[05:17:52] (PS1) AndyRussG: Improve squidline regex [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168932 [05:20:32] (CR) AndyRussG: "Works unless the UA field contains a tab (probably never really the case)." [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168932 (owner: AndyRussG) [05:21:12] (CR) AndyRussG: "See https://gerrit.wikimedia.org/r/#/c/168932/ for regex follow-up." [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168343 (owner: AndyRussG) [15:29:54] (CR) Ejegg: [C: 2] Improve squidline regex [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168932 (owner: AndyRussG) [15:39:08] (CR) Ejegg: "Not seeing the chartjs update - did you want to check in a submodule update too? Also, want to add the big-english component to the gulpf" [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/168603 (owner: Ssmith) [15:39:14] (CR) Awight: [C: 2] "Nice cleanup, definitely an improvement. I'm wondering, though, weren't there fields other than XFF which were sometimes completely empty" (2 comments) [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168932 (owner: AndyRussG) [15:40:22] (CR) Awight: [V: 2] Improve squidline regex [wikimedia/fundraising/tools/DjangoBannerStats] - https://gerrit.wikimedia.org/r/168932 (owner: AndyRussG) [15:46:06] hey pizzzacat :) [15:47:18] hola atgo [15:47:49] you're doing form things today, ya? [15:48:00] yep [15:48:15] that's what we agreed, right? [15:48:33] * awight puts up shields [15:49:06] no just making sure, since things change and I sometimes don't know it :) [15:49:31] * awight quickly grabs a card from the current sprint in mingle and feigns busy :p [15:49:44] I made a change but am in gerrit hell again [15:49:50] oh god [15:49:59] want me to remotely reboot for you? [15:50:13] whenever jenkins merges, it doesn't add a commit ID [15:50:23] ugh, yeah [15:50:26] which prevents me from being able to easily push gerrit reviews [15:50:42] wait, the jenkins automerge is an issue? [15:51:03] this is for Donation Interface ejegg [15:51:10] but yeah [15:51:13] ejegg: pizzzacat: hey. I want to log into the dash. Is that a thing [15:51:28] I think so? ejegg would know the status on that right now [15:51:41] I'm working on form stuff until it's finished so will be off the Dash [15:51:47] atgo ^ [15:51:49] awight: yeah, go for it! Just go to https://dash.frdev.wikimedia.org and click the log in button [15:52:01] yes pizzzacat [15:52:07] ahh that's so anooying [15:52:09] just seeing if you got my response [15:52:11] :) [15:52:15] yeah [15:52:23] ok so yes, forms. just wanted to check if you'd seen this as well: https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/2084 [15:53:06] pizzzacat ^ [15:53:10] no but I haven't looked in Mingle yet bc responding to the stuff Megan sent this morning [15:53:23] whatever I should do, should just be in this sprint in Mingle [15:53:36] if that's cool atgo [15:53:44] because otherwise I can't keep track :( [15:53:47] yeah sounds good [15:53:53] ejegg: pizzzacat: AWEZSOME [15:53:55] that's why I did this email thing first so I wouldn't lose it [15:53:57] adding roight now [15:53:58] but, it was visible before I logged in... [15:54:36] cool thanks atgo! [15:54:55] awight what's going on over there? [15:56:11] pizzzacat: I'm looking at DASH [15:56:18] yay [15:56:39] pizzzacat: only issue so far is, being logged out does not hide the interface [15:57:09] awight do you recommend me going through each commit that doesn't have a commit ID from jenkins, getting those ID'd, and then reviewing them, even though they're already in gerrit?? or is there an easier way? This is the second time I'm dealing with this this week :( [15:57:26] ah, I think that's an ejegg thing since he is working on that stuff [15:57:27] oh [15:57:28] god [15:57:45] they are in gerrit w/o change-ids? [15:57:48] I have a login flow branch [15:57:52] send an example URL? [15:58:00] umm I assumed so since I pulled [15:58:05] one sec [15:58:55] & sorry, in a meeting for a bit [15:59:17] it's unclear how to get that [15:59:21] I am blocked on this btw [15:59:35] want merge numbers? argh [16:00:40] pizzzacat: want to send me a zip of your local DonationInterface? Or just point me to the stuff in gerrit that's giving you trouble? [16:01:33] https://gerrit.wikimedia.org/r/#/c/166607/ [16:01:38] that's the first one [16:02:44] So it has a change id - what kind of trouble is that commit giving you? [16:02:54] ugh I extrapolated incorrectly [16:02:58] that's not the issue. [16:04:20] ejegg when I do a git log after gerrit complains about no change ID, I see several jenkins commits with no change ID's [16:04:33] so that made me think it was mad about those. but actually it's not. [16:04:38] ok [16:04:44] it's mad about a merge commit I did before that [16:04:54] oh, you might just need to rebase [16:05:28] So, I bet you had done some work locally, and then pulled master [16:05:34] yeah exactly [16:05:39] blah [16:05:44] so where do I rebase? [16:06:01] pizzzacat: sorry, it's the "pull" that messed you up [16:06:10] so first I fetched all [16:06:16] first of all, can you "git review -s" in the dir... [16:06:19] which is right, right? [16:06:30] yah fetch is good but pull is sketch [16:06:31] what does that do awight? -s [16:06:47] ok so when i fetch, what's my next step? because it doesn't do what I expect or want [16:06:47] it sets up the gerrit branch in your local, and installs hooks [16:07:01] yeah it only retrieves, no merging [16:07:14] instead of just doing a pull, do a pull --rebase [16:07:46] ok so my steps are 1) fetch all [16:07:55] 2) git pull —rebase (?) [16:08:10] yep! [16:08:17] err in that case, the fetch is redundant [16:08:28] so from now on I do it that way and I don't create 800 problems for myself [16:08:44] oh ok so just step 2? [16:08:48] yeah [16:08:55] I defer to ejegg on --rebase though [16:08:55] ah ok. sweet. that I can do. [16:09:01] the pull --rebase makes sure the new stuff from gerrit comes into your local copy the right way [16:09:33] so I won't run into the thing where I need to create a merge commit that gerrit then hates and rejects [16:09:55] without the rebase, it leaves your new local changes in, then tries to put the stuff from gerrit on top of those [16:10:09] ohh ok [16:10:14] gotcha [16:10:15] creating a new version of each gerrit commit [16:10:34] the rebase just says to take all the gerrit stuff, then replay your changes on top of them [16:10:50] great, that's exactly what I want and thought I was doing [16:11:27] so where I'm at now, what do I rebase? [16:11:41] I have my change, then under that the merge commit [16:11:53] but then like way down I have old merge commits that it also hates [16:12:31] Hmm, maybe just note the commit hashes of the local changes you want [16:12:57] check out a fresh branch tracking origin/master [16:13:06] make sure that's up to date with origin, [16:13:07] oh like we did the other day [16:13:14] and cherry-pick in the local stuff [16:13:20] yeah, that might be easiest [16:13:25] ok I *think* I remember those steps [16:13:40] rename -m oldbranch newbranch [16:14:36] another possibility is to do an interactive rebase and get rid of the merges, then do a pull --rebase. I think that would fix it in your existing branch, but awight may have a better method [16:15:02] ohh interesting [16:15:59] git rebase -i HEAD ? [16:16:36] or even ? [16:16:39] you can leave off the HEAD, but add a ^ on the end of the commit it's complaining about [16:16:51] yeah, actually, the bottom of the list is good [16:17:01] that way I can get rid of all offenders at once [16:17:03] with the ^ on the end for good measure [16:17:05] yeah [16:17:15] ok sweet [16:18:06] whoa I got like a million lines [16:18:29] huh, goes back pretty far, I guess. [16:18:43] hmm, I hope this works with merges [16:18:59] hmm don't see the offenders [16:19:00] I realize I've only done the interactive rebase with regular commits [16:19:12] oh, maybe i was wrong about using it for merges [16:19:27] dang, sorry [16:19:32] ah ok. no worries. [16:19:39] so a fresh branch + cherry-pick is probably the way [16:20:23] word [16:20:38] so remind me the tracking syntax? [16:20:43] setting up a new master [16:20:53] git branch --track newmaster origin/master [16:22:02] once you get a clean new master, I recommend making branches off of that to do all your local changes [16:22:23] so you're only ever pulling into master stuff that's merged into jenkins master [16:22:42] and then can cherry-pick into master when I'm ready to push to review? [16:23:10] nope, just push to review straight from the side branch [16:23:28] oh, will it go into regular review? [16:23:44] yeah, that's determined by the .gitreview file [16:23:54] I think there's a defaultbranch in there [16:24:09] which is master for everything except deployment branches [16:24:17] ok, cool [16:24:35] that flow really helps when you're working on a couple unrelated changes [16:25:05] so you can just jump between the side branches and keep updating version of your patchsets [16:25:07] ok! I will try it next time. it helps me keep it organized in my head too [16:25:36] I can't guarantee I won't mess it up the first time or two :-/ but will try! [16:25:45] best of luck! [16:25:59] so now I just do git cherry-pick and git review and I'm set? [16:26:07] yeah [16:26:32] (PS1) Ssmith: Form width/label height now fit w/mediawiki update [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/168985 [16:26:42] yay [16:26:50] thanks again ejegg! [16:27:03] will work on a new branch now, for the worldpay form updates. [16:27:15] might bug you out of worrying that I'm doing it wrong, later [16:27:15] ahh, css needs tweaks for 1.23? [16:27:28] well [16:27:40] the expiration date translation was pushing things around [16:27:53] ah, word [16:27:57] because the form width was too narrow to accomodate the new .875em font size [16:28:05] (CR) Ejegg: [C: 2] Form width/label height now fit w/mediawiki update [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/168985 (owner: Ssmith) [16:28:27] (Merged) jenkins-bot: Form width/label height now fit w/mediawiki update [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/168985 (owner: Ssmith) [16:28:39] so I both updated the width and then also made it so that pushing won't happen in the future if the translation is longer than it is now [16:28:45] thanks! [16:28:59] thanks for the explanation [16:29:01] Morning! [16:29:03] * AndyRussG waves [16:29:07] Hi AndyRussG [16:29:13] :) [16:29:58] the-wub: hi! is there anything technical (like maybe some banner JS you're working on?) that maybe I should look at ahead of our meeting? [16:30:00] (CR) Ejegg: [C: 2] include DonationInterface in Composer config [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/167320 (owner: Awight) [16:31:06] hi AndyRussG. I don't think so, just using the same code as previously for our next tests [16:35:48] the-wub: as in: B14_1015_enBE_dsk_full_y1, B14_1015_enBE_dsk_full_y2 and B14_1015_enBE_dsk_full_nn? [16:36:30] Ah I see more stuff now in the invite... [16:37:41] AndyRussG: yes, they're the ones. there will be one slight change: we'll assign people to bucket C or D randomly (rather than A=>C and B=>D) [16:40:30] the-wub: launching when? [16:41:24] AndyRussG: we aren't launching anything yet. I'm still working on banners and megan wants to get the bucket duration increased first [16:41:39] OK yes that sounds right :) [16:59:29] hey atgo, any tips on the easiest way to see all worldpay cards in mingle? [16:59:35] mmhmmm [16:59:38] hang on a sec [16:59:56] there's a supertask. you could filter tech tasks by that [17:00:05] ok cool [17:00:13] I will just move them into the current sprint, is that ok? [17:00:59] yeah sounds good [17:01:00] atgo is the supertask "form cleanup"? [17:01:08] mm... no. turns out i was wrong as i'm tinkering [17:01:12] with filters [17:01:27] wrong about which? [17:01:35] the supertasks [17:01:39] oh ok [17:01:46] * atgo shakes fist [17:01:51] miiiingle!! [17:01:52] just not clear on where they all are [17:01:54] word. [17:01:55] srsly! [17:02:07] so 2083 is the one jessica made [17:02:12] and as Aerosmith put it, I don't wanna miss a thing [17:02:21] hehe [17:02:58] 2084 is for the state field not having a label [17:03:23] ok [17:03:40] width issue i just saw and haven't made a card for, would oyu mind? [17:03:53] I will be working on those but if there are any others you can find, do you mind putting those in? [17:03:56] sure I can make that one [17:03:57] yeah [17:04:01] thanks! [17:04:04] i'm... 99% sure they're all in that one [17:04:07] ah ok [17:04:11] because i had jessica make it :) [17:04:12] cool [17:04:16] the one with the slide deck [17:04:32] so I will operate under the assumption that there are two cards [17:04:39] 2083 + 2084, yes [17:04:53] perfect [17:05:13] + we should track the width issue, so we should have a third [17:07:13] yep yep [17:07:16] doing it now [17:11:58] danke [17:17:43] the-wub: just saw the backscroll--I don't understand how it's OK to assign to C and D randomly... [17:17:50] So there are 4 groups... [17:18:00] can you discriminate between them? [17:18:27] the-wub: also, thanks for the burrito tunnel essay :) [17:19:45] awight: we want to be able to run "top banner" tests on buckets C & D too. so for that we need to split users randomly between them [17:20:15] But can u tell which treatment people get? [17:21:11] no. but it shouldn't matter too much, so long as we have enough people [17:21:30] I don't get it... [17:23:10] I'm not sure what the problem is. It's no worse than what we have at the moment, where we can't tell what banners were seen previously [17:24:05] There's probably a way you can get this data. [17:24:39] Otherwise, you could create separate tests to compare A/B, and C/D [17:26:22] (PS1) AndyRussG: Show a Limit traffic column to lists of campaigns [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/168997 [17:26:25] that's the idea. we'll just be running the tests at the same time [17:32:55] AndyRussG: your arbitrarily-named bucket proposal has been growing on me... [17:33:20] awight: hmmm [17:33:22] I want to recuperate the A/B bucket thing, though, and unpack the other stuff which has been attaching to it. [17:33:35] What about, buckets are for A/B, but... [17:33:54] there can be a new parameter sent to the banner chooser, "variant" [17:34:25] I'm torn about whether it's specific to a campaign, or the entire fundraising category. [17:35:01] it's also set using a cookie, so has the same behaviors as bucketing, except it's always set explicitly. [17:35:07] bucket is never set explicitly. [17:36:19] awight: that makes more sense from a modelling of A/B tests and campaigns point of view [17:36:25] however... [17:36:29] awight: What's up with the failmail? [17:36:32] That just got me. [17:36:45] AndyRussG: It's pretty much the same as yr idea, just decoupled from buckets [17:36:48] K4-713: looking... [17:36:58] Yeh... that makes sense... [17:37:04] O_o [17:37:05] I suspect that one we get to fully scoping out the requirements we'l find some edge cases that don't quite fit tho [17:37:40] My idea is that we do short, medium and long-term [17:37:52] AndyRussG: well, it already becomes unpleasant if we don't have test isolation... [17:37:55] anyway that's what I'm kinda tending to now [17:38:01] Hmm. [17:38:05] The long term is the whole hog [17:38:21] Fully scoped out requirements and domain model, everything modularized and reusable [17:38:27] that might take a while to grow on me :) [17:38:29] Short term is this December [17:38:54] with stable flexible bucket durations and the rest as is [17:39:12] Medium term is something more flexible than short term but still grafted on the current system mostly as is [17:39:40] K4-713: the audit file came in with blank column headers. [17:39:48] K4-713: I'll ask for a regen. [17:40:01] I think the solutions we're mostly talking about now would go in the medium-term... heheh.. bucket. [17:40:13] noooo! ;) [17:45:03] awight do you have some time today after the standup to chat about worldpay form things? [17:45:54] pizzzacat: sure! [17:46:17] sweet. I mean I'm trying to make multiple versions of forms to test things like button copy/colors [17:46:23] ohwow [17:46:45] so advice on how to set that up to be grabbable [17:46:51] by fr-online [17:47:48] pizzzacat: this is to make prototypes for fr-online to look at, and redirect people to using an URL param? [17:47:54] or you want to do actual A/B testing? [17:49:02] the latter [17:49:05] oh dear [17:49:12] uh oh [17:49:14] * awight squirms a bit [17:49:27] ok let's talk after standup [17:49:30] :) [17:50:07] ejegg: I'm a bit worried about that DI/crm thing going out, without other supporting stuff. You want to look at https://gerrit.wikimedia.org/r/#/c/163764/ ? [17:50:36] I just donno which class will take precedence, hopefully the non-autoloaded one. [17:50:45] Oh sure. Was just trying to figure out the consequenses of not updating contrib tracking in batch mode [17:51:30] right... I think CT is only supposed to be tracking user interaction. The fact that we were getting lines for recurring was an accident. [17:51:40] Ahh, OK [17:52:36] I'm more concerned about, how to cull the old entries :) [17:53:14] so they're just screwing up the source ratios right now [17:53:20] I think so [17:56:18] K4-713: are landing page tests... on donate exclusively, or sometimes on payments? [17:56:29] * awight fires self [17:56:46] and um about subscriptions vs contributions :p [17:56:49] Landing page... tests? [17:56:54] aargh [17:57:05] I thought they used to A/B test donatewiki LPs [17:57:07] Remember, "test" means like 8 things to us. [17:57:09] Now. [17:57:25] Ah. Banner++ tests. [17:57:57] We can test payments landing pages, no problem. [17:58:05] Just hard-code in the ffname. [17:58:21] rad. pizzzacat ^^ K4-713 answered all the fun [17:58:30] but I can still chat [17:58:31] That was fun? Oh good. [17:58:55] there's just no record of ffname in the donation, yet, so we have no test results... [17:59:06] that can change. [17:59:28] K4-713: R36...? [17:59:47] K4-713 so you make different versions of the same form in the same dir [17:59:57] and how do they get specified in the tests? [18:00:09] wait, ffname…. [18:00:19] standup. [18:00:24] rings a bell but been a while [18:00:25] yep [18:00:30] I'll get back to this in a bit... [18:00:37] the-wub should be able to tell you, though. [18:00:47] At least as far as the test specifications. [18:02:07] standup? [18:02:15] hey everyone standup time [18:02:18] ejegg, AndyRussG: standup? [18:04:25] pizzzacat, awight : we can just make different banners which point to different ffname forms. then get the results through the banner name [18:07:03] yep yep! [18:07:26] I had to get a refresher from K4 on how that works the-wub [18:07:45] pretty straightforward though. will sync up with you later, probably tomorrow [18:08:13] cool :) [18:09:45] awight still want to meet or you think I'm good? [18:11:40] pizzzacat: you sound good, in that case [18:11:49] awight: So, we have to do a composer.phar update in crm to pull down DonationInterface? [18:11:55] pizzzacat: Lemme know as you have questions, though [18:12:35] ok awight! thank you [18:19:11] Well I can't sit down and I can't sit down, 'cause my feet are all full of dancing around. [18:22:38] awight: I had to remove the empty DonationInterface directory from vendor/wikimedia for composer to be happy installing it. Do we want to check in a copy of the current dev-drupal branch there? Or can we remove the dir and run composer during deployment? [18:24:05] ejegg: there's corresponding commit in crm/vendor... yeah we have to run update locally, then check in the results. [18:24:22] ejegg: can't run composer during deployment, unfortunately. [18:24:26] probably not ever. [18:26:53] Ah, I seem to be missing a .gitmodules file in vendor [18:28:27] ejegg: hrm. ah [18:28:34] maybe I need to check that in. [18:28:55] And maybe I shouldn't have +2ed that commit without it ... [18:28:56] ejegg: actually, it's not a submodule :-/ [18:30:03] odd, commit 'Manage DonationInterface through Composer' lists it as a subproject commit [18:30:17] +Subproject commit 76b80e416e5b75e5d56a6b1a7f8529e2416266ed [18:30:41] and running 'git submodule' in vendor gives No submodule mapping found in .gitmodules for path 'wikimedia/DonationInterface' [18:32:36] hey AndyRussG https://plus.google.com/hangouts/_/wikimedia.org/bucket-talk?authuser=0&hceid=YWdvbWV6QHdpa2ltZWRpYS5vcmc.2sls23au3ranq5pvhlc041pom0 [18:49:29] (PS1) Ejegg: Fix DonationInterface submodule [wikimedia/fundraising/crm/vendor] - https://gerrit.wikimedia.org/r/169041 [19:10:35] (CR) Ejegg: [C: 2] "Should be OK for orphan adapter 'cause that one already assumes CT data exists." [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/166917 (owner: Awight) [19:11:08] (Merged) jenkins-bot: Don't do contribution tracking in batch mode [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/166917 (owner: Awight) [19:14:29] awight: we need to merge DI master into the drupal branch and point vendor/DonationInterface to that, right? [19:16:07] vendor/wikimedia/DonationInterface, yeah [19:16:20] right [19:16:39] how do you get the nice merge comment listing merged commits? [19:16:56] git cherry -v origin/deployment origin/master, for example [19:17:20] ah, thanks [19:17:26] AndyRussG can you put these 3 bullets your talking about into the google doc? [19:19:47] AndyRussG: high-five on successfully arguing for containing tests to their own campaigns! [19:20:34] (PS1) Ejegg: Merge remote-tracking branch 'origin/master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/169168 [19:22:46] (PS2) Ejegg: Fix DonationInterface submodule [wikimedia/fundraising/crm/vendor] - https://gerrit.wikimedia.org/r/169041 [19:34:17] hey pizzzacat if you could make sure to add an estimate to 2083 and 2084 at some point, that would be super [19:42:45] hey ejegg, could I bother you with a Worldpay sql question? [19:43:03] I think it should be quick, hoping for just a Yes/No :) [19:43:53] sure, go ahead [19:43:59] ok atgo [19:44:10] WP’s portal isn’t very good at telling us who failed a txn - us or them [19:44:23] so I ran this for a specific order id: select * from payments_initial pi LEFT JOIN payments_fraud_breakdown pb ON pb.payments_fraud_id = pi.id where gateway='worldpay' AND order_id = '20077733.1'; [19:44:52] if it says validation action = approved, payments_final_status = failed, that means WP failed it, right? [19:44:53] and not us? [19:45:37] Yep, that should be right [19:45:37] ah sorry! validation_action = process [19:45:39] okay cool [19:45:43] action=process, yeah [19:45:43] thank you! [19:45:48] yw [19:45:50] :) [19:46:12] awight AndyRussG while you're looking at cookies and buckets and the like, could you check this out? https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/2062 [20:22:16] Back in 20 or so... [20:43:14] ejegg: I'm a little concerned about this vendor submodule thing... [20:43:27] Deep in my heart, I know it's the right thing ;) [20:43:42] but Composer doesn't manage the directory as a submodule... [20:44:12] oh, should we let composer manage a copy? [20:44:16] hrm, ok now I think you're right, we just git add the submodule. verry nice [20:44:56] I mean, we'll have to update composer's autoload file when we update the submodule, so that could get messy [20:45:25] well, let's use Composer to manage the version, then git add all the resulting junk [20:45:48] ok, seems reasonable [20:47:05] We'll want to update the recurring test case [20:48:00] ejegg: where is commit a1944f2c21dcc24d05b0846ff8e5c8176aeffda1 ? [20:48:09] in review still [20:48:17] ah kay [20:48:23] https://gerrit.wikimedia.org/r/#/c/169168/ [20:48:27] thanks! [20:49:29] (CR) Awight: [C: 2] Merge remote-tracking branch 'origin/master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/169168 (owner: Ejegg) [20:49:47] (Merged) jenkins-bot: Merge remote-tracking branch 'origin/master' into drupal [extensions/DonationInterface] (drupal) - https://gerrit.wikimedia.org/r/169168 (owner: Ejegg) [20:49:53] ty [20:56:07] (CR) Awight: [C: 2] "Thanks!" [wikimedia/fundraising/crm/vendor] - https://gerrit.wikimedia.org/r/169041 (owner: Ejegg) [20:58:45] (CR) Awight: [V: 2] Fix DonationInterface submodule [wikimedia/fundraising/crm/vendor] - https://gerrit.wikimedia.org/r/169041 (owner: Ejegg) [20:59:52] BLARGH. I just remembered where I left all this last week. [21:02:17] (CR) Ejegg: [C: 2] "Looks like it should work! TODO: update recurring GC test" (2 comments) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/163764 (owner: Awight) [21:10:23] AndyRussG: can u chat buckets tomorrow 12-1 EDT? [21:13:42] awight: u bet! [21:13:56] I was gonna ask the same... [21:14:14] rad. encalendaring [21:14:42] ejegg: ^^ optionaled u [21:15:22] AndyRussG: [21:15:38] AndyRussG: I think we're on a roll with this named buckets/variants thing, just some details to work out. [21:16:10] Yeah! I think it's promising [21:16:31] Good point you made about the cache, I keep forgetting about that silly layer [21:16:38]