[00:01:23] (PS1) Ejegg: Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/444144 [00:01:43] (CR) Ejegg: [C: 2] Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/444144 (owner: Ejegg) [00:02:50] (PS1) Ejegg: Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/444145 [00:02:58] (CR) Ejegg: [C: 2] Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/444145 (owner: Ejegg) [00:07:09] (Merged) jenkins-bot: Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/444144 (owner: Ejegg) [00:07:11] (Merged) jenkins-bot: Merge branch 'master' into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/444145 (owner: Ejegg) [00:08:22] !log updated CiviCRM from e2c8ddd70e to d24c2bc08b [00:08:24] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [00:23:24] Fundraising Sprint Lactose is unusually tolerant, Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog: Write a specification for mapping banner/landing page impression event properties -> database schema - https://phabricator.wikimedia.org/T196563 (Ejegg) OK, the landing page sp... [00:23:28] Fundraising Sprint Lactose is unusually tolerant, Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog: Write a specification for mapping banner/landing page impression event properties -> database schema - https://phabricator.wikimedia.org/T196563 (Ejegg) OK, the landing page sp... [13:19:32] good morning fr-tech! i tried to post a WIP commit at the end of the day yesterday but composer fix wasn't working.. [13:19:56] it's still just hanging for some reason [13:22:58] morning mepps [13:23:05] sorry to hear you're having issues with composer [13:23:38] hmm it seems to be all composer commands.. [13:24:57] I just performed a composer update on a local project to check at my end and it's pulling down from github updates to symfony [13:25:18] the repo is github [13:25:45] (for the project I updated) [13:26:09] https://getcomposer.org/doc/03-cli.md#diagnose [13:27:01] gotta pop out [13:32:49] i'm confused, it's inheriting the source for updates? [13:43:34] urg i don't know why this is happening [13:44:17] (PS3) Mepps: WIP: Handle no status code for Ingenico api [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/440465 (https://phabricator.wikimedia.org/T194517) [13:46:44] (CR) jerkins-bot: [V: -1] WIP: Handle no status code for Ingenico api [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/440465 (https://phabricator.wikimedia.org/T194517) (owner: Mepps) [14:59:00] and i finally restarted my computer and composer is working again [14:59:23] wth lol [15:01:35] I had a late start today mepps, Roxy has run out of the preggiepops and she had another spell this morning [15:01:50] ejegg jgleeson if you look at my patch you can see the error i'm struggling with--actually i'm wondering, why are we going to approve payment if i mark it pending? [15:02:00] oh no jgleeson :/ that's rough [15:02:10] how many weeks along is she? [15:02:30] 11 I think [15:02:35] or we think [15:02:42] oh good, hopefully she'll feel better in a couple weeks then [15:02:48] but this is a rough time [15:03:08] just pulling down your patch to run the tests locally [15:10:38] can't see anything obvious mepps, I'm gonna do a cli debug run through the test to see if I can work out where it's missing the id [15:10:58] unless you've already solved it [15:11:12] well jgleeson it's missing the id in approvePayment [15:11:37] mepps I'll take a look [15:11:57] the issue is that if a payment is in_progress it doesn't return the whole payment object, just a json array {status: IN_PROGRESS} [15:12:21] mepps ahh, but we should only do 'approvePayment' if the status maps to PENDING_POKE.... [15:12:22] so i assume that approvePayment expected to get the id from the getStatus call [15:12:31] ejegg interesting... [15:12:37] why is this happening then? [15:15:21] mepps oh heck, looks like that gnarly transactionConfirm_CreditCard function doesn't actually call approvePayment within that switch on status [15:16:18] it relies on cases FAILED/REVISED/COMPLETE to set a problemflag or a cancelflag [15:17:04] argh, we really should have done this from scratch [15:18:47] so.... wtf to do here? [15:21:10] mepps you see the issue? [15:21:31] We call the get_orderstatus stuff in a loop (for whatever reason) [15:22:01] then break out of the loop with problemflag=false and cancelflag=false when we see a status of PENDING or PENDING_POKE [15:22:23] I got that far [15:22:38] and it fails within $this->do_transaction( 'SET_PAYMENT' ); [15:22:41] then outside of the loop, we call the approvePayment because both of those flags are false [15:22:54] wrapped by approvePayment [15:22:54] jgleeson right, or the ingenico equivalent [15:23:06] approvePayment is overridden in the ingenico adapter [15:23:16] to call do_transaction( 'approvePayment' ) [15:23:30] got it [15:23:47] but... when the status maps to PENDING, that means we're waiting for the donor or their bank [15:24:12] yes [15:24:16] status PENDING_POKE is the only one where we need to call approvePayment [15:24:39] anyway, so this logic seems like it's been wrong for the old API too [15:24:40] so this must be an issue for other places with ingenico both old and new, right ejegg? [15:24:43] right [15:25:12] so.... [15:25:15] I'm guessing we've gotten away with this due to the transaction not staying in a pending state for long? [15:25:24] jgleeson: I guess so [15:25:59] so should i re-do confirm cc for new ingenico or try to continue patching gc? [15:26:04] that, and the old API somehow not erroring as hard when we call SET_PAYMENT and aren't allowed to [15:26:12] mepps argh [15:26:19] that's a tough choice [15:27:05] A rewrite sure wouldn't be ready for another 1 hr test next week :( [15:27:21] but continued patching is what got us here [15:28:17] want to spend a couple hours looking at what a from-scratch version of that fn would be? [15:28:34] maybe just overriding that fn in the ingenico adapter? [15:29:20] I feel like we don't even need those loops, since we do api call retries at a lower level [15:29:46] sure ejegg [15:30:13] yeah the patch here would be very hacky unless i guess i add another flag called $pendingFlag? [15:30:28] argh, yeah, those loops are never even looped, right? [15:30:39] i don't think so, i was wondering about that [15:30:40] ALL of those switch cases end in 'break 2' [15:30:51] well hell [15:31:20] can we just drop the loops and the flags, and put the further processing in its own fns called directly in the switch cases? [15:38:14] hmm in my reading there is one case where we need the loop, so yeah that would make sense [15:38:32] where do we need the loop? [15:38:47] the do_finishpayment? [15:39:53] so in my reading if the case is pending poke there might be a case it would retry [15:40:03] i mean if the finalstatus is pending poke [15:43:03] but that's a pretty intense loop for that one case [15:46:53] right, and it's not even looping [15:47:09] that // FIXME:: explicit that we want to fall through comment [15:47:23] calls out the fact that it's actually going right through to the PENDING one [15:47:30] (CR) Jgleeson: [C: 2] "Tested again to confirm new 'skipAuthentication' field being sent to Smashpig during 'createHostedCheckout' call, which it is. Looks good " [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/441483 (https://phabricator.wikimedia.org/T176512) (owner: Ejegg) [15:47:30] where we break 2 anyway [15:47:48] mepps, I think we should unroll that bit [15:48:10] huh... [15:48:25] to a check on status code at the start [15:48:27] hmm [15:48:28] clearly i've wandered down the rabbit hole ;) [15:48:46] ...and found dragons [15:49:22] mepps, I thought that when ejegg suggested a refactor [15:49:23] (Merged) jenkins-bot: skipAuthentication flag disables 3DSecure for Connect [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/441483 (https://phabricator.wikimedia.org/T176512) (owner: Ejegg) [15:49:28] I thought "good luck with that" [15:49:29] :) [15:49:44] I hope you had good coffee [15:49:54] :) aww [15:50:08] always! [15:50:19] it's step one of every day [15:50:27] this definitely feels like a moment to consider the toxic code protocol [15:50:41] and not just add another band-aid [15:51:04] yes, there are times to be a hero and fix it [15:51:32] I applaud the courage! [15:52:20] * jgleeson imagines mepps charging at the while loop with a shield and sword in some braveheart-like fashion [15:53:16] for loop even [15:54:04] heh [15:55:01] so... this checking for status code 200 is a leaky abstraction [15:55:25] 200 is mapping to PENDING_POKE [15:56:04] but should map to a different normalized thing if it needs different treatment [16:55:21] (PS3) Ejegg: Just an extraction. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443886 (owner: Eileen) [16:55:26] (CR) Ejegg: [C: 2] Just an extraction. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443886 (owner: Eileen) [17:02:04] (Merged) jenkins-bot: Just an extraction. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443886 (owner: Eileen) [17:29:03] how have I not see this!!! https://www.youtube.com/watch?v=H5QCbogbOuc [17:29:13] I can't stop laughing dstrine [17:29:47] jgleeson: you're welcome :) [17:30:06] I really want one of those costumes [17:36:03] Fundraising Sprint Lactose is unusually tolerant, Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, Patch-For-Review: donor data export tool for civi - https://phabricator.wikimedia.org/T196511 (Ejegg) Huh, it's all mashed together for m... [17:36:34] (CR) Ejegg: "it's better than the blankness the blue printer icon gives me, but it's still nearly impossible to read. see pic uploaded on phab ticket." [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443542 (https://phabricator.wikimedia.org/T196511) (owner: Eileen) [17:39:40] dstrine, would it be appropriate to turn up to standup in one? [17:40:11] jgleeson: If you want to. You do you man ;) [17:40:35] I guess standup fancy dress could be a thing [17:40:41] what do you think XenoRyet ? [18:17:48] Maybe I should get one for retros [18:17:52] Dino-retro [18:19:15] :D [18:19:34] have a good weekend fr-tech, over and out [18:19:42] See you later [20:21:06] Fundraising-Backlog, MediaWiki-extensions-CentralNotice: CentralNotice api call generates lots of warnings - https://phabricator.wikimedia.org/T198996 (Ejegg) [20:23:21] Fundraising-Backlog, MediaWiki-extensions-CentralNotice: CentralNotice api call generates lots of warnings - https://phabricator.wikimedia.org/T198996 (Ejegg) Seems to involve BannerMessageGroup::registerGroupHook() Could this have to do with T198869 ? [20:42:17] Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, MediaWiki-extensions-CentralNotice: [Bug] CentralNotice: "Failed to load data blob" error when editing translatable messages - https://phabricator.wikimedia.org/T198869 (Ejegg) Here's the list of all the logstash things from t... [20:49:28] (PS1) Mepps: WIP Refactor of ConfirmCreditCard [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/444311 (https://phabricator.wikimedia.org/T194517) [20:51:37] (CR) jerkins-bot: [V: -1] WIP Refactor of ConfirmCreditCard [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/444311 (https://phabricator.wikimedia.org/T194517) (owner: Mepps) [20:53:19] mepps oh interesting, are we not getting a good code from the Ingenico Connect responses? [20:54:49] uups, this library closes early fridays. back in a bit! [21:00:26] fr-tech not the most productive afternoon, but i tried pulling out that loop and learned i can still complete a payment successfully, still need to work on this pending/approvePayment bug [21:00:29] and with that, i'm off [21:37:48] cwd: hey, if you're around, quick question... also, sorry to bug you about this 'cause I've asked before... 8p can u point me to the code that filters the current (legacy) firehose for the files created to be read by DjangoBannerStats? [21:38:31] Just wanted to check especially exactly what criteria from the general firehose we use to create the LandingPage ones.. [21:38:35] thx in advance!!! [21:38:45] AndyRussG: sure thing, do you have puppet checked out? [21:40:23] frpm1001:/var/lib/git/puppet.git [21:40:57] if you look at this file: hieradata/site/eqiad.yaml [21:41:02] and search for kafkatee [21:41:05] i think that's what you need? [21:48:43] cwd: hmmm looks close... I see it references a file, /usr/local/bin/landingpages-filter [21:49:00] should I look at that file then? [21:50:33] on americum? [21:50:57] AndyRussG: yep! [21:56:48] cwd: cool, got it!! thx :) [21:57:19] np! [21:58:04] hmm perl [21:58:09] Also, it sez, "# this is a test filter, looking for where the banner log pipeline breaks down" [21:58:14] is that right? [21:58:17] cwd: ^ [21:59:29] Or maybe the real filter on the firehose is upstream, before it gets to frack? [22:02:58] Fundraising Sprint Lactose is unusually tolerant, Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog: Write a specification for mapping banner/landing page impression event properties -> database schema - https://phabricator.wikimedia.org/T196563 (AndyRussG) OK, will add that!... [22:13:34] AndyRussG: looking [22:13:46] cwd: thx! [22:14:34] yeah that is mysterious [22:14:40] AndyRussG: did you look at weblog-filter ? [22:15:08] it doesn't look like a test filter to me [22:15:12] it looks like an actual one [22:15:29] cwd: where? [22:15:56] Ah I see same dir [22:16:16] yessir [22:17:42] hmmm [22:17:47] doesn't say how it's run [22:17:55] it has a command line arg to say what to filter [22:19:04] I guess it's not really urgent! [22:19:14] There are also other ways to check... [22:21:07] Fundraising Sprint Karma chameleons hide amongst us, Fundraising Sprint Lactose is unusually tolerant, Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, Patch-For-Review: Help switch over foundation pages to donate wiki - https://phabricator.wikimedia.org/T193663 (Krin... [22:22:41] ejegg: If I had to guess which change in core is the culprit, I'd imagine this one? https://github.com/wikimedia/mediawiki/commit/33258e04d1b84ce78cae64e97318c6f684adb987#diff-a0f7feeaae57e9d2c735c8919c16ad15 [22:23:00] oh? [22:23:25] 10 days ago, right timeframe [22:26:22] yurp [22:26:47] so we should try playing with MCR migration flags locally [22:27:33] scanning over that, nothing stands out as affecting CN namespace stuff in particular tho [22:28:09] ok, i guess I need to understand what exactly changes under MCR [22:29:18] AndyRussG: just forwarded you a messages from the ops list sent by Kinzler Jun 14th about MCR patches landing in core [22:30:15] Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, MediaWiki-extensions-CentralNotice: [Bug] CentralNotice: "Failed to load data blob" error when editing translatable messages - https://phabricator.wikimedia.org/T198869 (AndyRussG) Hmmm yeah...!! Comparing the stack traces fro... [22:31:02] ejegg: okok cool thanks! yeah looks very relevant [22:31:28] ooh, just read your comment. does feel like we're getting closer! [22:31:40] I think this is person to e-mail about this... I'll try adding him on the Fab task... (Didn't see him on IRC, though maybe for lack of knowing his handle) [22:31:46] yurp [22:31:55] (CR) Eileen: "ejegg - no it really is that ugly :-(" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443542 (https://phabricator.wikimedia.org/T196511) (owner: Eileen) [22:32:27] ejegg: gonna afk for about 20 min, back in a bit! [22:32:47] (PS5) Ejegg: Add basic print button to the forgetme screen. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443542 (https://phabricator.wikimedia.org/T196511) (owner: Eileen) [22:33:07] (CR) Ejegg: [C: 2] "Better ugly than invisible!" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443542 (https://phabricator.wikimedia.org/T196511) (owner: Eileen) [22:39:40] (Merged) jenkins-bot: Add basic print button to the forgetme screen. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443542 (https://phabricator.wikimedia.org/T196511) (owner: Eileen) [22:40:04] dang, my local mediawiki master setup broke on reboot... [22:41:03] no, i just mangled a db connection string when trying to add the MCR var to localsettings [23:04:08] ejegg: an omen? [23:08:28] heh, don't think so [23:09:00] well, I don't get to the same place stepping thru locally, probably because of no wikifarm settings [23:10:08] on my machine $canUseTitleNewFromId in RevisionStore::getTitle ends up true [23:10:23] and I think it needs to be false to get the issue. [23:10:49] could just hack that one var, but it'll probably depend on other multi-wiki things [23:16:04] yeah, that's not sufficient [23:16:10] hmmm [23:19:13] Is that where local execution diverges from the stack trace 4 u? [23:19:24] yep yep [23:19:28] hmm [23:19:29] K [23:19:36] I feel I need to read up on this MCR stuff [23:19:55] (glad u found that e-mai!) [23:20:38] yeah, me too [23:21:09] all i know is that it's supposed to help track the content of a page as a whole [23:21:25] so if pictures in the page change, that's a new revision [23:21:31] not sure how templates are handled [23:25:46] (PS2) Ejegg: Enotice fix - much as I would like it I can't give all fields titles in the metadata. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443887 (https://phabricator.wikimedia.org/T195611) (owner: Eileen) [23:25:50] (CR) Ejegg: [C: 2] Enotice fix - much as I would like it I can't give all fields titles in the metadata. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443887 (https://phabricator.wikimedia.org/T195611) (owner: Eileen) [23:26:08] Oh hmmm [23:31:49] (Merged) jenkins-bot: Enotice fix - much as I would like it I can't give all fields titles in the metadata. [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/443887 (https://phabricator.wikimedia.org/T195611) (owner: Eileen) [23:34:23] ejegg: https://docs.google.com/presentation/d/1nWe0Zuv26X-jOVXAokeNa8mODeMWR4qpNkvnNSIFKFw/edit#slide=id.p5 [23:35:00] Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, MediaWiki-extensions-CentralNotice: [Bug] CentralNotice: "Failed to load data blob" error when editing translatable messages - https://phabricator.wikimedia.org/T198869 (Ejegg) @daniel We added you because we're thinking this... [23:35:08] thanks! [23:35:15] :) likewise! [23:35:34] hmm it's not heavy on detail [23:36:11] sure looks like a massive change [23:38:23] also https://wikifarm.wmflabs.org/mcr/index.php/Main_Page#tab=Pert_Chart [23:39:54] ooh, nice trove [23:43:55] I'm guessing nothing has actually been enabled [23:44:13] Don't see any suspect changes in config, which is how I'd imagine something like this would be set to go out [23:44:40] Let's see, this must have gone out last week, since there was no train this week [23:45:52] (by "this" I mean, whatever the breakin' change was) [23:50:19] changes to this file last week are pretty massive: git difftool origin/wmf/1.32.0-wmf.8 origin/wmf/1.32.0-wmf.10 -- includes/Storage/RevisionStore.php [23:52:00] addshore: any thoughts on this bug? Seems related to recent changes in RevisionStore... https://phabricator.wikimedia.org/T198869 [23:52:09] thx in advance! [23:53:00] Fundraising Sprint Matt Damon to head up Space Force, Fundraising-Backlog, MediaWiki-extensions-CentralNotice: [Bug] CentralNotice: "Failed to load data blob" error when editing translatable messages - https://phabricator.wikimedia.org/T198869 (AndyRussG) Also adding @Addshore for the same reason, th...