? node --version, npm --version, grunt --version
[00:12:47] <AndyRussG>	 Krinkle: yes updated mwcore for this...
[00:13:08] <AndyRussG>	 $ node --version
[00:13:08] <AndyRussG>	 v0.10.29
[00:13:15] <Krinkle>	 AndyRussG: Also be sure to have MW_SERVER and MW_SCRIPT_PATH set in your bash env.
[00:13:31] <Krinkle>	 Unfortunately npm does not install MediaWiki. That's a prerequisite.
[00:13:32] <AndyRussG>	 $ grunt --version
[00:13:32] <AndyRussG>	 grunt-cli v0.1.13
[00:13:32] <AndyRussG>	 grunt v0.4.5
[00:14:05] <AndyRussG>	 Krinkle: hmmm that sounds like something
[00:14:38] <AndyRussG>	 Krinkle: yea thos are not set
[00:14:44] <Krinkle>	 e.g. export MW_SERVER='http://localhost' MW_SCRIPT_PATH='/w'
[00:15:41] <AndyRussG>	 Krinkle: the URL for the local install I'm working with is http://localhost/mw1/index.php/Main_Page
[00:16:14] <AndyRussG>	 $wgServer = "http://localhost";
[00:16:16] <awight>	 Krinkle: Perhaps the test harness should assert that those env vars are set...
[00:16:39] <AndyRussG>	 $wgScriptPath = "/mw1";
[00:16:57] <awight>	 It would be nice if there were a channel for sending configuration data from PHP -> js, also
[00:18:01] <AndyRussG>	 OK that's getting somewhere...
[00:18:02] <AndyRussG>	 grunt qunit
[00:18:02] <AndyRussG>	 Running "karma:main" (karma) task
[00:18:02] <AndyRussG>	 INFO [karma]: Karma v0.12.31 server started at http://localhost:9876/
[00:18:02] <AndyRussG>	 INFO [launcher]: Starting browser Chrome
[00:18:04] <AndyRussG>	 ERROR [launcher]: Cannot start Chrome
[00:18:07] <AndyRussG>	 	Can not find the binary google-chrome
[00:18:09] <AndyRussG>	 	Please set env variable CHROME_BIN
[00:18:11] <AndyRussG>	 Warning: Task "karma:main" failed. Use --force to continue.
[00:18:13] <AndyRussG>	 Aborted due to warnings.
[00:18:43] <Krinkle>	 AndyRussG: What OS, do you have Chrome installed?
[00:18:45] <awight>	 Krinkle: This could be part of a compile step... for now, it could run as a Composer script, and dump a .json file or something.
[00:19:03] <Krinkle>	 awight: phpunit?
[00:19:52] <AndyRussG>	 Krinkle: yes, on Debian
[00:19:58] <AndyRussG>	 BTW it worked!
[00:19:59] <awight>	 Krinkle: no, I was muttering that, it sounds like we need a compilation step which checks that the LocalSettings.php (etc) is synced to a .json file, which can be discovered by Grunt.
[00:20:08] <awight>	 AndyRussG: great!
[00:20:41] <AndyRussG>	 Or I mean, it ran tests and opened chrome
[00:20:53] <AndyRussG>	 in reverse order ^
[00:20:55] <AndyRussG>	 $ grunt qunit
[00:20:55] <AndyRussG>	 Running "karma:main" (karma) task
[00:20:55] <AndyRussG>	 INFO [karma]: Karma v0.12.31 server started at http://localhost:9876/
[00:20:55] <AndyRussG>	 INFO [launcher]: Starting browser Chrome
[00:20:56] <AndyRussG>	 INFO [Chrome 40.0.2214 (Linux)]: Connected on socket PYcXWzbNcI_a5LUJc0xO with id 28723612
[00:20:58] <AndyRussG>	 ................................................................................
[00:21:00] <AndyRussG>	 ................................................................................
[00:21:03] <AndyRussG>	 ..................
[00:21:05] <AndyRussG>	 LOG: 'Exception thrown by test.module1'
[00:21:07] <AndyRussG>	 ERROR: 'Error: expected', Error{}
[00:21:09] <AndyRussG>	 ................................................................................
[00:21:12] <AndyRussG>	 ...
[00:21:14] <AndyRussG>	 Chrome 40.0.2214 (Linux): Executed 261 of 261 SUCCESS (4.561 secs / 4.529 secs)
[00:21:16] <AndyRussG>	 Done, without errors
[00:21:17] <Krinkle>	 awight: The problem is with multi-wiki setups. If it's a single-wiki setup, one can just shell out to maintenance/eval.php and echo $wgServer, or maintenance/getConfiguration.php
[00:21:40] <Krinkle>	 However I didn't do that for now as in most environment that install mediawiki (e.g. jenkins, vagrant) can already pre-expose those variarbles.
[00:21:45] <Krinkle>	 And locally I just add them to bash_profile
[00:21:54] <Krinkle>	 next to the MW_DB you already need for maintenance scripts in extensions
[00:22:17] <Krinkle>	 AndyRussG: Nice. What did you do to make it work?
[00:22:40] <awight>	 Krinkle: ah, yeah multi-wiki has bitten me too.  FWIW, we have to deal with a new multiwiki scenario, where we check out an entire mw-core tree at a different revision than master (for payments.wikimedia.org vagrant).
[00:22:40] <AndyRussG>	 export CHROME_BIN='/usr/bin/chromium'
[00:22:42] <Krinkle>	 It couldn't auto-discover chrome the first time?
[00:22:47] <Krinkle>	 Ah, Chromium
[00:23:00] <Krinkle>	 Yeah, it only looks for Chrome (in about a dozen different places) but not Chromium.
[00:23:03] <Krinkle>	 I have a patch for that upstream.
[00:23:10] <AndyRussG>	 Ah hmmm :)
[00:24:15] <AndyRussG>	 Krinkle: out of curiousity, why does it say, "ERROR: 'Error: expected', Error{}"?
[00:24:27] <Krinkle>	 AndyRussG: part of some random unit test
[00:24:45] <Krinkle>	 That tests the ResourceLoader handling of exceptions
[00:24:50] <Krinkle>	 which qunit then logs to the console
[00:25:01] <Krinkle>	 Should be mocked in some way
[00:25:14] <AndyRussG>	 Aaaaah I see, cute :)
[00:25:37] <AndyRussG>	 I guess that's the explanation for the preceeding line, too, and also the fact that it ends with "Done, without errors"
[00:25:40] <awight>	 Krinkle: I'm shooting in the dark from the hip here (ahem...), but maybe the multiwiki configuration should be represented in the .json thing I'm imagining, so you can run "grunt --wiki default qunit" and so on.
[00:26:26] <Krinkle>	 AndyRussG: getConfiguration.php already supports that as being a maintenance script.
[00:26:40] <Krinkle>	 Maintaining and syncing a file seems messy.
[00:27:10] <Krinkle>	 So we could pass  --wiki on to that
[00:28:18] <awight>	 Krinkle: that would be perfect.
[00:29:49] <AndyRussG>	 Krinkle: awight: woot! I re-enabled CentralNotice on the wiki I'm testing with, and all the tests pass again!
[00:30:58] <AndyRussG>	 And yea it does run the CN tests... the number of tests run went from 261 to 335, and we have exactly 74 CN qunit tests
[00:35:01] <Krinkle>	 Great
[00:35:12] <AndyRussG>	 yeah!
[00:35:28] <AndyRussG>	 Now looking at the integration config again...
[00:45:18] <Krinkle>	 AndyRussG: So, you'll want to add an entry for '{name}-{ext-name}-qunit-karma' in the jobs section of mediawiki-extension and invoke it with CentralNotice. Similar to what it does for the non-karma qunit variant already.
[00:45:36] <grrrit-wm>	 (PS4) Awight: Use common code for Adyen form handling [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189136 (https://phabricator.wikimedia.org/T86251)
[00:45:50] <Krinkle>	 And then enable it in zuul.layout.yaml for the 'test' and 'gate' pipeline. And probably non-voting for now as it's still in flux while I finetune it.
[00:46:10] <AndyRussG>	 Krinkle: yurp...
[00:46:26] <AndyRussG>	 I was also looking at mwext-VisualEditor-qunit-karma, but I guess that has stuff that's special for VE
[00:47:37] <AndyRussG>	 (L204+
[00:47:38] <AndyRussG>	 )
[00:48:56] <Krinkle>	 Yeah
[00:49:03] <Krinkle>	 The '{name}-{ext-name}-qunit-karma' template should work for CN
[00:51:27] <AndyRussG>	 K
[00:52:11] * AndyRussG studies new-for-me files
[00:54:09] <AndyRussG>	 In that layout.yaml for CN we have currently only this:
[00:54:09] <AndyRussG>	   - name: mediawiki/extensions/CentralNotice
[00:54:09] <AndyRussG>	     template:
[00:54:09] <AndyRussG>	       - name: extension-qunit
[00:54:09] <AndyRussG>	       - name: extension-unittests
[00:54:12] <AndyRussG>	       - name: extension-rubylint
[00:56:14] <AndyRussG>	 others have no template prop but gate-and-submit, test and check, and some have all four... ah there's aslo one with experiemntal:
[00:56:21] <AndyRussG>	 bear with me, I'll get it in a sec :)
[00:59:15] <grrrit-wm>	 (CR) Ejegg: [C: 2] "Nice! I wonder if GatewayResultPage should be a subclass of GatewayPage with this default handleRequest behavior." [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189136 (https://phabricator.wikimedia.org/T86251) (owner: Awight)
[00:59:37] <grrrit-wm>	 (Merged) jenkins-bot: Use common code for Adyen form handling [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189136 (https://phabricator.wikimedia.org/T86251) (owner: Awight)
[00:59:42] <grrrit-wm>	 (PS11) Ssmith: Data retrieval/superclassing. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343
[01:02:54] <AndyRussG>	 Krinkle: awight: https://gerrit.wikimedia.org/r/#/c/194773/
[01:04:23] <AndyRussG>	 ah right, non-voting...
[01:05:08] <grrrit-wm>	 (PS1) Ejegg: Clean up some Adyen cruft [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/194774
[01:07:01] <grrrit-wm>	 (CR) Ssmith: [C: 2] Lint all js except node_modules and bower_modules [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/194389 (owner: Ejegg)
[01:13:47] <AndyRussG>	 Mmmm eats! backsoon... Thanks much Krinkle awight
[01:17:19] <K4-713>	 I thought he was a bassoon.
[01:17:21] <K4-713>	 Huh.
[01:40:20] <grrrit-wm>	 (CR) Ejegg: "Some initial observations, will look again in the morning" (10 comments) [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343 (owner: Ssmith)
[01:40:36] <ejegg>	 good night all
[01:43:21] <atgo>	 hey awight
[01:43:25] <atgo>	 good night ejegg|away!
[01:52:47] <awight>	 atgo: hi!
[16:49:44] <AndyRussG>	 Hey awight... How's Mari doing?
[16:51:34] <awight>	 Oh thx!  We're giving her the eye drops, she's doing well.
[16:56:35] <AndyRussG>	 Ah glad to hear :)
[16:57:46] <AndyRussG>	 awight: do you think it's not too late to write a proposal for Wikimania for the more general Editor Campaigns idea?
[16:58:31] <AndyRussG>	 I bet they'd not say now
[16:58:32] <AndyRussG>	 no
[17:00:33] <awight>	 AndyRussG: We should probably ask the selection committee first... it's definitely past the deadline, and they're rejecting proposals out of hand, cos: https://wikimania2015.wikimedia.org/wiki/Submission_review#Disqualified
[17:00:46] <AndyRussG>	 Oooh! really darn
[17:01:03] <awight>	 There might be an alternative, though... unconference sessions or something?
[17:03:28] <awight>	 It's a talk that could be given again and again, until everyone is on board :)
[17:03:53] <AndyRussG>	 awight: yeah that sounds better, then. Drat! sorry for my inactionly tardiness bla bla etc
[17:04:10] <AndyRussG>	 Even disqualifying WMF employees!
[17:04:20] <AndyRussG>	 Well, I guess it's a popular venue then
[17:05:20] <awight>	 Me too
[17:05:32] <AndyRussG>	 awight: all the disqualifieds have no abstract
[17:05:34] <awight>	 I barely squeaked my crypto-branching proposal through ;)
[17:05:39] <awight>	 AndyRussG: that is true.
[17:05:57] <awight>	 I wouldn't mind writing / collaborating on an abstract anyway, and we can reuse if rejected...
[17:06:50] <AndyRussG>	 Hahah I was still copy editing and watching the clock, and I had one minute left, and was wondering if the server clock might be 30 seconds ahead of my laptop's
[17:07:26] <awight>	 oh dear :)
[17:07:49] <awight>	 I see someone who got disqualified for submitting before midnight D.F. time!
[17:07:57] <awight>	 (organizers meant, UTC)
[17:08:17] <awight>	 AndyRussG: you have something written?
[17:08:26] <AndyRussG>	 No...
[17:08:54] <AndyRussG>	 ...but ther'es doc that'd be a fine starting point... sounds like a good idea! We could write the submission, and ask the committee if they want to consider it, and if not, what they suggest
[17:09:11] <AndyRussG>	 And could also ask Quim if there might be a place for it at the Wikimania hackathon
[17:09:55] <awight>	 I like the idea of presenting the whole package, of several interoperating components... e.g. include Gather and Workflow in the vision.
[17:10:05] <awight>	 hehe s/e.g./i.e./
[17:10:11] <AndyRussG>	 awight: that sounds like a great idea
[17:10:19] <awight>	 +1 sneaking in via the hackathon
[17:10:56] <AndyRussG>	 We could write it up but not put it immediately on the Wikimania wiki but somewhere else draft-ish
[17:11:02] <AndyRussG>	 and see what they sez?
[17:11:15] <AndyRussG>	 Or should we put it where all the submissions go?
[17:17:43] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Create CI slave instance for CiviCRM testing - https://phabricator.wikimedia.org/T89894#1095906 (awight) Created a m1.small instance, and following these instructions:     https://wi...
[17:19:12] <awight>	 AndyRussG: donno, we might as well try to sneak it into the middle of the pile of submissions ;)  DaTE: feb 15
[17:49:45] <AndyRussG>	 awight: so how should we proceed? when is a good time for you to work on something?
[17:50:03] <AndyRussG>	 i.e. should we each write some notes, maybe then do a hangout, etc?
[17:50:13] <AndyRussG>	 or whatever u think would work best
[17:50:52] <awight>	 Any time!  I'm spending the day working on setting up a CI slave on labs, it's a good thing to interrupt frequently ;)
[17:52:10] <AndyRussG>	 awight: OK! How about in a couple hours or so?
[17:52:20] <awight>	 AndyRussG: pls send me the etherpad link we were working on... different computer...
[17:52:23] <awight>	 sure!
[17:52:58] <AndyRussG>	 awight: http://etherpad.wikimedia.org/p/ContinuingEditorCampaigns
[17:59:59] <grrrit-wm>	 (PS13) AndyRussG: Preload JS comes with no baggage [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/181244 (owner: Awight)
[18:17:35] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Create and provision CI slave instance for CiviCRM testing - https://phabricator.wikimedia.org/T89894#1096112 (awight)
[18:28:10] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Create and provision CI slave instance for CiviCRM testing - https://phabricator.wikimedia.org/T89894#1096172 (awight)
[18:28:32] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Create and provision CI slave instance for CiviCRM testing - https://phabricator.wikimedia.org/T89894#1096173 (awight) Open>Resolved
[18:28:33] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, § Fundraising Sprint Abba, § Fundraising Sprint Beastie Boys, § Fundraising Sprint the Cure, and 5 others: Deploy CiviCRM integration job to WMF integration server - https://phabricator.wikimedia.org/T86374#1096174 (awight)
[18:30:00] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Configure Jenkins to run CiviCRM builds on Fundraising CI slave instance - https://phabricator.wikimedia.org/T89895#1096178 (awight)
[18:30:18] <wikibugs>	 Wikimedia-Fundraising-CiviCRM, Continuous-Integration, Fundraising Tech Backlog, Release-Engineering, and 2 others: Configure Jenkins to run CiviCRM builds on Fundraising CI slave instance - https://phabricator.wikimedia.org/T89895#1048130 (awight)
[18:33:28] <grrrit-wm>	 (PS1) Awight: Move integration scripts into this repo [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915
[18:33:59] <grrrit-wm>	 (PS2) Awight: Move integration scripts into this repo [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100)
[18:34:15] <awight>	 ejegg|errant: If you have a minute to kickme, ^^
[18:53:47] <grrrit-wm>	 (PS3) Awight: Move integration scripts into this repo [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100)
[18:55:02] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[18:57:06] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[18:59:58] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[19:10:53] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[19:13:17] <ejegg|errant>	 awight: sure, one sec
[19:13:39] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[19:13:55] <awight>	 ejegg|errant: actually, I'm still tweaking on those...
[19:14:40] <grrrit-wm>	 (PS4) Awight: WIP Move integration scripts into this repo [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100)
[19:15:27] <ejegg|errant>	 ahh, I see.  lmk when they reach steady state
[19:15:35] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[19:15:59] <awight>	 ejegg|errant: It's been fun :)  I think we're only a few days away from Civi CI, though!
[19:18:03] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[19:20:43] <ejegg|errant>	 woohoo!
[19:21:17] <grrrit-wm>	 (PS12) Ssmith: Data retrieval/superclassing. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343
[19:25:08] <grrrit-wm>	 (Abandoned) Ssmith: Save configuration for fraud gauge. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/189840 (https://bugzilla.wikimedia.org/86836) (owner: Ssmith)
[19:52:44] <grrrit-wm>	 (PS1) Awight: HHVM compatibility when using "php -r" [wikimedia/fundraising/civicrm-buildkit] - https://gerrit.wikimedia.org/r/194937
[19:54:24] <awight>	 ejegg: if u can, ^
[19:54:39] <ejegg>	 sure thing, just playing with php -r !
[19:56:34] <ejegg>	 heh, they really like PHP.  I guess some of that logic would be more complicated in pure shell script
[19:56:38] <awight>	 wat.
[19:56:44] <awight>	 oh, yeah exactly
[19:56:51] <awight>	 hehe I misread that as "I really like PHP"!
[19:56:57] <ejegg>	 heh
[19:57:11] <ejegg>	 haven't drunk that much of the koolaid yet
[19:57:32] <awight>	 I think this change is better for even the Zend PHP, cos it protects against arguments that might have special meanings 'n' stuff.
[19:57:41] <awight>	 Yeah, they don't make kool-aid strong enough.
[19:58:01] <grrrit-wm>	 (CR) Ejegg: [C: 2] HHVM compatibility when using "php -r" [wikimedia/fundraising/civicrm-buildkit] - https://gerrit.wikimedia.org/r/194937 (owner: Awight)
[19:58:10] <grrrit-wm>	 (CR) Ejegg: [V: 2] HHVM compatibility when using "php -r" [wikimedia/fundraising/civicrm-buildkit] - https://gerrit.wikimedia.org/r/194937 (owner: Awight)
[19:58:11] <awight>	 I hope u haven't looked at the src code for PHP?
[19:58:32] <ejegg>	 no, why?
[19:58:36] <ejegg>	 That horrifying?
[19:58:42] <awight>	 yeah
[19:58:53] <awight>	 it's like a high school compiler-writing exercise
[19:59:04] <ejegg>	 i'll continue in my blissful ignorance then
[19:59:25] <awight>	 ifelse OPCODE='if' then do some hair-raising bloated C thing...
[19:59:55] <awight>	 hehe, at least they resisted the urge to optimize.
[20:00:44] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:04:37] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:10:45] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:14:14] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:16:39] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:29:15] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:32:00] <grrrit-wm>	 (CR) Awight: "check experimental" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100) (owner: Awight)
[20:32:04] <awight>	 ahem...
[20:37:47] <AndyRussG>	 ahem?
[20:38:27] <AndyRussG>	 awight: I was gonna look at the EditorCampaigns/Edu stuff finally, if u have a chance...
[20:39:12] <AndyRussG>	 Just gonna add some brainstorming to the etherpad
[20:42:06] <awight>	 AndyRussG: we were just about to lunch, how about 4:30 EST?
[20:42:21] <AndyRussG>	 awight: sure, that's perfect :)
[20:42:28] <awight>	 ok, thanks!
[20:42:32] <AndyRussG>	 likewise! :)
[20:42:41] <AndyRussG>	 Does that mean that standup is unstood?
[20:44:28] <grrrit-wm>	 (PS1) Awight: update packages [wikimedia/fundraising/civicrm-buildkit/vendor] - https://gerrit.wikimedia.org/r/194943
[20:45:15] <grrrit-wm>	 (PS1) Awight: composer package updates [wikimedia/fundraising/civicrm-buildkit] - https://gerrit.wikimedia.org/r/194944
[20:48:52] <awight>	 AndyRussG: good point, I'll try to feign attendance :)
[20:49:30] <AndyRussG>	 awight: heh I'm sure it's not that important!
[20:51:26] <awight>	 In case I miss it--my standup is: I'm unblocked on the Civi continuous integration, I've got a slave instance provisioned and linked to a job which can be triggered by a comment on patches.  It doesn't run correctly yet, so I'll be fixing that for a day or two.
[20:52:25] <awight>	 ejegg|bike: more CI patches above...
[20:53:20] <grrrit-wm>	 (PS5) Awight: WIP Move integration scripts into this repo [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/194915 (https://phabricator.wikimedia.org/T78100)
[20:57:56] <grrrit-wm>	 (CR) Ejegg: [C: 2 V: 2] update packages [wikimedia/fundraising/civicrm-buildkit/vendor] - https://gerrit.wikimedia.org/r/194943 (owner: Awight)
[20:58:30] <awight>	 ejegg: it's worse than I thought.  Some of the civicrm-buildkit/vendor/* are not being checked in for random git reasons...
[20:58:42] <ejegg>	 aah, thrilling
[20:59:05] <ejegg>	 oh yeah, i see the git packages
[20:59:10] <awight>	 u do?
[20:59:13] <ejegg>	 dagnabbit
[20:59:18] <ejegg>	 I mean in the composer.json
[20:59:20] <awight>	 totten/amp is on purpose, but where do you see the others
[20:59:21] <awight>	 oh
[21:01:50] <awight>	 what a disaster.  See vendor/phpunit/phpunit-selenium/.gitmodules
[21:03:20] <AndyRussG>	 ejegg: awight: following pizzzacat's wise lead, shall we call standup an email?
[21:03:28] <awight>	 oh ok
[21:03:50] <AndyRussG>	 I'm OK joining the hangout too if u prefer :)
[21:04:09] <ejegg>	 email is good!
[21:04:35] <AndyRussG>	 ejegg: coo' !
[21:04:42] <AndyRussG>	 (short for "cool")
[21:04:57] <AndyRussG>	 awight|lunch: buen provecho :)
[21:40:35] <awight>	 AndyRussG: hi!  I'm available any time...
[21:45:36] <AndyRussG>	 awight: K one sec :)
[21:46:20] <awight>	 Mari just had her first bike accident... 0mph, but still got gravel and leaves in the face!
[21:46:54] <ejegg>	 aww
[21:47:04] <awight>	 I suppose it was a tricycle accident.
[21:47:22] <ejegg>	 you got that thing from your folks' place fixed up?
[21:47:45] <awight>	 I did!  that's even more dangerous, tho.  This is a tricycle with stealth parent handle built in.
[21:48:04] <ejegg>	 ah, nice
[21:48:29] <awight>	 A generic one of these: http://ecx.images-amazon.com/images/I/411iWN1oTIL._SY300_.jpg
[21:49:19] <ejegg>	 that parent handle doesn't look too stealthy
[21:49:39] <awight>	 It's easy to ignore things at your back :)
[21:50:27] <awight>	 ejegg: fyi, this is ridiculous, but I'm considering actually using composer from the CI box.  nothin' to lose, eh?
[21:50:49] <AndyRussG>	 awight: aww
[21:50:51] <ejegg>	 o rly?
[21:50:57] <awight>	 can we do that?
[21:50:59] <awight>	 hmmmm
[21:51:11] <AndyRussG>	 awight: also all set anytime re: campaigns etc
[21:51:19] <ejegg>	 you mean on each checkin?
[21:51:23] <awight>	 There's already some stuff which downloads "civix"...
[21:51:33] <awight>	 ejegg: yeah, just on the CI box, to provision the testing env
[21:51:41] <awight>	 AndyRussG: ok, hanging out...
[21:52:19] <awight>	 you don't even want to see the weather here.
[22:08:00] <ejegg>	 awight: looks like crm/civicrm branch dev-v4.4 last had our master patches merged in March of last year.  Shall I merge in the stuff we've done since then, or do you have other plans?
[22:14:28] <pizzzacat>	 awight what's the trick I'm supposed to do where gerrit complains I don't have a change ID?
[22:17:23] <awight>	 pizzzacat: mmm... so this is not the latest commit?
[22:18:20] <awight>	 ejegg: Sure, if you think the changes are sane.  I honestly can't remember whether those were good or not.
[22:18:49] <awight>	 Most of that was just bugfixing and compatibility stuff, which might have been fixed in a more recent civi upstream 4.4. release
[22:18:55] <AndyRussG>	 pizzzacat: as a temporary measure, if the gerrit complaint contains a suggested ID, you can copy it, then do git commit --amend and paste it into the commit message
[22:19:02] <ejegg>	 awight: cool, probably don't need most of 'em
[22:19:10] <awight>	 I hopenot!
[22:19:17] <ejegg>	 Oh, and I should do those on top of your contrib-v4.4 merge
[22:19:54] <AndyRussG>	 awight: here's what Anke, the OPW intern, has been working on. Her stint with us is ending Monday, BTW https://www.mediawiki.org/wiki/User:Sputniza/Interview_guide_line
[22:20:06] <AndyRussG>	 She also has a final report in the pipeline
[22:21:30] <grrrit-wm>	 (PS13) Ssmith: Data retrieval/superclassing. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343
[22:21:40] <awight>	 ejegg: That sounds much better!
[22:22:42] <awight>	 AndyRussG: What is that?  An exit poll method for EP students?
[22:23:04] <AndyRussG>	 awight: no, it's Anke's interviews for research design... Results at the bottom
[22:23:18] <AndyRussG>	 Or mostly undistilled results
[22:23:27] <awight>	 But who gets interviewed?
[22:23:34] <AndyRussG>	 I think they were EP professores
[22:23:46] <awight>	 OK i see, teachers and students
[22:24:00] <awight>	 and it looks like, after they have used EP for a semester
[22:24:51] <AndyRussG>	 She e-mailed a draft version of the final report but I dunno if it's on-wiki yet, I should ask before sharing I guess
[22:25:27] <awight>	 The 3 examples are nice
[22:25:38] <awight>	 AndyRussG: you want to continue hanging out?
[22:26:16] <awight>	 I pinged yurik, but it was the end of the day, so no luck locating his .js workflow thing.
[22:26:23] <AndyRussG>	 awight: sure, one sec, getting coffee
[22:26:34] <awight>	 I think he said it was for the Zero portal... which unfortunately means we cannot access it, afaik
[22:26:37] <awight>	 k
[22:27:35] <awight>	 hehe, or I could dump the db :)
[22:28:16] <pizzzacat>	 the problem was that I had too many change id's because it was a rebase/merge
[22:28:21] <AndyRussG>	 awight: I haz coffee
[22:28:26] <awight>	 :)
[22:29:19] <awight>	 pizzzacat: yeah, rebase -i might not work on a merge commit, so my advice is to checkout the offending commit, then "git commit --amend", which will give it a Change-Id, then go ahead and cherry-pick the rest of your branch.
[22:30:29] <awight>	 AndyRussG: u in there?  I see your avatar :)
[22:30:55] <pizzzacat>	 yeah I just had to delete the other commit msgs that weren't the one being actually committed
[22:31:02] <pizzzacat>	 and then it worked
[22:31:10] <pizzzacat>	 err change id's
[22:31:35] <awight>	 whoa
[22:31:40] <awight>	 delete?
[22:31:46] <pizzzacat>	 yep
[22:31:50] <AndyRussG>	 One sec
[22:31:51] <awight>	 That sounds a bit odd, but I'm glad it worked for now :)
[22:31:53] <awight>	 k
[22:31:59] <pizzzacat>	 everything was squashed into one commit msg
[22:32:05] <pizzzacat>	 from the other squashed commits
[22:32:06] <awight>	 oooh right.
[22:32:15] <AndyRussG>	 awight: I'm there again now
[22:32:19] <pizzzacat>	 so I just got rid of the unnecessary ones
[22:32:43] <awight>	 Yes, now it all makes sense!
[22:32:53] <pizzzacat>	 :)
[23:55:07] <awight>	 AndyRussG: I tossed a short abstract in that etherpad, fwiw
[23:56:01] <AndyRussG>	 awight: nice!
[23:56:34] <AndyRussG>	 I'm turning domestic for a bit but I'll get to it again sooooooooon :)
[23:56:48] * AndyRussG prepares to scrub floors
[23:57:42] <awight>	 lol, that's usually my job, too
[23:58:14] <AndyRussG>	 the best way to find interesting stuff!
[23:59:00] <awight>	 :D don't spend it all in one place...