[00:25:17] (PS2) Awight: Rewrite all "LIKE" queries on civicrm_prevnext_cache [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 [00:25:31] (CR) jenkins-bot: [V: -1] Rewrite all "LIKE" queries on civicrm_prevnext_cache [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 (owner: Awight) [00:26:29] (PS3) Awight: Rewrite all "LIKE" queries on civicrm_prevnext_cache [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 [00:48:19] (CR) Ejegg: [C: -1] "This ought to help! Just need to delete some trailing percent signs." (2 comments) [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 (owner: Awight) [01:03:14] (PS4) Awight: Rewrite all "LIKE" queries on civicrm_prevnext_cache [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 [01:03:21] (CR) Awight: "Thanks! I found a third one, even..." [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 (owner: Awight) [01:25:06] (CR) Ejegg: [C: 2] "These functions sometimes get called with _alphabet already appended to $cacheKey, but it shouldn't hurt that we tack another one on." [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/180656 (owner: Awight) [01:25:18] ejegg: ooo, good point [01:25:24] rats, that might sort of matter [01:25:29] oh? [01:25:39] cos then we aren't deleting... [01:25:42] It would have the same effect, I think [01:25:51] just an unnecessary OR [01:26:19] ok yeah i checked, and it's only on select or delete [01:26:57] there's some stuff with totally different cache keys in refillCache too [01:27:06] but that's only called from the built-in dedupe merge [01:27:14] (PS1) Awight: update civicrm submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/180694 [01:27:16] which we've disabled [01:27:42] u know, I think I tracked that down and it's never called? Or... I found some important code path within that which is never run... [01:27:45] getting old [01:28:19] (CR) Awight: [C: 2 V: 2] update civicrm submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/180694 (owner: Awight) [01:28:33] oh, the code path never is hit, is it? [01:28:47] (PS1) Awight: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/180695 [01:28:56] yah but I'd have to dig around in irc logs for the details :( [01:29:26] nonstandard key happens only when the cacheKeyString arg is null, but the only call to the fn uses a non-null cache key string [01:30:04] (CR) Awight: [C: 2] Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/180695 (owner: Awight) [01:30:12] (Merged) jenkins-bot: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/180695 (owner: Awight) [01:30:36] ejegg: I'm looking forward to being on the same git repo as Civi development, so we can really backport some fixes... [01:31:06] yeah, even 4.4 will be there, right? [01:31:18] yep! [01:31:21] !log update crm from f1e558592ee98ff8fc84d19ff2c0435619e11242 to ae7b2381667dd65d68812c58f61e3ea66fa9fa6f [01:31:25] Logged the message, Master [01:32:16] ok, searching still works at all... [01:32:22] I'll leave that alone, now [01:38:47] ejegg: WAT [01:38:53] I just read your UnionPay comment [01:39:28] srsly. [01:39:41] "just install this exe, and type your bank details into it" [01:39:47] "best to advise that... must be installed" [01:40:04] sounds like a blockbuster [01:40:36] okbye! [01:40:37] maybe they've got a .deb for red flag linux? [16:17:32] (PS4) Ejegg: Backend for x-by-y widget with 'group by' [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/179340 [16:18:18] (CR) Ejegg: [C: 2] fix up composer.json [wikimedia/fundraising/SmashPig] - https://gerrit.wikimedia.org/r/180577 (owner: Awight) [17:22:03] (PS1) Ejegg: Update Composer autoload for switchable WmfFramework [extensions/DonationInterface/vendor] - https://gerrit.wikimedia.org/r/180822 [17:25:49] (CR) Awight: "Cool! Looking forward to seeing how it works in the parent repo :)" [extensions/DonationInterface/vendor] - https://gerrit.wikimedia.org/r/180822 (owner: Ejegg) [17:29:13] (PS1) Ejegg: Add vendor submodule and update composer files [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180825 [17:29:15] (PS1) Ejegg: Use class_alias to select WmfFramework [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 [17:29:44] (CR) jenkins-bot: [V: -1] Use class_alias to select WmfFramework [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [17:29:59] (CR) jenkins-bot: [V: -1] Add vendor submodule and update composer files [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180825 (owner: Ejegg) [17:30:06] dang, 0 for 2! [17:31:37] aargh, jquery.json dependency breaking tests [17:31:43] omg [17:31:49] that's oold news... [17:32:06] hey, you got the composer validate job wired up! [17:32:16] did I? [17:32:26] mos def [17:32:35] Must have done it in my sleep [17:32:42] hehe 17:29:25 stderr: fatal: reference is not a tree: 4c18e544fce5f14c09487f0cdd3ac39b40402ffc [17:32:45] 17:29:25 Unable to checkout '4c18e544fce5f14c09487f0cdd3ac39b40402ffc' in submodule path 'vendor' [17:33:07] right, not till we merge it anyway! [17:33:37] weird. I tought other jobs let you reference unmerged patches [17:33:58] awight: only if it's a git ancestor, i think [17:35:07] I don't actually see that treeish in gerrit... [17:35:09] fwiw [17:35:47] oh, thought that was what I just reviewed in /vendor. let me check [17:36:48] yeah, https://gerrit.wikimedia.org/r/#/c/180822/ [17:37:26] oh [17:37:28] :) carry on [17:43:18] Why is it running so many more tests on my commit? [17:44:31] good question [17:44:41] those are core tests [17:44:48] looks like same suite.xml [17:45:27] aargh [17:45:40] suite.xml commit yesterday [17:45:43] ejegg: yep. [17:45:44] 61a30753a698408680557c018544df7349d174da [17:45:53] exactly what i'm looking at [17:46:06] so, that breaks us [17:46:52] hashar has been off of IRC for a while now, it's distracting or something [17:47:04] like months [17:47:07] heh, convenient [17:47:49] grr, it's a self-merge [17:47:50] GRR [17:48:04] ouch [17:48:07] also, [17:48:08] https://gerrit.wikimedia.org/r/#/c/180496/ [17:48:14] ERROR! It seems this job introduced a regression although it might be caused by a previous commit. in 4m 15s (non-voting) [17:48:17] wt [17:48:17] f [17:48:19] (PS3) AndyRussG: WIP Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 [17:48:40] dude, not cool [17:48:50] I'll revert... [17:51:32] ilar to this one, but with a '/' and the list name appended. 12:47 < ejegg> heh, convenient [17:51:34] 12:47 < awight> grr, it's a self-merge [17:51:37] 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined #wikimedia-fundraising [17:51:40] 12:48 < grrrit-wm> (PS3) AndyRussG: WIP Minor improvements in test fixtures [17:51:43] 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined #wikimedia-fundraising [17:51:46] 12:48 < grrrit-wm> (PS3) AndyRussG: WIP Minor improvements in test fixtures [17:51:49] 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined #wikimedia-fundraising [17:51:52] 12:48 < grrrit-wm> (PS3) AndyRussG: WIP Minor improvements in test fixtures [17:51:55] 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined #wikimedia-fundraising [17:51:58] wth [17:51:58] 12:48 < grrrit-wm> (PS3) AndyRussG: WIP Minor improvements in test fixtures [17:52:01] aargh [17:52:04] mouse went crazy [17:52:13] ejegg: that's what they all say... [17:52:17] (jk) [17:53:04] t> wth [17:53:04] 12:52 < Andg> 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined [17:53:07] g> 12:48 -!- atgo [~atgomez@wikimedia/agomez-wmf] has joined [17:53:19] goldangit [17:53:30] ejegg: maybe just +1 this revert [17:53:45] yep, will do [17:55:18] (PS4) AndyRussG: Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 [17:56:47] I wonder why the CentralNotice commits are not suffering the same issue [17:56:55] the "structure" tests aren't running there. [17:57:07] (CR) jenkins-bot: [V: -1] Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 (owner: AndyRussG) [17:57:17] bah, gonna downgrade kernel to the one where my mouse works... [17:58:02] ejegg|fubuntu: revert!!! [17:58:20] oh, now they are [17:58:34] AndyRussG: heads-up, see the backscroll. That test failure is not yr fault [17:59:25] awight: thanks, te: backstroll you mean the test failure? Hmm just trying to grok it... [17:59:45] s/te: backstroll/re:backscroll/ [17:59:46] AndyRussG: ejegg|fubuntu: erm, actually these tests might be exposing actual issues in our code [17:59:51] hehe +1 for backstroll [18:00:04] It's IRC's version of the moonwalk [18:00:11] ow! [18:00:31] awight: the jquery.json may be an issue, but the replacement doesn't exist in the mw version we're running [18:01:01] I'm playing the game in which I try to figure out wtf is going on in here without clicking on any links or anything. Today's limited backscroll has been notably good for this hour. [18:01:10] ejegg|fubuntu: ./DonationInterface.php: 'dependencies' => array( 'mediawiki.util', 'jquery.json' ), [18:01:19] I think that's our bad... [18:01:30] K4-713: :) doing what we can with a small amount of material [18:10:13] K4-713: I'll try to create interesting backstcroll taking that into account [18:10:38] Thanks, I'd appreciate that. [18:10:47] :) [18:10:54] (PS1) Awight: Register some missing autoload classes, to satisfy structure tests [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180842 [18:11:09] Actually, I could just write a bot to inject nonsense into the conversation two minutes after I join the channel. [18:11:29] More fun if someone else does it without telling you [18:11:36] Oooo. [18:11:36] (CR) jenkins-bot: [V: -1] Register some missing autoload classes, to satisfy structure tests [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180842 (owner: Awight) [18:12:09] * awight appears persecuted [18:12:15] whooa snowcat: http://imgur.com/a/5f2c4 [18:12:26] K4-713: for example ^ noo don't click [18:12:31] Already did. [18:12:44] (PS2) Awight: Register some missing autoload classes, to satisfy structure tests [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180842 [18:12:53] That thing is all like O.O [18:13:07] At least it dared to go out [18:14:50] awight: looks like the qunit fail on that one is a jenkins bork [18:16:16] (PS1) Awight: Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 [18:16:39] AndyRussG: nah it was a typo, surprising that phplint didn't catch it too... PS2 should fix it [18:16:56] awight: ah yea I see... [18:17:21] (CR) AndyRussG: [C: 2] "Tests are snow cats" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180842 (owner: Awight) [18:18:06] (Merged) jenkins-bot: Register some missing autoload classes, to satisfy structure tests [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180842 (owner: Awight) [18:20:02] (PS5) AndyRussG: Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 [18:20:09] so DI Javascript doesn't actually use jquery.json, as far as I can tell [18:20:23] yeah it looks very crufty [18:20:45] (CR) jenkins-bot: [V: -1] Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 (owner: AndyRussG) [18:21:03] there's a little bit of mw.util though [18:21:31] (CR) AndyRussG: "Rebased..." [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 (owner: AndyRussG) [18:22:57] (PS1) Awight: remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 [18:24:03] (PS6) AndyRussG: Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 [18:26:20] (PS2) Awight: Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 [18:27:31] ejegg: that ContributionTracking patch will need some affirmation as well [18:28:15] ok, will take a look [18:29:53] soon as I tame this damn mouse. maybe HW issue? [18:33:08] that sounds horrible [18:33:41] I have the dumpstered mouse at home... which I very frequently need to open up, to tickle the batteries... [18:33:51] when I found it, there was a ton of corrosion... [18:34:10] the only redeeming quality is that it makes it harder for my kid to hack into all open apps [18:34:15] heh [18:34:43] I'm actually using trackpad & trackpoint, but they started acting up like crazy [18:35:18] so far so good this reboot though. 3.16.0-26-generic might be the one [18:35:33] uggh [18:35:44] keeping up with the Jonses, I see [18:36:23] heh, totally necessary [18:36:40] Yeah I have not reaped any particular benefit from compiling kernels [18:37:19] why 'jquery.contributionTracking' ? [18:40:08] oh, probably because of all the jQuery [18:40:14] it's terrible [18:40:32] we should only do that if it's actually designed to be a jquery plugin [18:40:34] it should still depend on jquery though [18:41:04] (CR) jenkins-bot: [V: -1] Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [18:41:11] nggg there goes the mouse again. wtf. [18:41:28] i need a 'rmmod psmouse && modprobe psmouse' hotkey [18:42:01] ejegg: maybe it's a mouse issue... I think it's been like at least a decade since I had to fiddle with mouse drivers on Debian... [18:42:10] I mean, like maybe it's the mouse hardware [18:42:21] Just a thought... [18:43:02] (PS2) Awight: remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 [18:43:15] (CR) jenkins-bot: [V: -1] remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [18:43:17] maybe, but i suspect not. both trackpad and trackpoint fail simultaneously [18:44:08] hmm, jquery is never listed as a dependency, even in jquery plugins. sloppy! [18:44:20] huh, ok... [18:44:31] meh, I added it in PS2 [18:48:01] ejegg|biab: hmm... how can you tell which is failing if both are plugged in? [18:49:13] IRQ conflict [19:03:23] They're both built in to the laptop [19:03:39] usbmouse seems to be fine [19:04:47] by trackpoint u mean one of those little round stubbly things in the middle of the keyboard? [19:08:52] yep [19:09:31] yrrrp, I've never met one of those that didn't start misbehaving eventually [19:20:11] (PS1) Awight: (TR #2121) fr and he TY updates [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/180864 [19:20:19] ejegg: awight: K4-713: anyone have any idea why our browser test campaigns aren't showing up on http://en.m.wikipedia.beta.wmflabs.org/wiki/Main_Page ? [19:20:41] They _are_ here: http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [19:20:59] And I don't see anything wrong with the campaign's config: http://meta.wikimedia.beta.wmflabs.org/w/index.php?title=Special:CentralNotice&method=listNoticeDetail¬ice=CN+browser+tests [19:21:14] AndyRussG: awkward! [19:21:25] I thought I enabled mobile devices... [19:21:43] awight: u did! [19:21:46] AndyRussG: there was something the mobile team did to disable CN, in some places [19:21:50] that must be the thing [19:22:15] awight: makes sense, K lemme ask on -mobile, thanks! [19:22:29] AndyRussG: https://gerrit.wikimedia.org/r/173119 [19:22:35] ejegg: Do you have anything to check in about today? [19:22:49] AndyRussG: :( [19:23:07] K4-713: nothing I can think of just now [19:23:18] AndyRussG: afair, they were running an a/b test and they wanted to eliminate interference. We might be able to revert that, now [19:24:01] awight: Hmmm beta.wmflabs != beta site enabled via prefs on enwiki [19:24:22] also if in the jsconsole I go to http://en.m.wikipedia.beta.wmflabs.org/ [19:24:27] ejegg: Okay, in that case, want to skip the checkin? You can just put something on my calendar if you think of something later. [19:24:52] I get: [19:24:52] > mw.cnBannerControllerLib.choiceData [19:24:55] < [] [19:25:41] If CN modules were disabled, I don't think we'd get even that... [19:25:48] K4-713: sure, works for me! [19:26:19] Cool. [19:26:25] Okay, I'm going to food then. [19:29:16] (CR) Ejegg: "Looks like we need to roll in removal of the jquery.json dependency to get this merged." [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [19:32:23] (CR) Ejegg: "So, jquery is an illegal dependency? Guess PS1 was better." [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:36:35] (CR) Legoktm: "The 'jquery' module should not be dependend upon, it is always loaded and depending upon it should throw an error client-side (iirc)." [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:38:29] (CR) Ejegg: "Sorry, guess the error's coming from ContributionTracking!" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [19:38:41] (CR) Legoktm: "See https://www.mediawiki.org/wiki/ResourceLoader/Developing_with_ResourceLoader#Registering and http://www.gossamer-threads.com/lists/wik" [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:40:53] awight: gonna hijack your CT patch just to clear up the logjam [19:41:07] (PS3) Ejegg: remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:41:33] AndyRussG: browser test campaigns? [19:41:42] we are not running banners on beta, i believe. [19:42:04] atgo: yes, on beta.wmflabs.org [19:42:13] http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [19:42:17] what are you talking about? banners? [19:42:20] (CR) Ejegg: "Thanks Legoktm! Back to PS1." [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:42:25] At the top of the screen there you get "one", "two" or "three" [19:42:31] Those are banners [19:42:44] hm [19:42:51] They're for our automated cross-browser testing service [19:43:11] (CR) Ejegg: [C: 2] remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:43:18] i believe beta mobile has banners disabled [19:43:22] i could be wrong, but that sounds familiar [19:43:49] atgo: yeah awight also noted that issue... I'm just testing locally now to see if that's it, before going to bug more people [19:43:57] i think it's known [19:44:00] and supposed to be that way [19:44:24] atgo: well it's disabled on their "beta" skin, which is shown, but not by default, on the real production sites [19:44:54] i guess i'm still not seeing the problem [19:45:12] so the beta cluster, which is a mock-up of the real cluster with URLs that end in beta.wmflabs.org, should actually mimic production on that count [19:45:16] but AFAIK we are intentionally not running banners on Beta [19:45:20] One sec I'll explain... [19:45:23] like.. they disabled [19:45:36] atgo: https://integration.wikimedia.org/ci/view/BrowserTests/view/CentralNotice/ [19:46:09] ok. what i'm saying is that i am fairly sure they disabled CN or banners or something on beta [19:46:23] atgo: so there's this service, browser sauce or something, that runs tests automatically on different browsers and operating systems [19:47:04] i get that [19:47:09] what i'm trying to say is that this is intentional [19:47:15] (Merged) jenkins-bot: remove unused js dependency [extensions/ContributionTracking] - https://gerrit.wikimedia.org/r/180847 (owner: Awight) [19:47:15] For testing our code, they start up browsers in virtual machines and point them to http://en.wikipedia.beta.wmflabs.org/ [19:47:17] right [19:47:24] atgo: I think there's a confusion as to what "beta" means [19:47:41] I think for mobile "beta" does not mean that site [19:47:48] ? [19:48:13] (CR) Ejegg: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [19:48:28] ok [19:48:38] I think it's a special fetaure of the mobile site for seeing beta features ahead of time: http://www.mediawiki.org/wiki/Beta_Features/Minerva [19:49:52] AndyRussG: yes, and users that opt into beta should not get banners [19:49:56] the mobile team worked on this recently [19:50:01] atgo: exactly [19:50:06] so what's the issue? [19:50:06] (CR) Ejegg: [C: 2] Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [19:50:17] So that's where banners were disabled [19:50:30] yes [19:50:56] And as far as I should know, they should be still available on the mobile version of the beta wmflabs site, which is where we expect to have them for our browser tests [19:51:20] hm [19:51:21] ok [19:51:35] might it make sense to check with the mobile team to see if they did it in both places? [19:51:36] Unless somehow the beta skin is automatically enabled there, that's what I thought I'd check... [19:51:58] Yeah exactly... I'm just doing a bit of testing locally in case it's something I can tell without bugging them [19:55:10] alrighty [19:55:15] :) [19:55:18] i'm happy to ping them, it's probably a quick answer :) [19:55:46] atgo: thanks so much!! just confirmed it, that's not the issue... [19:56:10] Probably more like something strange about the way CN is set up on the beta cluster [19:57:01] awight: ^ if I enable beta features on http://en.m.wikipedia.beta.wmflabs.org/ (via user prefs) and then say, mw.cnBannerControllerLib.choiceData, I get Uncaught TypeError: Cannot read property 'choiceData' of undefined [19:57:32] whereas on the beta cluster without beta mobile features enabled, I just get the empty array of no campaign choices available [] [20:05:44] (CR) jenkins-bot: [V: -1] Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [20:07:28] wtf, jenkins? [20:07:35] npm aborted? [20:07:49] (CR) Ejegg: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [20:15:04] anyone know who's responsible for the setup of *.beta.wmflabs.org? Just pinged Coren, no answer so far... [20:37:31] AndyRussG|brb: that's spooky [20:38:09] (CR) Ejegg: Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [20:38:27] (CR) Ejegg: [C: 2] Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [20:38:45] * awight munches popcorn [20:39:30] * ejegg wipes brow with towel [20:39:34] (Merged) jenkins-bot: Satisfy structure tests [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180844 (owner: Awight) [20:39:40] TKO! [20:40:46] (PS2) Ejegg: Use class_alias to select WmfFramework [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 [20:40:48] (PS2) Ejegg: Add vendor submodule and update composer files [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180825 [20:40:51] whew [20:41:09] (CR) jenkins-bot: [V: -1] Use class_alias to select WmfFramework [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [20:41:42] (CR) Awight: [C: 2] Add vendor submodule and update composer files [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180825 (owner: Ejegg) [20:42:20] (CR) Awight: [C: 2] Update Composer autoload for switchable WmfFramework [extensions/DonationInterface/vendor] - https://gerrit.wikimedia.org/r/180822 (owner: Ejegg) [20:42:36] (CR) Awight: [V: 2] Update Composer autoload for switchable WmfFramework [extensions/DonationInterface/vendor] - https://gerrit.wikimedia.org/r/180822 (owner: Ejegg) [20:42:57] sorry, didn't get that auto-submit set up [20:43:03] (CR) Awight: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [20:43:15] ejegg: no rush. yeah the phplint job is probably wise [20:43:41] (Merged) jenkins-bot: Add vendor submodule and update composer files [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180825 (owner: Ejegg) [20:44:09] true. but is it just 'phplint' ? Or --phplint? [20:45:15] (CR) Awight: "Nice one. But is it not possible to $wgautoload the WmfFramework file?" (1 comment) [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [20:45:55] ejegg: yah repo-phplint [20:46:04] thanks! [20:46:07] actually, I think you might have to define that job as well [20:46:35] or mebbe not [20:47:11] ejegg: see ./jjb/wm-fundraising.yaml: [20:47:14] oops [20:48:14] wow, I'm not understanding how that glue works [20:48:26] whatever happens, let's make it more complicated :) [20:48:37] heh, yup, puzzled the heck out of me. [20:48:58] we still need a job-template statement, I don't see where we declare those [20:49:20] isn't "just run phplint on this repo" a pretty simple use case? [20:49:39] blaaargh [20:49:54] u must ask the emperor of byzantium [20:50:28] he is ensconced in his hareem, try again next month [20:50:53] So... {name}-phplint is defined in jjb/job-templates.yaml [20:51:17] we can rely on that to create a job in jjb/wm-fundraising.yaml [20:51:28] then attach to triggers from layout.yaml [20:51:29] :( [20:51:44] * AndyRussG moistens towel with brow [20:52:54] awight: is it? I think the beta cluster is supposed to mimic production... so I'm glad it did that... I was just asking for input on wikimedia-qa when I had to unexpectedlybrb [20:52:56] (CR) Awight: [C: 2] "I think it'll work. require_once seems evil, but I like the sounds of deprecating in favor of autoload.php in the near future." [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [20:53:31] AndyRussG: I've lost track of what the mobile team was disabling... even the backstroll didn't help :) [20:53:34] awight: oh, was just about to change that! [20:53:36] (Merged) jenkins-bot: Use class_alias to select WmfFramework [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180826 (owner: Ejegg) [20:53:41] oops! [20:54:03] ehh, got the rest of the changes stashed and ready [20:54:27] awight: aww hmm ... yeah in the mobile view u go to user preferences and check "enable beta", so you use the beta mobile skin, that's where it's disabled [20:54:47] why? [21:00:21] awight: 'cause that's where they were running an a/b test. They disabled campaigns so they wouldn't interfere with an a/be test they ran on production [21:00:39] oh ok that sounds familiar. Do they want to revert that, now? [21:01:33] awight: no idea, but it's not a problem AFIK [21:02:57] Ah, I see. So we don't care about beta features and skin. But the choiceData is empty on en.m.beta [21:03:04] ping [21:03:08] ? [21:03:20] oh stoodup [21:13:57] (CR) Awight: [C: 2] "Much more readable!" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 (owner: AndyRussG) [21:14:26] awight: right [21:14:29] also, thanks! ^ [21:14:45] (right wrt choiceData being empty) [21:14:46] (Merged) jenkins-bot: Minor improvements in test fixtures [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/180650 (owner: AndyRussG) [21:15:05] AndyRussG: well lmk if you need a fingerpuppet to voice ideas to :) [21:17:44] awight: will do thanks... I think I already mentioned the main changes: bringing default values into the test fixture data and adding expected results from the various stages in the process (for testing more methods) as well as sets of alocatee values (country, device, logged in, etc.) to test with [21:19:28] you've clearly got that stuff under control--I meant more, debugging the en.m.beta fail [21:23:05] Ah right! thanks... just now in the wikimedia-qa channel on that also, BTW [21:28:12] awight: should our stomp wrapper be in the composer autoload list? [21:29:59] (PS1) Ejegg: Use Composer-managed Stomp and Minfraud classes [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180894 [21:30:17] (CR) jenkins-bot: [V: -1] Use Composer-managed Stomp and Minfraud classes [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180894 (owner: Ejegg) [21:31:13] ahh right, tests need to include composer autoload too [21:32:59] hey K4-713 awight ejegg we're sending API calls to GC without any associated order ID [21:33:06] forwarded you the message just now [21:33:14] yikes! [21:33:16] atgo: Uh, how? [21:33:45] just forwarded [21:33:52] GC brought it up with us [21:34:04] making a card, too [21:34:05] Looks like somebody busted the orphan slayer or something. [21:34:12] Anyway: On it. [21:37:11] K4-713: FYI: https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/2305 [21:37:14] atgo: Yep, this is completely limited to the orphan slayer. [21:38:59] atgo: Also, this isn't the call that makes a payment. It's the call that gets the status. [21:39:18] so no donors would've seen any issue? [21:39:19] It's nice that they let us know, though. [21:39:25] atgo: Nope. [21:39:27] yeah!! i'm so pleased with them about that [21:39:59] Absolutely no impact to actual people, other than orphan transactions will have to wait for the nightly audit to get their thank you emails. [21:40:06] I don't think we have many of those anymore anyway. [21:40:10] awight: did you say the french & hebrew TYs went out? [21:40:16] oh rad, thanks K4-713 [21:40:19] yep [21:40:31] I'm still going to go cross-eyed at this for a bit, because... weird. [21:42:10] awight: Did I see something recently about... changing the stomp libs in donationinterface so they come through composer? [21:42:45] K4-713: got a patch set it for it, but not merged [21:43:00] hm. There goes that theory, then. [21:44:18] (PS2) Ejegg: Use Composer-managed Stomp and Minfraud classes [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180894 [21:45:28] (CR) jenkins-bot: [V: -1] Use Composer-managed Stomp and Minfraud classes [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180894 (owner: Ejegg) [21:47:30] ejegg: Oh man. Stomp libs *and* minfraud in one commit? [21:47:53] K4-713: well, they're both in the Composer autoload... [21:48:00] Scariful. :) [21:48:15] Whoever does that deploy should probably be really ready for the undeploy. [21:48:16] seems we need to tell jenkins specifically to check out the submodule though [21:48:36] K4-713: we don't start using the composer libs in that commit, tho [21:48:55] awight: in that one we do ^^^ [21:48:56] atgo: no, the TYs are pending CR [21:49:00] ejegg: oh :) [21:49:01] good. [21:49:42] just as soon as we can get Jenkins to check out /vendor before running tests [21:50:23] (PS1) Awight: Get rid of unused colspan=2 cruft [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/180902 [21:53:13] atgo: jfyi, I got to the bottom of the issue. [21:53:17]