[06:28:21] (PS1) Jdlrobson: Hygiene: Officially deprecate global functions [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213986 [06:28:23] (PS1) Jdlrobson: Hygiene: Simplify the module you add to the page [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213987 [06:28:25] (PS1) Jdlrobson: Hygiene: Use this keyword (it's your friend) [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213988 [06:28:27] (PS1) Jdlrobson: Simplify selection of random banner [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213989 [06:28:29] (PS1) Jdlrobson: Save 5.3k whenever a campaign is not running [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213990 [06:29:33] (CR) jenkins-bot: [V: -1] Simplify selection of random banner [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213989 (owner: Jdlrobson) [06:29:42] (CR) jenkins-bot: [V: -1] Save 5.3k whenever a campaign is not running [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213990 (owner: Jdlrobson) [06:30:09] (PS2) Jdlrobson: Hygiene: Officially deprecate global functions [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213986 [08:41:59] Wikimedia-Fundraising, Mobile: Mobile fundraising banners appearing within other apps, moving up the screen - https://phabricator.wikimedia.org/T100421#1314017 (Wittylama) Looking at landing pages of those links: do you mean to say that a you're STILL running same "€2 = a cup of coffee" story that everyon... [14:38:16] heading to boulder, back shortly [15:05:54] Fundraising Sprint Kraftwerk, MediaWiki-extensions-CentralNotice: [bug] Banners adding themselves to some buckets - https://phabricator.wikimedia.org/T99915#1314642 (atgo) [15:09:34] Fundraising Sprint L, Fundraising-Backlog, § Fundraising Sprint Abba, § Fundraising Sprint Beastie Boys, and 3 others: Update GC audit to read WX file - https://phabricator.wikimedia.org/T86090#1314655 (atgo) [15:15:17] Fundraising Sprint Indigo Girls, Fundraising Sprint Kraftwerk, Fundraising Tech Backlog, Astropay Integration, Patch-For-Review: Make form for Astropay Brazilian credit cards - https://phabricator.wikimedia.org/T97056#1314668 (atgo) @ejegg I just confirmed with Megan that there are not addition... [15:16:26] Fundraising Sprint Indigo Girls, Fundraising Tech Backlog, MediaWiki-Templates, MediaWiki-extensions-DonationInterface, and 2 others: Mustache renderer needs to support i18n - https://phabricator.wikimedia.org/T97149#1314670 (atgo) @ejegg this one is blocking T97056 - should we put it int he spri... [15:19:30] Fundraising Tech Backlog, Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM: Make JP Morgan Import a bit more robust - https://phabricator.wikimedia.org/T95487#1314674 (RLewis) @atgo and @awight I had a call with JP Morgan about this and it seems as though they updated their systems after we create... [15:26:07] Fundraising Sprint Kraftwerk, MediaWiki-extensions-CentralNotice, Documentation: CN documentation: review status quo and write a new overview - https://phabricator.wikimedia.org/T99011#1314684 (atgo) Thanks for doing this, @andyrussg. The plan looks good to me, I think we should clean up as we go. Ha... [16:13:50] morning K4-713, back in SF? [16:13:58] cwdent: Yes! [16:14:09] Though, I have no idea how I woke up and went to the office this morning. [16:14:12] But I did. [16:14:19] impressive [16:14:31] we'll see how you feel at 5pm :) [16:16:16] Oh, probably "extremely caffeinated". [16:16:29] I practically jumped out of bed at 6am. [16:16:51] Yesterday was terrible. [16:17:11] A group of us had to catch a taxi out of the venue at 4:30am for our first flight. [16:17:37] ugh [16:17:58] Which would have been harsh enough after the ways we all punished ourselves at the hackathon, but then, the flight out of amsterdam was delayed for five and a half hours or something. [16:18:27] brutal [16:18:37] I was straight-up hallucinating from sleep-deprivation at one point. The kind where stationary things in your peripheral vision are moving. [16:19:04] But, yeah: Four of us were on the same flight. [16:20:01] do you feel good about the hacking part of the hackathon? [16:20:10] i was watching the tickets fly and wishing i had something to contribute [16:20:21] Anyway, I kind of hope the jetlag hits soon, or I will be forced to say insane things in the future like: "The best way to conquer jetlag is to not sleep more than three hours a night on the whole trip". :/ [16:20:38] So, the hacking thing... I feel kind of good about some of it. [16:21:03] Day 1 was basically awight and myself digging ourselves out of a configuration hellhole. [16:21:23] Day 2 started where I thought we already were before Day 1. [16:22:07] Day 3 was basically me thinking "Wow, I need to split the first part of this task up in to like eight more specific subtasks". [16:22:34] So... I feel good about this, in that we will be ready for the majority of the team to be working on the thing at the wikimania hackathon. [16:22:47] Also, I met some cool people I had only ever talked to over IRC. [16:22:52] So, that was nice. [16:23:26] that's cool [16:24:07] yeah, the devil's in the details [16:24:41] happens to me basically every time i open a text editor [16:24:49] spend the whole day on step 0 [16:24:55] seriously. [16:25:22] I guess that's why people who seem to do best at hackathons usually start with something that didn't exist at all before. [16:25:35] that's a good point [16:27:07] browser behaviors i have never desired include making the website huge because i grazed the ctrl key while scrolling [16:28:31] I thought everybody liked that. :p [16:28:51] The operating system one that drives me *insane*, is "paste formatted". [16:28:55] Never. In. My. Life. [16:29:08] Why is that default. [16:29:14] Who enjoys this. [16:29:24] haha totally [16:29:34] now the 2nd half of my email is green italics [16:29:45] * cwdent goes back to mutt [16:38:57] protip: in FF about:config, switch mousewheel.with_control.action to 1 to disable that zoom behavior [16:39:36] i was able to get much more pleasant trackpad action in general by tweaking a few things in there [16:49:52] * K4-713 waves at XenoRyet [16:49:59] Hi there [16:50:06] How was the trip? [16:51:01] Great! And terrible. [16:51:08] But mostly great. [16:51:24] The airport stuff was the only terrible thing, and that's... well, there's only so good that crap can get. :) [16:52:09] The rest of it was nice. [16:52:45] I haven't really gone to Europe for so few days before, either, so that was interesting. [16:53:09] How have you been? RapidHTML still a giant hairball? [16:56:14] (PS1) Cdentinger: make a/b widget return some fake data [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/214084 [16:57:46] * AndyRussG waves at everyone :) [16:57:53] Yea, still detangling that mess. Though I think in the latest blind alley I've chased down I found a bunch of wholly unused javascript. [16:58:05] Valid thing to do to take that out? [16:59:26] Heeeee. [16:59:35] XenoRyet: in CentralNotice, the approach we took for that kind of stuff was to put the removal of cruft in a separate commit/Gerrit change [16:59:55] Then we deployed out the pairing down/clearing out before any new stuff [17:00:03] XenoRyet: Jeff and I were talking at the hackathon, about maybe putting a bounty on removing total cruft. [17:00:09] We both think it's a great idea. [17:00:21] Yea, was planning to do this in it's own commit. [17:00:24] * AndyRussG knows less than nothing about RapidHTML [17:00:46] K4-713: that could lead to corrupt practices... 8p [17:01:04] Well... what doesn't? :D [17:01:14] "I'll write you some cruft and you give me a cut of the cruftbounty" [17:01:21] heh also true [17:01:37] bienvenue BTW [17:01:50] Who wants to write a script that pulls down the last month's commits that got merged to master for a list of extensions, counts up lines removed, and tells you who won? [17:01:51] We could have a whole underground cruft trade network. [17:01:57] bahaha [17:02:23] Possibly we already do. [17:02:28] Maybe later we could accept donations in cruftcoin [17:02:31] Just need to leverage it. :p [17:02:34] haha [17:02:42] cruftcoin [17:02:47] I love it. So much. [17:02:57] Right? Somebody get on that. [17:03:11] we just missed the cryptocurrency bubble [17:03:23] cruftocurrency [17:03:25] Maybe it missed us. [17:03:34] * K4-713 dodges [17:03:40] yeah [17:03:43] This time it won't be a bubble, it'll be a hairball [17:03:50] hehe [17:03:59] You know... I'll be a lot more inclined to take that stuff seriously on Round 2. [17:04:10] Which is true with me and pretty much anything. [17:04:23] (PS6) Ejegg: More logic for mustache forms [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/212728 [17:04:36] Particularly when we're talking about donor money... [17:04:50] (CR) jenkins-bot: [V: -1] More logic for mustache forms [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/212728 (owner: Ejegg) [17:05:15] i have a local branch of mediawiki-vagrant i was hoping to get up for review before the meeting this aft, but i don't think i have write access to the repo. anyone besides awight have any info? [17:05:16] what is money, anyway? [17:05:29] AndyRussG: abstracted gold. [17:05:50] speak of the devil [17:05:54] Or possibly dubloons. [17:06:07] anyone need anything from other teams? [17:06:08] ohai? [17:06:08] cwdent: You should be able to send up a patch to any repository with git review [17:06:11] No, wait. That's recursive. [17:06:17] awight: hi! [17:06:28] ejegg: want to bring up Mustache? [17:06:29] ejegg: hi! thx, not much [17:06:50] AndyRussG: morning! Please poke me if I should be reviewing things [17:06:55] awight: good call [17:07:09] sorry, meeting until 10:30 [17:07:40] (CR) Jdlrobson: [C: -1] "Tests need updating" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213989 (owner: Jdlrobson) [17:07:46] when i try to git review it asks for creds but mine don't work and i get: fatal: Authentication failed for 'https://gerrit.wikimedia.org/r/mediawiki/vagrant/' [17:07:57] (CR) Jdlrobson: [C: -1] "Tests need updating" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213990 (owner: Jdlrobson) [17:08:01] if i try to browse to that repo i get "not found" [17:08:13] which is what github does when you don't have access...so i wondered if it was the same thing [17:11:17] Fundraising Dash, Fundraising Sprint L: make the a/b test widget return data - https://phabricator.wikimedia.org/T99867#1314982 (cwdent) https://gerrit.wikimedia.org/r/#/c/214084/ maybe not an actual use case but this does plug some stuff together so the widget returns data. still waiting for clarificat... [17:12:50] Fundraising Dash, Fundraising Sprint Kraftwerk: make the a/b test widget return data - https://phabricator.wikimedia.org/T99867#1314983 (cwdent) [17:15:31] cwdent: https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/vagrant ? [17:15:43] Very little is not public here, thankfully [17:16:23] cwdent: https://gerrit.wikimedia.org/r/#/admin/projects/ [17:16:59] huh, yeah that the remote for my repo. i wonder why it won't take my creds? [17:17:32] Once you have used the filter on that once, and have the filtered URL in your browser history, just type "filter" (or anything else that works) into the URL bar to get it back, and replace the value of the fulter URL param with something... avoid the wait [17:18:40] cwdent: if it's https then you cloned the repo using https, copying the clone command when you were logged out of Gerrit [17:18:58] You can log in again, then go to the repo's page, to see the ssh remote repo URL [17:20:10] cwdent: XenoRyet: I'm just fixing my settings for automatically becoming a reviewer, do you want me to add you both for CentralNotice? https://www.mediawiki.org/wiki/Git/Reviewers [17:20:25] doh, yep that was it. thanks AndyRussG [17:20:48] cwdent: np :) glad it worked :) [17:20:50] and sure i'd love to be a default reviewer. digging into my reviews now [17:21:00] Yea, sure. Sounds good [17:21:05] K will do :) [17:22:53] (PS7) Ejegg: More logic for mustache forms [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/212728 [17:34:25] Fundraising Sprint Indigo Girls, Fundraising Tech Backlog, MediaWiki-Templates, MediaWiki-extensions-DonationInterface, and 2 others: Mustache renderer needs to support i18n - https://phabricator.wikimedia.org/T97149#1315030 (Ejegg) a:pizzzacat>Ejegg [17:36:28] Fundraising Dash, Fundraising Sprint Kraftwerk, MediaWiki-Vagrant: create vagrant role for fundraising dash - https://phabricator.wikimedia.org/T99869#1315031 (cwdent) cloned through ssh and got this up for review: https://gerrit.wikimedia.org/r/#/c/214088/ i've got a few things to sort out still [17:39:14] hrm i think i shouldn't have actually pushed that last one for review. it's definitely not ready, i just wanted awight|meet to see it before i go any further. [17:39:58] AndyRussG: is there a better procedure for that? [17:41:17] cwdent: honored tradition dicates you put "WIP" at the beginning of the commit message [17:41:41] Fundraising Sprint Indigo Girls, Fundraising Sprint Kraftwerk, Fundraising Tech Backlog, MediaWiki-Templates, and 3 others: Mustache renderer needs to support i18n - https://phabricator.wikimedia.org/T97149#1315040 (Ejegg) [17:41:41] Gerrit actually has a "draft" option but we don't... like it... [17:42:16] ok, going to rebase it anyway so i'll throw WIP on there. thanks! [17:43:16] cwdent: np! [17:44:03] cwdent: probably also awight|meet has told you, you're supposed to separate rebases and real changes into different patch sets (though in the case of just adding "WIP" I think it's OK) [17:45:15] cwdent: awesome, looking now [17:45:24] https://gerrit.wikimedia.org/r/#/c/214088/ [17:45:27] ty! [17:45:53] just need to figure out how to passthru a custom port (8080 is all kinds of taken) and write an init script [17:48:24] cwdent: I've had trouble with that too, see: https://gerrit.wikimedia.org/r/#/c/208861/2/puppet/modules/role/settings/fundraising.yaml,unified [17:48:29] That... didn't quite work [17:48:51] awight: how was meet? In answer to your question about stuff for review, the ball is currently out of my court for both of these: https://gerrit.wikimedia.org/r/#/c/210408/ https://gerrit.wikimedia.org/r/#/c/202310/ [17:49:10] * cwdent isn't going to code any more today, is going to review [17:49:55] Also here's some low-hanging fruit: https://gerrit.wikimedia.org/r/#/c/212834/ [17:50:02] well weds is the day the old sparkfun crew meets up for lunch [17:50:12] AndyRussG: ok, thanks! [17:50:19] cwdent: cool :) [17:50:23] awight: yw :) [17:50:31] (PS2) Awight: Clean cruft in CNBannerChoiceDataResourceLoaderModule::getChoices() [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/212834 (owner: AndyRussG) [17:50:35] (CR) Awight: [C: 2] Clean cruft in CNBannerChoiceDataResourceLoaderModule::getChoices() [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/212834 (owner: AndyRussG) [17:51:25] (Merged) jenkins-bot: Clean cruft in CNBannerChoiceDataResourceLoaderModule::getChoices() [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/212834 (owner: AndyRussG) [17:52:28] i'm going to head over there now, sometimes it can go a little long cause there are a bunch of us and the service is slow, but i'll be back in plenty of time for spring chicken and irc will buzz my phone [17:55:43] AndyRussG: Donno if I should be reviewing https://gerrit.wikimedia.org/r/#/c/202310 ? It's still blocked on the getDependencies thing, and includes that placeholder mixin, so I assume you don't want to merge yet? [17:55:55] aah: branch [17:55:58] I keep doing that. [17:56:05] heh branch branch indeed [17:56:36] It still won't merge unless I force though, cos of the RL incompatibility. No word on the getContext hypothesis? [17:56:51] It won't merge even into the branch due to depending on the getDependencies adaptation, which definitely shouldn't merge yet [17:57:03] k, I'll leave it alone then [17:57:25] awight: no word yet, Krink.le is all post-hackathon I think [17:57:51] awight: I think if you'd like to look at the most recent changes and give your opinion, even in the form of a +1, that'd be great [17:57:56] (or a -1) [17:57:56] k [17:57:58] thx! [17:58:12] AndyRussG: wrt https://gerrit.wikimedia.org/r/#/c/210408/ yes, I meant a CN-specific module to do the key/value stuff. [17:58:16] files are cheap :p [17:58:33] awight: ah hmm well yes, since they're all RL-munged [17:58:53] yep, and the cognitive gain from splitting up is pretty big [17:59:57] Since you mention it, maybe it is wise to ask around to see if there's a similar library in mw-core already. [18:00:15] resources/lib/jquery/jquery.jStorage.js ? [18:01:03] Also jdrobson and mobile have an interest in us reorganizing bannerController code so that less is loaded at the top... https://phabricator.wikimedia.org/T100372 [18:01:16] creepy, mediawiki.js uses the raw localStorage rather than jStorage [18:01:25] AndyRussG: sounds great! [18:01:27] awight: I did poke around, and looked at that one... Yes also true [18:01:47] There must be some localStorage roadmap ideas out there [18:02:31] ooh yeah https://gerrit.wikimedia.org/r/#/c/213990/1/includes/CNBannerChoiceDataResourceLoaderModule.php,unified [18:10:35] Fundraising-Backlog: Stop retrying certain types of GC recurring failures - https://phabricator.wikimedia.org/T100536#1315111 (CCogdill_WMF) NEW [18:13:47] (CR) Awight: [C: 2] "Looks really good!" (3 comments) [extensions/CentralNotice] (campaign_mixins) - https://gerrit.wikimedia.org/r/202310 (https://phabricator.wikimedia.org/T90915) (owner: AndyRussG) [18:14:09] AndyRussG: want to merge this stuff to the branch? [18:14:20] I wouldn't mind not reviewing this patch again :p [18:54:53] awight: hey... wrt the merge, it actually can't merge to the branch, because a dependency is the patch that updates the RL getDependencies method sig [18:55:11] because of merge branch fun fun, we have actually two copies of that patch [18:55:28] The real one is on master, and should be merged along with the core change, when that happens [18:55:58] There is a temporary cherry-picked one on the feature branch that is just there so we can smoke test the patches that need it [18:56:26] However, once the master one merges, we should instead merge master into the feature branch again, and rebase the mixin patches on that instead [18:56:29] I think [18:56:34] Anyway, that was the plan.... [18:57:23] I think the +2 on the current version is good for now, tho! Shouldn't be any more actual review work [18:57:35] (unless we... rewrite the whole thing.... bwahahahahaha) [18:57:42] thx also [18:58:48] hi! [18:59:06] :) [18:59:08] AndyRussG: we can force a merge, it's up to you [18:59:54] u remove Jenkins as a reviewer, then click Submit real quick, and up-up-up-down-down-down-left-right-left-right-left [18:59:56] awight: Is there much point? It'd leave a nonfunctional period in history, no? [19:00:12] well it's already got non-production stuff such as the placeholder [19:00:32] the master history will always be functional, as long as we merge the mixins branch at a functional point. [19:00:33] awight: true but at least it works as promised! [19:01:05] I'm for not merging but moving the card to "awaiting deploy" :) [19:02:26] Even on the feature branch I feel it's a bit off to have a patch like that [19:02:48] Maybe I'm just being persnickity [19:03:18] no problem, it's entirely a matter of preference [19:04:37] cool thx! [19:35:04] awight: ok to deploy the merged hackathon fixes? I'm hoping to put some part of the astropay forms up tomorrow [19:35:33] I know we need to add the new 'enabled' globals [19:35:38] ejegg: oof [19:35:43] yeah let's do it [19:35:46] I can [19:35:57] only if you wanna! [19:36:02] sounds fun :) [19:36:14] ok, i'll ride shotgun [19:37:49] awight: wrt https://gerrit.wikimedia.org/r/#/c/210408/ , I'd like to breaking out the KV stuff into a separate component. The time is either ripe or nearly ripe, I think, for a bigger reorg of all that code, and that's essentially what jdrobson is looking for, too [19:37:59] So I'd like to wait and do that all at once [19:38:09] Thoughts? [19:39:21] It seems really easy to move it into a separate file and add to the bannercontroller scripts array... it doesn't have to be its own module until you're ready [19:39:42] what did you think of jQuery jStorage? [19:40:31] ok, it doesn't save you much [19:40:44] nvm [19:42:26] sigh, I remember marking something on master to not be merged into deployment, but can't find it now [19:44:27] hrm, lemme see [19:44:52] (PS1) Awight: Merge master into deployment [extensions/DonationInterface] (deployment) - https://gerrit.wikimedia.org/r/214126 [19:45:33] ejegg: I just diffed the merge and it's clean, actually [19:45:42] nice [19:49:52] seems to be working locally [19:50:06] hmm, dd-nl doesn't load. form settings change? [19:50:17] ouch [19:52:10] awight: do you know which version of python does DjangoBannerStats run on? Also maybe these days I have the power to see 4 myself? (It's on erbium?) [19:52:31] oh hey, that's been commented out for a while. right. on purpose [19:52:32] AndyRussG: it's on... barium I believe [19:52:56] python -V [19:52:56] Python 2.7.3 [20:05:14] (PS2) Gilles: Save 5.3k whenever a campaign is not running [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213990 (owner: Jdlrobson) [20:06:01] (CR) jenkins-bot: [V: -1] Save 5.3k whenever a campaign is not running [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/213990 (owner: Jdlrobson) [20:42:04] Fundraising Dash, Fundraising Sprint L: make the a/b test widget return data - https://phabricator.wikimedia.org/T99867#1315578 (cwdent) [20:42:08] Fundraising Sprint Indigo Girls, Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, and 2 others: Make form for Astropay Brazilian credit cards - https://phabricator.wikimedia.org/T97056#1315579 (Ejegg) [20:42:27] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, Wikimedia-Fundraising, Patch-For-Review: Upgrade payments mw-core to REL1_25 - https://phabricator.wikimedia.org/T94528#1315580 (awight) [20:42:27] Fundraising Dash, Fundraising Sprint L, MediaWiki-Vagrant: create vagrant role for fundraising dash - https://phabricator.wikimedia.org/T99869#1315581 (cwdent) [20:42:27] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, Astropay Integration, Patch-For-Review: Write confirmation message listener for Astropay, which rescues donations from the limbo queue - https://phabricator.wikimedia.org/T90506#1315582 (Ejegg) [20:42:41] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface: Provide access to limbo messages without knowing keys - https://phabricator.wikimedia.org/T99152#1315583 (awight) [20:42:55] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface, Patch-For-Review: Orphan slayer reads from Memcache rather than ActiveMQ - https://phabricator.wikimedia.org/T99017#1315584 (awight) [20:42:59] Fundraising Dash, Fundraising Sprint L: Create "A/B testing" board in Dash - https://phabricator.wikimedia.org/T98927#1315585 (cwdent) [20:43:02] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, Astropay Integration, Patch-For-Review: Add Astropay forms to DonationInterfaceFormSettings.php - https://phabricator.wikimedia.org/T90505#1315586 (Ejegg) [20:43:08] Fundraising Sprint Indigo Girls, Fundraising Sprint James Brown, Fundraising Sprint Kraftwerk, Fundraising Sprint L, and 2 others: Document how the new queuing stuff will work - https://phabricator.wikimedia.org/T96434#1315587 (awight) [20:43:13] Fundraising Dash, Fundraising Sprint L: "Get Test" in A/B testing header - https://phabricator.wikimedia.org/T89990#1315588 (cwdent) [20:43:26] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface: Confirm orphan slaying works on Memcache, stop writing to ActiveMQ - https://phabricator.wikimedia.org/T99018#1315589 (awight) [20:45:13] Fundraising Sprint House of Pain, Fundraising Sprint Indigo Girls, Fundraising Sprint James Brown, Fundraising Sprint Kraftwerk, and 3 others: Change errors on GC forms from popup to red text - https://phabricator.wikimedia.org/T86214#1315601 (XenoRyet) [20:46:03] Fundraising Sprint Flaming Lips, Fundraising Sprint Grandmaster Flash, Fundraising Sprint House of Pain, Fundraising Sprint Indigo Girls, and 6 others: Implement campaign associated mixins per findings from T90913 - https://phabricator.wikimedia.org/T90915#1315604 (AndyRussG) [20:47:33] Fundraising Sprint James Brown, Fundraising Sprint Kraftwerk, Fundraising Sprint L, MediaWiki-extensions-CentralNotice, Patch-For-Review: CentralNotice: deploy core change required for campaign mixins - https://phabricator.wikimedia.org/T98924#1315608 (AndyRussG) [20:49:01] Fundraising Sprint House of Pain, Fundraising Sprint Indigo Girls, Fundraising Sprint James Brown, Fundraising Sprint Kraftwerk, and 3 others: CentralNotice: key-value storage for new campaign-associated mixins/banner history - https://phabricator.wikimedia.org/T94763#1315615 (AndyRussG) [20:51:08] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Analytics-Cluster, operations: Verify kafkatee use for fundraising logs on erbium - https://phabricator.wikimedia.org/T97676#1315643 (atgo) [20:51:12] Fundraising Sprint Kraftwerk, Analytics-Cluster, operations: Verify kafkatee use for fundraising logs on erbium - https://phabricator.wikimedia.org/T97676#1315644 (AndyRussG) [20:51:54] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Analytics-Cluster, operations: Verify kafkatee use for fundraising logs on erbium - https://phabricator.wikimedia.org/T97676#1249877 (AndyRussG) [20:53:57] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, Fundraising-Backlog: Send announcement to CN-Admins about upcoming technical and UI changes - https://phabricator.wikimedia.org/T98053#1315658 (AndyRussG) [20:59:49] Fundraising Sprint L, Astropay Integration: Validate data in mustache form - https://phabricator.wikimedia.org/T100562#1315677 (Ejegg) NEW a:Ejegg [21:00:59] Fundraising Sprint L, Astropay Integration: Validate data in mustache form - https://phabricator.wikimedia.org/T100562#1315677 (Ejegg) [21:02:16] Fundraising Sprint L, Astropay Integration: Validation and error display for mustache form - https://phabricator.wikimedia.org/T100562#1315697 (Ejegg) [21:03:55] Fundraising Sprint L, Astropay Integration: Validation and error display for mustache form - https://phabricator.wikimedia.org/T100562#1315698 (Ejegg) [21:15:46] Fundraising Sprint Kraftwerk, Fundraising Sprint L, Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface: Revoke ActiveMQ read permissions for payments cluster - https://phabricator.wikimedia.org/T92923#1315724 (awight) [21:25:55] Fundraising Dash, Fundraising Sprint L, Fundraising-Backlog: Search and return results for Campaign name - https://phabricator.wikimedia.org/T89980#1315782 (atgo) [21:36:48] MBeat: shall I get the deployment ready to go? [21:38:58] (PS3) Awight: Rebrand WorldPay -> Worldpay [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/213269 [21:39:11] ejegg: K4-713: ^^ hey can someone kick that? [21:40:42] sure, 1 sec [21:41:00] ejegg: Am I right that Astropay should be disabled in production? [21:43:50] awight: yeah, at least till this form stuff gets merged [21:44:00] k [21:45:54] d'oh, we have i18n strings for test adapter titles? [21:46:08] yuck [21:46:24] friends, i am so wiped. unless you all need me, i think i'm going to crash out [21:46:25] how does that... [21:46:38] atgo: see ya! [21:46:46] ciao! thanks for takng the time early today [21:46:47] ejegg: where you looking? [21:46:53] bonne nuit [21:47:01] ciao ciao [21:47:10] awight: just grepped all of DI for the old casing [21:47:47] some languages have "testingworldpaygateway": "Prueba de WorldPay" [21:47:58] in worldpay_gateway/i18n [21:48:21] ejegg: I don't see it get used, we should delete and see if it breaks the tests [21:48:31] maybe in another commit... [21:49:44] yah, another commit [21:50:01] (CR) Ejegg: [C: 2] Rebrand WorldPay -> Worldpay [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/213269 (owner: Awight) [21:50:39] (Merged) jenkins-bot: Rebrand WorldPay -> Worldpay [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/213269 (owner: Awight) [21:51:47] ah man - deleting it /does/ break tests [21:53:01] yeesh [21:54:05] deployment merge looks good locally. were you trying to fold that Worldpay rebranding in? [21:54:13] yeah I think so [21:54:57] Fundraising-Backlog: Stop retrying certain types of GC recurring failures - https://phabricator.wikimedia.org/T100536#1315857 (atgo) I wonder if this has any relation to this: https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/1887 Perhaps we can extend the functionality there to inc... [21:57:39] (PS1) Awight: Merge deployment into master [extensions/DonationInterface] (deployment) - https://gerrit.wikimedia.org/r/214258 [22:02:25] (CR) Ejegg: [C: 2] "not exploding things over here!" [extensions/DonationInterface] (deployment) - https://gerrit.wikimedia.org/r/214258 (owner: Awight) [22:02:40] oh thanks! [22:05:33] (Abandoned) Awight: Merge master into deployment [extensions/DonationInterface] (deployment) - https://gerrit.wikimedia.org/r/214126 (owner: Awight) [22:08:12] who is pgehres? [22:08:24] * cwdent checked the staff page [22:08:45] cwdent: nice guy we used to work with [22:09:16] He was planning to build out something like the analytics dashboard, which is why the silly banner impressions stuff is so baroque and based on Django [22:09:48] ah ha! but it looks like current data? [22:11:37] or maybe a subset of dev_analytics [22:12:18] hmmm, so is this impression data coming from centralnotice? [22:14:02] cwdent: soort of. CN causes a web request back to Special:RecordImpression, currently, and then the pageview is intercepted [22:14:36] the baroque thing munges those web request logs into the "pgehres" db [22:14:48] as aggregate, so mind the "count" column. [22:15:30] gotcha. looks like a bunch of the same tables are in dev_analytics [22:15:42] that sounds stale [22:15:53] err yeah. the data is old [22:16:40] do you think all the data needed to derive banner a/b results is available on lutetium? [22:18:25] awight: K4-713: ejegg: should I request access to barium? If I did that could I see directly how DjangoBannerStats runs instead of having to ask other people for details? (I'm just trying to re-set it up to run locally on this computer...) [22:18:49] AndyRussG: IMO yeah [22:19:17] AndyRussG: Totally. [22:19:22] awight: K4-713: cool thanks! [22:19:39] barium is indeed the one, right? [22:19:50] AndyRussG: here's the script that runs every 15 min: [22:19:50] export PYTHONPATH=/etc/fundraising [22:19:50] python /srv/DjangoBannerStats/manage.py LoadLPImpressions --verbose --recent [22:19:53] python /srv/DjangoBannerStats/manage.py LoadBannerImpressions2Aggregate --verbose --top --recent [22:22:09] https://www.youtube.com/watch?v=lGVoZuMbTI4 [22:22:11] AndyRussG: oh. and some config magick. [22:22:15] dammit [22:22:20] Janet [22:23:12] Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface: In the event that credentials for a gateway are not supplied, that gateway should go in to local development mode. - https://phabricator.wikimedia.org/T99954#1315932 (Qgil) Did someone work on this project during #Wikimedia-Hackathon-2015?... [22:23:34] AndyRussG: emailed. [22:25:14] cwdent: sorry, I'll try to review the vagrant patch right after this deployment, assuming success... [22:25:51] Fundraising Tech Backlog, MediaWiki-extensions-DonationInterface: In the event that credentials for a gateway are not supplied, that gateway should go in to local development mode. - https://phabricator.wikimedia.org/T99954#1315975 (K4-713) I did, yes. But I was not able to get the patch I was working on... [22:29:16] K4-713: confirming, we should leave all gateways but GC disabled on p4? [22:29:26] Well... [22:29:36] Here's the thing. [22:29:44] I want that to go back to being a staging machnie. [22:29:46] *machine [22:30:04] Jeff and I were grumbling about this in the same direction over the weekend. [22:30:08] Shouldn't we not do orphan stuff on a staging machine? [22:30:12] Eh. [22:30:20] I don't really care either way. [22:30:24] k [22:30:35] i have opinions on that [22:30:39] awight: thanks much! [22:30:47] AndyRussG: any time! [22:30:50] there should be a place to stage orphan code too [22:30:55] ^ [22:31:11] How would we do that? [22:31:14] Yeah I think in fact that's the second time I've had to ask you for such cruft :P sorry for the repetitiveness, I'll save this somewhere good now [22:31:16] especially since we need orphan processor code to be in step with paymentswiki code [22:31:30] that part of deployment is incredibly annoying [22:31:39] awight: You remember how much of a PITA it was to generate orphans? [22:31:43] K4-713: whole parallel infrastucture :) [22:31:45] yes! [22:31:59] The question still stands. [22:32:09] Populated with what orphans? [22:32:30] Hook staging paymentswiki up to processor sandboxes [22:32:39] Staging could allow some params to cause different failures and orphanage? [22:32:39] So, here's my thoughts on this. [22:32:52] the "orphan maker" [22:32:53] hahahahahaha [22:33:00] If we had enough people working on things such that this was our biggest outstanding issue, I'd be all for it. [22:33:01] But. [22:33:19] ...there is a looooooooooot of things we have to knock over before this is the biggest fish to fry. [22:33:40] Also, have we looked to see how many orphans we've picked up lately? [22:33:46] nope [22:33:58] (PS1) Ejegg: Quit wasting translators' time with test pages [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/214263 [22:34:00] Yeah, so, unless it's rather a lot, we should probably spend the time on something else. [22:35:09] Right now, I'm just trying to get the settings files in sync between payments4 and the other boxes. [22:35:31] But I think ejegg's right that when we eventually set up a staging box, it should be full-featured [22:35:32] aaaagh [22:35:56] Don't worry, I'm making a patchfile... [22:35:57] awight: So, because we are where we are right now, there are some settings that are intentionally not the same. [22:36:02] I realize that [22:36:03] i mean, i agree there's a ton of more important stuff [22:36:05] okay [22:36:12] You can review the patch if you have time [22:36:16] but it's no big deal [22:36:19] awight: How are you patching? [22:36:22] "patch" [22:36:29] ah [22:36:50] anyway, it's just that my config stuff makes a bunch of things explicitly disabled, when previously we might have been relying on implicit enabledness. [22:37:08] We shouldn't have been. [22:37:13] hmmm [22:37:29] well, Paypal and Amazon were enabled explicitly, which crept me out, that's why I wanted to get yr opinion [22:37:30] I at least had been operating under the assumption that "fail closed" was the default. [22:37:34] nah [22:37:49] What do you mean "nah"? [22:37:58] Is that "I went in and changed it all"? [22:38:06] urrrgh [22:38:13] And anyway, we whitelist special pages. [22:38:16] no, it was all enabled by default, until my patch [22:38:28] So, that at least had to be done manually. [22:38:40] * awight changes from deployment music to polite conversation music :) [22:39:11] ahhh, you know what, that explains why this happened--people must have enabled gateways to test on staging. [22:39:18] probably me :) [22:39:32] That explains things? [22:39:34] Oh good. [22:40:04] K4-713: check out what I'm talking about, though, https://gerrit.wikimedia.org/r/#/c/213003/11/DonationInterface.php [22:40:07] line 729 [22:40:17] all the gateways default to enabled [22:40:23] or--*did* default to enabled [22:40:34] Yeah, see how the comment says "define as fail closed"? [22:40:39] That makes me think somebody changed it. [22:40:46] that's creepy [22:41:05] I'm about to go have a blame festival. [22:42:00] omg it was me [22:42:06] Oh shit! I just realized what stupid thing I did at the hackathon. [22:42:11] I think. [22:42:35] commit a173f8fd4756f3ec9366c2a81d9cb4c9de3ed5cc [22:42:53] And I did it all with a "few more tweaks in an attempt to get Jenkins to run our tests" [22:42:56] fired. [22:43:16] aha! [22:43:26] ...yeah. What's the date on that? [22:43:33] Within the last... year? [22:43:34] mmm old enough to know better [22:43:36] sort of [22:43:41] recently fired :p [22:43:42] oh hey, we have selenium test [22:43:44] s [22:43:51] whoa [22:44:08] orly? [22:44:10] oh man, in that weird html format [22:44:44] yeah, in tests/macros, formatted as html tables with the ID to find, the value to put in, and the thing to expect [22:50:24] campaign name "your mother's arse" [22:51:23] oh man - guess they're in the "RecordImpression should die in a fire" camp [22:51:40] lol... cwdent: Did you just find one of those? [22:51:42] Because... [22:51:57] * K4-713 looks 100% innocent [22:52:31] I mean, "yourmom" wasn't something they were actually going to use. [22:52:48] hehe, i assume pcoombe [22:52:53] for the british spelling [22:53:09] https://wikimediafoundation.org/wiki/Staff_and_contractors [22:53:13] hmm [22:53:53] oh. the whole everything [22:54:32] * cwdent prays he didn't somehow cause this [22:57:16] cwdent: You knew about this, right? http://frdata.wikimedia.org/ [22:57:55] nope! [22:58:09] The campaign vs amount spreadsheet shows basically everything. [22:58:19] That last row is probably entirely my fault. [22:58:45] also love how ours is the only working server :) [22:59:08] huh [22:59:09] oh nm. stuff's back [23:01:25] everything was broke for a minute. no localization cache for the mw version apparently [23:02:25] where is this? [23:02:33] dev box? [23:02:42] no like...wikipedia [23:02:46] O_o [23:02:50] was saying [033fc78d] 2015-05-27 22:57:17: Fatal exception of type MWException [23:03:00] somebody wins a shirt! [23:03:05] ooo [23:03:30] That's the best exception [23:03:37] There should be an official "you got in trouble" dange. [23:03:39] *dance [23:03:46] of type "stupid subclass of Exception" [23:04:05] everyone gets the t-shirt, eventually. [23:04:12] No such thing as a stupid subclass [23:04:19] baahaaha [23:04:35] Even if my parents are stupid? :p [23:04:47] I got a mini but portable t-shirt 'cause I only broke mobile [23:05:10] I didn't actually get my t-shirt yet, I'm a little put off. Do I need to break this shit again but louder? [23:05:18] OO evolution? [23:05:25] hehe one can hope [23:06:49] K4-713: so this spreadsheet...is this generated? [23:06:56] Yep. [23:07:00] Every 15 minutes or something. [23:07:12] We may have dialed that down. [23:07:24] nice. can i find the code that does it? seems like a good place to start gathering queries [23:07:37] cwdent: python /srv/tools/FundraiserStatisticsGen/fundstatgen.py -c /etc/fundraising/fundstatgen.cfg /tmp/FundraiserStatistics [23:08:05] thanks! [23:08:35] fwiw, it runs on barium, and that's the wikimedia/fundraising/tools repo [23:08:45] emailing the .cfg now :) [23:08:57] aah, if only onboarding were actually this easy ;) [23:09:04] (PS2) Ejegg: WIP Astropay Listener [wikimedia/fundraising/SmashPig] - https://gerrit.wikimedia.org/r/207526 (https://phabricator.wikimedia.org/T90506) [23:09:06] heheh [23:09:09] hey speaking of cron [23:09:33] i've gotten about 10k emails from cron so far. i was fixing to just comment every print statement out of a couple jobs [23:09:42] cron considers any output to be email worthy [23:09:43] no please [23:09:57] Something should be done, but put your ideas in a card [23:10:09] so we can all talk about it [23:10:20] of course [23:10:42] IMO yeah, we should be stderr'ing errors to cron, and piping regular stdout to logs [23:10:44] i was also thinking there must be a wedge you can put between cron and email to categorize [23:11:14] afaik it doesn't differentiate, but we could make something that does [23:11:18] sadly, I think we've all dialed up gmail filters until we stopped caring :) [23:11:19] and pipe all it's output [23:12:13] the highest priority email thing is, to never send donor details. I think we accidentally send details in some error cases. [23:14:07] yeah so something that i was considering wrt the dash is making views on lutetium to get at the data so that the queries the dash does don't return any sensitive data [23:14:12] and then test data is no big deal [23:15:33] cwdent: we have some interesting issues with that, actually. like, we don't ever want to publicize exact timestamp, country, and amount [23:15:48] in tandem? [23:15:50] I'm not sure we've documented this well [23:15:52] yeah [23:15:57] that makes sense [23:16:34] MPaulson told us to chill out a while ago, I'm not sure we fully digested that memo [23:17:13] like, people with access to bank records don't need to correlate with our frdata spreadsheets to figure it out [23:17:21] Yeah, the only reason we can have a bunch of stuff in public is that it's all aggregated. [23:17:26] And, there is no country data in there. [23:17:40] Long story short, we need to have any data reviewed before it's published [23:17:45] yep [23:17:56] Very, very reviewed. [23:18:02] countries are... we should talk elsewhere [23:18:07] hmm yeah, that does sound difficult then. regarding dash. [23:18:15] yeah for sure [23:18:17] those 3 data points are important [23:18:21] in tandem [23:18:31] right [23:18:43] making the dash public is a big deal [23:18:55] Yeah. Basically, we can't. [23:19:01] https://github.com/wikimedia/wikimedia-fundraising-tools/blob/master/FundraiserStatisticsGen/fundstatgen.cfg [23:19:04] * cwdent wipes brow [23:19:08] well, without serious battles [23:19:20] hehe [23:19:23] Aw. Matt. :) [23:19:41] I wonder if he made it to the moon yet. [23:19:53] * K4-713 looks at the sky [23:20:08] (PS1) Awight: update DonationInterface submodule [core] (fundraising/REL1_23) - https://gerrit.wikimedia.org/r/214267 [23:20:11] awight: Also: I'm hella dum. [23:20:19] (CR) Awight: [C: 2 V: 2] update DonationInterface submodule [core] (fundraising/REL1_23) - https://gerrit.wikimedia.org/r/214267 (owner: Awight) [23:20:19] The thing I thought wasn't working, totally is. [23:20:50] good to hear! [23:21:04] As for me, I'm stumped just trying to respond to the deployment scripts with 4-character codes [23:21:12] I just need to put a URL in the form thingy so I can talk to myself. [23:21:17] Officially. [23:21:28] sounds very useful :p [23:21:43] Then we can all officially talk to ourselves all day long. [23:21:53] That's what carpentry was like [23:21:56] The badass thing is that I only had to override something in one place. [23:21:57] I dig it [23:22:07] always a good feeling! [23:22:13] I was like "YES I MADE A GOOD DECISION ONCE" [23:22:17] $dumb = false [23:22:21] aha! [23:22:44] ok... I'm setting the payments settings on fire. [23:22:51] heh [23:22:53] good thing we don't have source control there. [23:22:59] I was about to ask you about something that doesn't really matter. [23:23:07] But: Carry on. [23:23:09] k [23:24:03] !log updated payments from 858b87319daa3d66f62eb32e08cefc6b061748d1 to aa66797553fbcfb63f7cf29abccc44d060b65db0 [23:24:08] crap [23:24:11] Logged the message, Master [23:24:25] ? [23:24:56] it should be a syntax error if the last line in an array doesn't end with a comma. [23:25:09] oof [23:25:11] it's all on fire [23:25:23] * K4-713 raises sysbrow [23:25:28] heh [23:25:40] That is also a nice typo [23:26:49] !log payments rolled back to 858b87319daa3d66f62eb32e08cefc6b061748d1 [23:26:53] What's with all the Uncommitted DB writes? [23:26:57] Logged the message, Master [23:27:08] * awight wakes up with a massive lump on head. [23:27:11] waaat happened [23:27:25] uncommitted writes are a stupid but normal side-effect of rsyncing [23:27:35] donno why. probably kills any active php threads [23:27:37] We used to get one per server. [23:27:42] There are a lot more than that in there. [23:28:19] p3 did it three times. [23:28:20] :/ [23:29:27] looks like the 1st was in the initial rsync, then 2x on roll back [23:29:50] well everyone is going home so i'm getting kicked out of this office [23:30:03] but i'll be on again later and ramble about dash some more [23:30:03] aw, okay. [23:30:16] you'll miss my firing [23:30:23] cwdent: Isn't it after 5 over there? [23:30:44] yeah [23:32:17] i just have a bunch o questions [23:32:24] groovy. [23:32:26] that i don't know if there are really answers for right now [23:32:38] I can probably make something up. :p [23:32:52] i haven't heard anything back from ellery but i feel like i'm not going to be very productive until i get a data source for the tests [23:32:59] or i'll just spin off in the wrong direction [23:34:44] anyway i'll see if i can effectively communicate this through phab [23:34:55] gotta roll! later [23:44:19] Going in for another deployment, it was a dumbass typo [23:44:29] s/Enable/Enabled/ [23:44:52] gah! [23:45:07] * awight glares coldly at former self [23:45:28] No idea why I decided to prioritize grammar over, getting it to work right [23:45:45] at least I proved that we indeed fail closed now :) [23:46:18] !log updated payments from 858b87319daa3d66f62eb32e08cefc6b061748d1 to aa66797553fbcfb63f7cf29abccc44d060b65db0 [23:48:46] d'oh. i did exactly the same locally [23:50:25] I did too [23:57:41] K4-713: did you still want to ask why I did something stupid? :) [23:57:47] er... [23:57:49] Oh. [23:58:07] I was just wondering why you went with BadTitleError everywhere that isn't enabled. [23:58:16] ahh. Yeah I think that was a mistake [23:58:28] It says something about that wiki page not existing. With is technically correct, but confusing as hell. [23:58:52] I wanted an error that was already understood by MediaWiki, and therefore translated, but that doesn't quite fit. [23:59:03] heh [23:59:12] Okay, that makes sense. [23:59:39] I can add a new error message and we can use it later, once it's translated...