[00:13:47] § Fundraising Sprint Devo, § Fundraising Tech Backlog: Logging should be handled outside of core gateway class - https://phabricator.wikimedia.org/T86266#1033034 (atgo) Testing something... will put this back in 2 seconds [00:13:55] § Fundraising Tech Backlog: Logging should be handled outside of core gateway class - https://phabricator.wikimedia.org/T86266#1033035 (atgo) [00:14:12] § Fundraising Sprint Devo, § Fundraising Tech Backlog: Logging should be handled outside of core gateway class - https://phabricator.wikimedia.org/T86266#964763 (atgo) [00:17:10] (PS4) Awight: New PaymentResult class for passing user interaction hints [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 [00:17:22] (PS3) Awight: Consolidate GatewayPage logic in the base class [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189133 (https://phabricator.wikimedia.org/T86251) [00:17:27] (PS2) Awight: Don't senselessly repeat base class code [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189134 (https://phabricator.wikimedia.org/T87043) [00:20:24] (PS2) Awight: DO NOT MERGE: octopus for development. [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189629 [00:20:26] (PS6) Awight: WIP doPayment() [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189153 [00:23:03] (CR) jenkins-bot: [V: -1] New PaymentResult class for passing user interaction hints [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [00:23:05] (CR) jenkins-bot: [V: -1] Consolidate GatewayPage logic in the base class [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189133 (https://phabricator.wikimedia.org/T86251) (owner: Awight) [00:23:39] (CR) jenkins-bot: [V: -1] Don't senselessly repeat base class code [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189134 (https://phabricator.wikimedia.org/T87043) (owner: Awight) [00:24:53] (CR) Awight: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [00:24:57] (CR) jenkins-bot: [V: -1] WIP doPayment() [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189153 (owner: Awight) [00:25:01] GRR [00:25:35] (CR) jenkins-bot: [V: -1] DO NOT MERGE: octopus for development. [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189629 (owner: Awight) [00:25:52] (CR) Ejegg: [C: -1] "Thanks for getting board details in setDisplayPage! You can delete the other set of $.get calls for board details now. Also, do you want" (1 comment) [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/188674 (owner: Ssmith) [00:32:14] (CR) Awight: "for the record, CI failure is unrelated to this patch..." [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [00:33:47] hah, composer.lock is not up to date [00:34:35] ejegg: James_F was explaining on the operations channel, it's https://gerrit.wikimedia.org/r/#/c/190119/ and RoanKattouw is currently deploying the fix... [00:34:47] tx for info! [00:34:57] (CR) Awight: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [00:35:02] * awight kicks the can a bit [00:48:12] (CR) Awight: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [01:04:35] (PS1) Ejegg: Add push and pop fns to DonationLoggerContext [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190134 [01:04:37] (PS1) Ejegg: Use DonationLogger in GatewayPage subclasses [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190135 [01:15:29] awight: let's merge the DI work so far into the drupal branch tmw morn and see if we broke the orphan adapter [01:16:06] yikes! [01:16:20] I think the orphan adapter is only run on payments4 [01:16:36] ah sorry, standalone mode, right [01:16:41] sounds good! [01:17:27] cool [01:18:46] o right, i meant recurring [01:18:49] oh, but looking at your change, you did mean the orphan adapter... [01:18:56] ? [01:19:09] i'm interested in both! [01:19:37] (CR) Awight: [C: -1] "The local keys thing is too hard to understand :)" (2 comments) [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190134 (owner: Ejegg) [01:20:23] (CR) Awight: [C: 2] Use DonationLogger in GatewayPage subclasses [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190135 (owner: Ejegg) [01:20:49] (CR) Awight: "recheck" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/185598 (owner: Awight) [01:24:00] (CR) Ejegg: "Yeah, lemme ditch those keys..." (2 comments) [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190134 (owner: Ejegg) [01:30:05] (PS7) Awight: WIP doPayment() [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189153 [01:39:08] (CR) Awight: "breadcrumbs: The remaining issue with WorldPay is probably caused by sloppy implementation of checkTokens(). The token call should not al" [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189153 (owner: Awight) [01:43:26] (PS2) Ejegg: Add push and pop fns to DonationLoggerContext [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190134 [01:43:28] (PS2) Ejegg: Use DonationLogger in GatewayPage subclasses [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190135 [03:22:15] (PS8) AndyRussG: WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 [03:22:55] (CR) jenkins-bot: [V: -1] WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 (owner: AndyRussG) [03:23:06] blrrrrg [03:24:06] (PS9) AndyRussG: WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 [03:28:14] Well at least that was easy [03:28:31] * AndyRussG talks to self IRL [03:28:39] AndyRussG: can haz IRL? [03:28:41] hahahaha [03:50:16] (PS10) AndyRussG: WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 [03:55:23] (PS11) AndyRussG: WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 [03:55:56] (CR) jenkins-bot: [V: -1] WIP: Test fixtures: expand to cover more scenarios [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 (owner: AndyRussG) [03:58:50] (CR) AndyRussG: "This PS contains an intentional test failure to smoke test CI. (The passing QUnit logs, for example, for the previous PS, show an error as" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188687 (owner: AndyRussG) [11:51:16] MediaWiki-extensions-CentralNotice, Wikimedia-Fundraising: Special:RecordImpression should die in a fire - https://phabricator.wikimedia.org/T45250#1033913 (faidon) Can someone give me a high-level summary of why a separate call to record impressions is needed, rather than just running analytics on our logs?... [15:15:37] (CR) AndyRussG: [C: 1] "Smoke tested on desktop and mobile... Locally, work as advertised :)" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/188394 (https://phabricator.wikimedia.org/T45250) (owner: Ejegg) [15:46:40] MediaWiki-extensions-CentralNotice, Wikimedia-Fundraising: Special:RecordImpression should die in a fire - https://phabricator.wikimedia.org/T45250#1034358 (AndyRussG) Re: priority, @faidon, you're completely right. Apologies and thank you for saying this! I've smoke tested [[ https://gerrit.wikimedia.org/r/... [15:55:03] AndyRussG: Hi. Is the stats of banner loading available to CN admins? [15:55:19] Hi Glaisher :) [15:55:23] Good question! [15:56:40] I think maybe.... not? except perhaps by special request? I'm not exactly the person to ask but I can help you look into it :) especially a bit later when SF people are up and about :) [15:57:00] Do you have a banner running that you want to check the stats on? [15:59:27] Yeah, someone just asked about how many users are seeing the steward elections banner. [16:01:35] Glaisher: right! I was just looking into that. [16:01:50] It looks like there's some errors in code in the banner itself [16:02:36] Yeah, the same code has been in use since 2011(?) and no one bothered to update it. [16:02:54] I made only some minor changes to it this year. [16:05:41] Glaisher: ah OK, yeah it happens... :) [16:06:48] Is your first concern debugging that? And then after it works, seeing about the stats, if necessary...? [16:08:47] Glaisher: you can check that a browser is receiving a banner by going into the developer tools and the console and looking at mw.cnBannerControllerLib.choiceData [16:09:25] AndyRussG: It's working as expected (which is why no one has updated it) so I don't think it's a priority for now. :S [16:09:39] Just some outdated JS stuff [16:09:45] ResourceLoader things [16:10:05] How is it expected to work? When I log in on enwiki, I don't see it, even though according to stuff in the console, I should [16:10:43] (on Firefox, or really Iceweasel, 35.0) [16:11:04] Right now I see two banners available to logged-in desktop users on enwiki [16:11:28] When that happens CentralNotice randomly chooses which one to show based on priorities and some arcane logic [16:11:39] You can force the randomness with a URL param though [16:11:51] The JS checks whether a user is eligible to vote and based on that it shows the banner to that user. [16:12:00] with css [16:12:11] I guess that's why it's showing up in the console [16:12:14] Glaisher: ah OK no wonder... [16:12:28] You're not eligible, then I guess. :P [16:13:00] Yeah when I say https://en.wikipedia.org/wiki/Main_Page?random=0.2 I do get the other banner [16:13:10] But not when I say Yeah when I say https://en.wikipedia.org/wiki/Main_Page?random=0.8 [16:14:34] So if it's showing up in the console, the stats on the server side would be unreliable in this case, right? [16:14:59] Glaisher: it depends on what's in the banner [16:15:26] In fact we're just talking about this issue, so if you'd like to comment on https://phabricator.wikimedia.org/T45250 := [16:15:29] :) [16:16:08] Basically there is a thing called Special:RecordImpression that's used by fundraising to tell when a banner was loaded but was not actually shown to the client based on Javascript logic contained in the banner itself [16:16:12] Which is exactly your use case [16:16:22] It wasn't clear until now the use cases outside Fundraising, in fact [16:16:45] Yeah, I'm cc'd to that ticket. [16:19:37] Glaisher: cool! [16:20:29] Glaisher: Hmmm just got notice of a more worser ticket! Please let me know if I can do anything more for you, please feel free to reach out anytime [16:20:59] sure [16:21:05] and thanks! :) [17:43:42] AndyRussG: so, something's still hitting BannerRandom? [17:44:05] (PS1) Awight: Remove server-side banner random [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/190239 [17:44:48] (CR) jenkins-bot: [V: -1] Remove server-side banner random [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/190239 (owner: Awight) [17:46:08] ejegg: something is at least apparently calling code there! [17:56:15] hey AndyRussG - did you see my email & invite justnow? [17:57:58] (PS2) Awight: Remove server-side banner random [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/190239 [17:58:34] (CR) jenkins-bot: [V: -1] Remove server-side banner random [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/190239 (owner: Awight) [17:58:58] atgo: thanks for the ping! sure, be there in a sec [17:59:04] thanks [17:59:05] :) [18:12:56] awight: ejegg: as per call with meganhernandez and atgo a minute ago, for the RecordImpression it sounds like the best immediate fix will be just to turn it off if a user is not targetted [18:13:20] so no calls with hide:empty [18:13:21] ? [18:13:24] Sure, as long as we're certain we don't need to debug that case [18:14:32] ejegg: I think... Or we could just turn it off when the _server_ thinks there's no available campaign, as in, when choices are empty. That might be safer [18:14:52] awight: yeah for now I think :) [18:17:12] AndyRussG: no choices at all, or no choices after the first round of filtering? [18:17:28] ejegg: yes [18:17:39] ejegg: yes that's the question [18:18:00] k [18:27:35] ejegg: awight: I guess we could turn of S:RI only if choices are empty before any filtering, as the first option [18:28:10] That would definitely turn things down several notches and still give us debugging for client-code-related show/hide errors (like device stuff or logged-in status) [18:28:37] After that we could see what ops says... what think? [18:29:18] hell yeah. [18:29:36] Although--having the villagers all here with there pitchforks was almost useful... [18:29:44] there->their [18:30:24] pitchforks... mmm [18:30:47] cos we have a huge haystack that needs baling... [18:32:21] good to have so many tests when baling fast [18:33:10] oh yeah, I'm bowling against the bumpers at the moment ;) [18:34:12] What I meant to say though, was that the ops attention on Special:RecordImpression might be to our advantage. That's the only down side to hiding the problem... [18:40:48] Hmmmm [18:51:51] (PS3) Awight: Remove server-side banner random [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/190239 [18:53:12] AndyRussG: okay, that thing is ready for smoke testing ^^ [18:57:56] ejegg: Hey, are you available for checking in, or are you also wearing the fireman shorts? [18:59:39] K4-713: i'm available [18:59:46] mostly just watching the fire trucks [18:59:54] Nice. [19:00:02] Ah, give me a minute to find somewhere to be. [19:00:28] OH WAIT. [19:00:31] We got a room! [19:00:33] Good for us. [19:00:36] Going there. :) [19:00:51] woohoo [19:01:51] AndyRussG: I'm wandering off for now, please lmk if you see issues with my haphazard butchery! [19:02:04] (by wandering off, I mean standing right here and doing other things :) [19:02:16] awight: thanks! I'm sure it's near-surgical precision [19:02:25] hehehehe [19:02:34] unwarranted trust [19:02:44] Want to buy an old car? :p [19:03:31] Yes! [19:03:46] ...if it comes with free shipping [19:04:07] ejegg: ...I have apparently contracted calendar gremlins. [19:04:29] heh, i thought you just wanted to meet 1/2 hr early [19:04:42] works for me, if you're all set up [19:04:50] I am not. [19:05:00] I can't get in the room until I actually have it. [19:05:33] At least this set of calendar gremlins makes me early and not late. [19:05:45] Early, I can work with. [19:05:58] only wasting your own time, i guess [19:06:07] Exactly. [19:06:10] I'm really good at that. [19:09:11] Fundraising-Backlog, § Fundraising Sprint E: Change errors on credit card forms from popup to red text - https://phabricator.wikimedia.org/T86214#1034835 (atgo) [19:10:03] Wikimedia-Fundraising-CiviCRM, Fundraising-Backlog, § Fundraising Sprint E: Make Civi Reminders work - https://phabricator.wikimedia.org/T86345#1034838 (atgo) [19:12:39] hey ccogdill i don't think you've moved this one, but have you? https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/790 [19:13:05] I haven’t atgo [19:13:14] I emailed asking if a phab task existed [19:13:28] and didn’t follow up… [19:13:30] :) [19:15:22] cool. i'm doing it now :) [19:15:31] thanks! can you cc me? [19:15:45] ccogdill: yes. do we need to get translations going for this as well? maybe we shold look in jessica? [19:16:13] I guess so! I wasn’t sure how long the technical side would take so I hadn’t involved her yet [19:16:22] we may also want Megan to take a look at the proposed language [19:16:28] before translating [19:17:21] yeah [19:20:07] ccogdill: would you be able to take the lead on the language? [19:22:05] how so? I guess I’m not sure what stage we’re in [19:22:41] should I sent a draft of all the proposed responses to Megan and Jessica and ask her to translate? Is tech able to do all the proposed form changes we want? [19:24:15] Fundraising-Backlog, § Fundraising Sprint E: Error-specific messaging on "transaction could not be completed" page - https://phabricator.wikimedia.org/T89371#1034870 (atgo) NEW a:CCogdill_WMF [19:25:29] ccogdill: yeah we should review each of the messages with them [19:25:37] and then jessica probably knows how to get them translated [19:25:53] okay so we start with a tech review and then take to Megan and Jessica [19:25:58] right? [19:28:04] for the messaging i think you can go straight to megan/jessica [19:28:36] § Fundraising Sprint Devo, § Fundraising Dash: BUG: Fraud widget numbers don't seem right - https://phabricator.wikimedia.org/T87810#1034893 (Ejegg) @CCogdill_WMF, can you compare the SQL query you're running to get the data directly to the SQL query displayed in the fraud widget? After you set the filters o... [19:29:10] okay.. [19:42:52] Fundraising-Backlog, § Fundraising Sprint E: Error-specific messaging on "transaction could not be completed" page - https://phabricator.wikimedia.org/T89371#1034918 (atgo) [19:46:43] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Error popup is not translated - https://phabricator.wikimedia.org/T89376#1034930 (awight) NEW [19:47:57] awight: we're going on a food quest [19:48:07] mmm! [19:49:47] § Fundraising Sprint Devo, § Fundraising Dash: BUG: Fraud widget numbers don't seem right - https://phabricator.wikimedia.org/T87810#1034956 (CCogdill_WMF) @ejegg, here's the query I use: select * from payments_initial where gateway = 'worldpay' AND country = ‘IL' AND date > TIMESTAMPADD(DAY, -1, NOW()) I e... [19:53:59] ejegg: awight: maybe somehow adding a new dependency to bannerController (mw.Uri) made things load in a different order sometimes, making this false sometimes? https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FCentralNotice.git/82eb579bba58ed5050863642b51cc4b69b9b87fa/modules%2Fext.centralNotice.bannerController%2FbannerController.js#L125 [19:57:04] AndyRussG: did we confirm there actually was an actual increase in hits to BannerRandom? [20:09:29] ejegg: "confirm" is a strong word. Strongly suspect, more like...we caught at least one of them causing the HHMV error [20:10:09] ejegg: https://phabricator.wikimedia.org/P289 [20:11:19] Right, but that could have been a cached thing that's been hitting all along [20:11:28] and only making noise now [20:11:54] i don't think the loading order is to blame [20:15:22] ejegg: could it? From November 25th of last year? at many requests per second? [20:17:58] hmm, i don't know [20:23:47] ejegg|away: the uncensored logs should tell us. We'll see there if there was an increase in SpecialBanneRandom following the deploy, or not [20:35:30] AndyRussG|school: I'd like to know! [20:36:18] ejegg|away: I need some WorldPay test settings help, when you're back... [20:43:40] (PS1) Ssmith: WIP x-by-y data retrieval/parsing [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343 [20:44:13] (CR) jenkins-bot: [V: -1] WIP x-by-y data retrieval/parsing [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/190343 (owner: Ssmith) [20:49:28] awight: whom should I ask? [20:51:49] AndyRussG: I think the unsampled logs are in hive, I would ask on the analytics channel... [20:52:05] awight: Yeah I did... no answer... [20:52:22] awight: but I didn't ping anyone in particular, just called out in an empty field [20:52:36] Make it sound more serious :) [20:53:02] "fatal error" "fundraising"... :) [20:53:10] "your paychecks" [20:53:29] ewulczyn would be a good candidate, he's around today [20:53:34] just not in IRC :-/ [20:54:38] right hmmm [20:54:43] lemme gchat I guess [20:54:46] sorry I don't have names [20:54:51] np :) [20:58:47] hey tech - FYI i will be missing standup and will send an update by email [21:00:10] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Error popup is not translated - https://phabricator.wikimedia.org/T89376#1035151 (atgo) @awight could you give a screenshot? As I recall, when I talked to @k4-713 about this, she said that this is not an error a donor s... [21:06:18] atgo-food: ah OK [21:06:23] :) [21:06:44] ok we've officially waived standup :) [21:06:55] awight: ah OK got.... [21:06:59] I'm working on the DI refactor, almost done with my doPayment patch [21:07:15] Ah hmm [21:07:20] * AndyRussG leaves hangout [21:07:27] ?? we were just in there [21:07:32] oops [21:07:37] not patient enuf! [21:07:50] Maybe I was in the wrong one? [21:07:52] we did the blitz version [21:08:00] donno, there were all kinds of google errors trying to log in [21:08:12] Hmmm [21:08:13] sorry to waste your time! [21:08:21] Ah no waste at all! :) [21:08:42] I'm seeing how to move forward with these deploy things, I'm sure I'll be bugging you more soon [21:08:55] yes, hopefully I can help [21:09:05] it's the least I can do after wasting your week fixing my bugs! [21:18:46] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Expiration date error message is wrong when date < now - https://phabricator.wikimedia.org/T89388#1035175 (awight) NEW [21:29:57] awight: I'm looking at your anti-RandomBanner patch... So far it looks fine, but it also looks like a lot of moving parts [21:30:06] AndyRussG: it really is... [21:30:52] How about just a simple change to the JS, and something to not fill up the error log, for starters? [21:32:14] AndyRussG: I was trying to balance the two approaches... What about this argument--if we're going to want to eventually cut out everything I've removed here, do you think there is merit in deploying a medium size change once, rather than spreading potential destabilizing patches out over many weeks? [21:32:47] awight: yea... however that supposes that all patches can be equally destabilizing [21:32:51] I'm happy to write it either way, but I want to make sure we aren't just taking the conservative road by default. [21:51:35] Wikimedia-Fundraising, Quality-Assurance, Continuous-Integration: Create unit and integration tests for Fundraising extensions to identify breaking MediaWiki changes - https://phabricator.wikimedia.org/T89404#1035380 (bd808) NEW [21:54:34] § Fundraising Sprint Devo, MediaWiki-extensions-DonationInterface, § Fundraising Tech Backlog: Move any remaining controller logic out of the form classes - https://phabricator.wikimedia.org/T86251#1035389 (awight) [21:55:48] (PS8) Awight: Move remaining controller logic into GatewayAdapter::doPayment() [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/189153 (https://phabricator.wikimedia.org/T86251) [22:04:39] (PS1) Awight: WIP move param harvesting to the form layer [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190359 [22:04:53] (CR) jenkins-bot: [V: -1] WIP move param harvesting to the form layer [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/190359 (owner: Awight) [22:15:34] Wikimedia-Fundraising, Quality-Assurance, Continuous-Integration: Create unit and integration tests for Fundraising extensions to identify breaking MediaWiki changes - https://phabricator.wikimedia.org/T89404#1035427 (atgo) Adding fr-tech (feel free to unsub if you'd like) [22:17:08] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Expiration date error message is wrong when date < now - https://phabricator.wikimedia.org/T89388#1035429 (atgo) If this isn't part of the Janelle doc, it should be. Megan said that you guys would look into the messagin... [22:24:35] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Error popup is not translated - https://phabricator.wikimedia.org/T89376#1035452 (atgo) p:Triage>Low [22:28:25] (PS7) Awight: Clean up database switching [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/181238 [22:28:41] ejegg: awight: it's still so mysterious. Looks like there was no change in S.BannerRandom number of requests following our deploy [22:28:47] and it's no where near the number of hhmv log errors [22:29:57] bizarre [22:30:34] hmm, are browser tests hitting real meta somehow? [22:33:47] ejegg: no, they hit the beta cluster. Even so, they're only a few per day [22:34:25] aren't they running all the wmf extension unit tests on each checkin now? [22:35:09] ejegg: ah the qunit tests? [22:35:30] huh [22:35:30] ah, guess not [22:35:51] Well lemme check... [22:37:36] Wikimedia-Fundraising-CiviCRM, § Fundraising Sprint E: Undo a recent merge of two contact records in Civi - https://phabricator.wikimedia.org/T89408#1035484 (atgo) p:Triage>High a:atgo>None [22:37:39] § Fundraising Sprint Devo, § Fundraising Dash: BUG: Fraud widget numbers don't seem right - https://phabricator.wikimedia.org/T87810#1035487 (Ejegg) @PPena, @CCogdill_WMF: The joins explain it all. The widget data is (# of fraud filter rejections) / (# of attempts run through the fraud filter). Your data is... [22:38:01] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Error popup is not translated - https://phabricator.wikimedia.org/T89376#1035488 (awight) {F40501} [22:39:56] atgo-meeting: I was saying :p, I'm also interruptible any time, never concentrating on anything! [22:40:00] ejegg: intersting theory, but running qunit tests on my local setup, I don't get any calls to Special:BannerRandom, or Special:BannerLoader, for that matter [22:40:13] hmm [22:40:20] :P [22:40:36] damn, that WAS a good theory [22:42:59] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Error popup is not translated - https://phabricator.wikimedia.org/T89376#1035490 (awight) I see that the validation errors are translated correctly, so maybe this only happens for internal errors? [22:46:05] ejegg: awight: I think we got it. It looks like the difference in volume is due to lots of retries from varnish following the 503 [22:46:15] oho! [22:46:24] That was a crazy suprise ending... [22:46:39] and... the chicken dies. The end. [22:47:01] wait, so bannerRandom was failing in a way that made varnish think it would work if it just tried hard enough? [22:47:10] * ejegg reads -operations [22:47:22] That's what qchris found following quick grep [22:47:32] It might try other nodes on the cluster [22:47:44] I mean, I have no idea, but if I were varnish, it'd cross my mind [22:48:10] The moral of the story: I'll continue reviewing awight's ELO for BannerRandom, and unless a reason not to crops up, we'll go with that? [22:48:31] yeah, I was looking at that too [22:48:42] didn't see anything wrong at first glance [22:50:07] yeah me neither [22:50:11] § Fundraising Sprint Devo, Wikimedia-Fundraising-CiviCRM: BUG: Undo a recent merge of two contact records in Civi - https://phabricator.wikimedia.org/T89408#1035504 (atgo) [22:50:32] First tho lemme hack a bit of RecordImpression then to see if we agree on an approach [22:50:41] sound good? [22:50:42] § Fundraising Sprint Devo, Wikimedia-Fundraising-CiviCRM: BUG: Undo a recent merge of two contact records in Civi - https://phabricator.wikimedia.org/T89408#1035472 (atgo) Ok so I know this isn't technically a bug, but I think it fits in the criteria for our purposes (interruption rather than planned). [22:51:47] sure - you're starting by just not logging when there's no choiceData? [22:56:14] § Fundraising Sprint Devo, MediaWiki-extensions-DonationInterface, § Fundraising Tech Backlog: Move any remaining controller logic out of the form classes - https://phabricator.wikimedia.org/T86251#1035530 (awight) Oops... I still need to refactor the resultswitcher pages, but it's less urgent, cos these will... [22:59:03] (PS1) Ejegg: Merge branch 'master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/190371 [22:59:22] (CR) jenkins-bot: [V: -1] Merge branch 'master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/190371 (owner: Ejegg) [23:04:31] § Fundraising Sprint Devo, Wikimedia-Fundraising, MediaWiki-extensions-DonationInterface: Tech Review of LATAM processor - https://phabricator.wikimedia.org/T87046#1035542 (awight) OKAY, now we're reviewing the "Streamline API v1.7" document. This integration will get us credit card, online bank payment, Bol... [23:05:35] § Fundraising Sprint Devo, Wikimedia-Fundraising, MediaWiki-extensions-DonationInterface: Tech Review of LATAM processor - https://phabricator.wikimedia.org/T87046#1035547 (awight) Open>stalled (I'm just using the "stalled" status to indicate that we're waiting for answers here.) [23:06:13] (PS2) Ejegg: Merge branch 'master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/190371 [23:13:06] § Fundraising Tech Backlog, § Fundraising Sprint E: BUG: GC CC donations inconsistent in when they reach Civi - https://phabricator.wikimedia.org/T89016#1035575 (atgo) [23:14:20] atgo: I don't know about this "BUG" thing, it's sort of distracting me :) [23:14:33] Should we relabel the RecordImpression thing a bug also? [23:14:33] aaww [23:14:40] and I like bugs :) [23:15:17] Not to be a total throwback, but the old regime sort of worked, where "Emergency" priority meant that it snuck up on us. [23:15:24] Maybe it's a crustacean [23:15:42] That would go along with our shrimp theming [23:16:12] iinteresting [23:17:16] awight: ejegg: BTW qchris is helping dig more into the Special:BannerUnpredictable on #wikimedia-analytics [23:18:27] hehe [23:26:29] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Client-side validation breaks credit card button - https://phabricator.wikimedia.org/T89412#1035620 (awight) NEW [23:29:04] (CR) Ejegg: [C: 2] Merge branch 'master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/190371 (owner: Ejegg) [23:36:02] Special:BannerIHaveNoIdeaJustGiveMeWhatever [23:46:35] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Client-side validation breaks credit card button - https://phabricator.wikimedia.org/T89412#1035679 (atgo) @awight I believe this is why the "continue" button appears. Are you saying that there's no solution or that wha... [23:51:31] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Client-side validation breaks credit card button - https://phabricator.wikimedia.org/T89412#1035689 (awight) Oh, interesting... Well, now that the radio buttons are styled like pushbuttons, they should have behavior con... [23:52:47] AndyRussG: ejegg: anybody need CR? [23:52:51] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Client-side validation breaks credit card button - https://phabricator.wikimedia.org/T89412#1035697 (atgo) p:Triage>Low [23:53:08] awight: nope, just some composer advice. composer really doesn't like vendor folders to be actual submodules - it wants their .git to be a real gitdir. [23:53:10] MediaWiki-extensions-DonationInterface, Wikimedia-Fundraising, § Fundraising Tech Backlog: Client-side validation breaks credit card button - https://phabricator.wikimedia.org/T89412#1035620 (atgo) Thanks! Given that this isn't blocking payments, I'm rating it low. I agree we should do something about it when... [23:53:18] AndyRussG: I remember you were saying something about CentralNotice patches getting stale, but I don't see them. maybe they floated off the recent queue? [23:53:27] ejegg: aargh yeah [23:53:30] so I'm about to make those things in crm/vendor/wikimedia back into regular directories [23:53:40] mmm ok sure [23:53:49] it's either that or manually update composer.lock, which seems like a terrible idea [23:53:56] awight: if ur bored you could start peeking at my WIP CN tests... I'm just about to send a modified patch on ejegg client sampler to only sample when there'z no choice data [23:53:58] please test migrating that locally, though... I do worry that we might all have to rm -rf to recover [23:54:37] ejegg: well, afair the alternative is to just wrangle the submodule situation manually. I see what you mean about the lockfile, though. [23:54:44] AndyRussG: sure! [23:55:09] awight: ah, guess I could temporarily move the submodules.. [23:55:29] heinous! [23:55:55] awight: thanks! a lot more tests fixtures are going back in there in a bit, including stuff that was in there and is temporarily out (while I refactored the data struct) https://gerrit.wikimedia.org/r/#/c/188687/ [23:56:09] ejegg: grr, I hate when tools try to own us. c.f. computers [23:56:23] * AndyRussG takes "heinous" as a complement [23:56:55] those submodules are definitely wrong. I mean, we're checking in copies of all the third party code from other repos [23:57:03] AndyRussG: hehe, that jab was meant for Composer [23:57:29] Ah hmmm u should try Maven [23:57:36] * awight short-circuits [23:57:45] I've been there :[ [23:58:40] mvn2-managed project that used Rhino, a port of the mozilla engine for running javascript written in Java. [23:58:49] the horror! [23:59:04] ejegg: what do you mean, the submodules are wrong? [23:59:31] vendor shouldn't have submodules under it if we want composer to work/