[00:14:24] see you all tomorrow! [00:58:31] (PS1) Eileen: Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399545 [00:59:16] (CR) Eileen: [C: 2] Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399545 (owner: Eileen) [01:02:37] (Merged) jenkins-bot: Update CiviCRM submodule [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399545 (owner: Eileen) [01:05:17] (PS1) Eileen: Merge branch 'master' of https://gerrit.wikimedia.org/r/wikimedia/fundraising/crm into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/399547 [01:08:07] (CR) Eileen: [C: 2] Merge branch 'master' of https://gerrit.wikimedia.org/r/wikimedia/fundraising/crm into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/399547 (owner: Eileen) [01:08:54] (Merged) jenkins-bot: Merge branch 'master' of https://gerrit.wikimedia.org/r/wikimedia/fundraising/crm into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/399547 (owner: Eileen) [01:16:54] !log update CiviCRM civicrm revision changed from 9d6ba74f57 to ffa9d7fc7a, config revision is da6fa4cce9 [01:17:06] Logged the message at https://wikitech.wikimedia.org/wiki/Server_Admin_Log [01:17:39] cwd are you about - assume it's late for you now [01:18:01] eileen: i am here, what's up? [01:18:32] not something super important but I thought it might be worth deleting some files from /tmp before holiday kicks in [01:18:35] just in case [01:19:30] oh, good idea [01:19:44] there is one file in use [01:19:45] Raw Recipient Data Export Dec 20 2017 16-00-05 PM 1842.csv [01:19:53] cool [01:19:59] and i assume one of those per day? [01:20:03] & others with same filename before the . [01:22:23] eileen: i can delete everything but that one though? [01:23:47] yep [01:23:52] I think so [01:24:00] I know about the silverpop ones [01:24:03] not the ingenico ones [01:25:31] cwd the file date on that one is Dec 20 16:17 [01:25:44] so maybe just use date rather than file - not big files after that date [01:26:22] but this one is big (before it) Dec 20 12:01 Raw Recipient Data Export Dec 20 2017 12-00-03 PM 156.csv [01:27:53] sure are a lot of files [01:28:01] what if i just delete everything before december? [01:30:08] /bin/mv: Argument list too long [01:30:14] :| [01:33:28] yeah - or even Dec 15 [01:33:57] (don't ask me for command help - my bash skills are painstaking :-) [01:34:58] hehe no problem bash is a bad habit of mine :) [01:35:12] lol [01:35:18] for f in Raw\ Recipient\ Data\ Export\ Oct\ * ; do sudo mv "$f" ~/todelete/ ; done [01:36:04] yeah that [01:36:30] bash is so fucking terrible [01:36:52] :-) [01:36:58] the name is fitting [01:37:23] the thing that annoys me most about the merciless gutting of linux by redhat is that they are not even bothering with bash [01:37:56] btw I just did a minor civi deploy - queues look fine but I'm going to be offline in 90 mins & going to the beach so just flagging it could be rolled back if any issue [01:38:13] awesome [01:38:15] jealous [01:38:31] well it's a beautiful place & it's beach weather [01:38:46] hehe [01:38:49] but it's not a perfect situation [01:38:49] it's dark and cold here [01:39:13] the karate instructor is running a 'holiday programme' & today's activity is going to the beach & camping [01:39:48] but, we don't actually trust him to supervise them at the beach properly so we are tag-teaming to be there ourselves - despite paying him to look after them :-) [01:40:07] (kids are good swimmers but beaches have rips sometimes) [01:40:09] so it will be you and a bunch of tiny samurais? [01:40:15] what could go wrong [01:40:24] well there are not many kids gonna be there [01:40:41] our kids love his holiday programme because it is everyone a holiday programme should not be [01:41:10] poorly supervised, highly likely to include r-rated cartoons & bulk-ice-cream friendly [01:41:39] sounds perfect :) [01:41:45] :-) [01:42:52] needless to say we don't really recommend the programme to friends. THey might not be so tolerant of their kids coming home telling obscene jokes as we are [01:44:23] He likes this beach for the reasons it makes me nervous - good surf http://www.doc.govt.nz/parks-and-recreation/places-to-go/northland/places/waipu-area/things-to-do/uretiti-beach-campsite/ [01:45:18] so beautiful [01:45:26] i really hope we do an fr-tech thing in NZ sometime [01:45:38] I totally support that idea! [01:45:39] meee too! [01:45:54] although WMF has funny ideas about where meet ups should be [01:46:19] *cough* also the office [01:47:47] * cwd shows self out [01:47:57] :-) [01:48:49] anyhoo i moved all those files out of the way, and will delete them tomorrow as long as nothing starts on fire [01:49:13] cool [02:12:48] (PS4) Eileen: [wip]Allow Benevity currencies [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399252 (https://phabricator.wikimedia.org/T182808) [02:55:53] (PS5) Eileen: Allow Benevity currencies [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399252 (https://phabricator.wikimedia.org/T182808) [02:57:45] Fundraising Sprint Winter Wanderland, Fundraising-Backlog, Wikimedia-Fundraising-CiviCRM, Patch-For-Review: Different currencies for Benevity import - https://phabricator.wikimedia.org/T182808#3853218 (Eileenmcnaughton) There is now a version of this code on staging available for testing [10:49:40] Fundraising-Backlog: Error while ordering a payment slip from Switzerland to do a donation - https://phabricator.wikimedia.org/T183455#3853678 (Gehel) [11:56:52] Fundraising-Backlog: Document severity of the various forms of outage - https://phabricator.wikimedia.org/T183293#3849435 (Pcoombe) There's also https://collab.wikimedia.org/wiki/Fundraising/Engineering/Outages, not sure how up to date it is. [12:04:48] (PS2) Rafidaslam: Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) [12:16:21] Fundraising-Backlog: Error while ordering a payment slip from Switzerland to do a donation - https://phabricator.wikimedia.org/T183455#3853851 (Pcoombe) @Ilario, could you or someone else from Wikimedia CH take a look at this? [12:47:20] (CR) D3r1ck01: "recheck" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [12:48:42] (CR) jerkins-bot: [V: -1] Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [12:48:48] (CR) D3r1ck01: "composer.lock is not supposed to be in the PS, can you remove it?" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [12:53:10] (PS3) Florianschmidtwelzow: Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [12:53:25] (PS4) Florianschmidtwelzow: Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [12:54:26] (CR) jerkins-bot: [V: -1] Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [13:04:42] (PS5) Florianschmidtwelzow: Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [13:06:06] (CR) jerkins-bot: [V: -1] Add minus-x to "composer test" [extensions/FundraisingEmailUnsubscribe] - https://gerrit.wikimedia.org/r/399480 (https://phabricator.wikimedia.org/T175794) (owner: Rafidaslam) [13:28:47] good morning fr-tech! [13:30:19] hi mepps :) [13:39:39] how's it going jgleeson? [13:43:47] pretty good so far, just listening to the lord of the rings symphony working through a small patch to resolve a glitch in the `composer test` script on the civi project, before moving back to a bug fix for the prometheus issue discussed yesterday [13:44:50] I love this music https://youtu.be/ybZINoDv8I8?t=419 [13:45:19] just fills me with immense happiness :) [13:46:29] how about you? [14:10:18] nice! yesterday i deployed my dash patches so kind of poking around now [14:10:51] might go back to trying to get vagrant up though that felt like a black hole last time (because virtual machine just wasn't running on my computer, not your patch) [14:46:48] hi all! [14:47:07] hey AndyRussG :) [14:47:13] :) [15:06:28] hi jgleeson and mepps! [15:06:47] and AndyRussG ! [15:12:04] ejegg: hey :) [15:19:33] hey ejegg :) [15:20:41] ejegg, what's the best way to pull in a package that has been forked but is yet to be merged back into the original package? packagist won't let me add it to be pulled due to this [15:21:34] I've add a patch for jsonlist that we're using in the civi project "composer test" script block to support multiple file arguments here https://github.com/jackgleeson/jsonlint/commits/master [15:22:15] it's part of a patch to restore the parallel-lint command also [15:41:41] jgleeson: you can list your github URL in the 'repositories' section of composer.json [15:42:11] we're still doing that with the amazon payments sdk in smashpig and donationinterface [15:44:50] thanks! [15:47:32] cool change! [16:11:44] (CR) Mepps: [C: 2] Stop inserting extra 'A/B testing' boards [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/327792 (owner: Ejegg) [16:11:49] (PS2) Mepps: Stop inserting extra 'A/B testing' boards [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/327792 (owner: Ejegg) [16:15:46] hey ejegg, let me know if there's anything you need help with today, otherwise i'll go back to working on upstreaming the activities change to civi and trying to get vagrant working [16:22:16] https://i.redd.it/8uvcmqtpv8501.png [16:25:41] mepps upstreaming sounds good! [16:25:54] I'm just exploring various Amazon donation quirks [16:25:56] sweet ejegg [16:26:18] AndyRussG: lol [16:27:28] lol AndyRussG :) [16:38:07] AndyRussG but that feature just haaaad to go out [16:45:22] Fundraising-Backlog, FR-Amazon, MediaWiki-extensions-DonationInterface: Amazon errors showing up as [Object object] - https://phabricator.wikimedia.org/T183477#3854479 (Ejegg) [16:46:40] Fundraising-Backlog, FR-Amazon, MediaWiki-extensions-DonationInterface: Amazon errors showing up as [Object object] - https://phabricator.wikimedia.org/T183477#3854495 (Ejegg) FR-tech: to reproduce this error locally, try uncommenting the 'seller authorization note' on line 318 of the amazon adapter. [16:47:28] (PS1) Ejegg: Properly serialize Amazon error messages [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/399659 (https://phabricator.wikimedia.org/T183477) [18:09:20] fr-tech are civicrm packages managed via a package manager or manual updates? /org.wikimedia.civicrm/civicrm/packages [18:20:05] jgleeson: those are manually updated [18:20:40] thanks [18:21:42] fr-tech anyone want to get set up for amazon pay testing? [18:24:34] ejegg, just so I'm clear what I'm updating.. can you confirm that the crm repository under fundraising is the drupal project? Also the civi submodule within it, is that our own flavour of the civi drupal module? [18:25:05] (Abandoned) Jgleeson: Merging master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/398310 (owner: Jgleeson) [18:27:21] (PS1) Jgleeson: Removed use of deprecated "= &new" being reported by phplint on crm project. [wikimedia/fundraising/crm/civicrm] - https://gerrit.wikimedia.org/r/399669 [18:32:28] jgleeson: so, the 'crm' repo contains a submodule for drupal and one for civicrm [18:32:43] the root of the website is crm/drupal [18:33:26] the civicrm submodule is mostly stock upstream civi at this point, but still does contain a few local modifications [18:34:04] it's the civi code as distributed in a zip file, rather than the civi code as checked out directly from github [18:34:06] do you have a minute to jump on hangouts? [18:34:11] sure [18:34:17] great see you in the usual room [18:36:48] (PS1) Jgleeson: Restored `composer test` script [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/399670 [18:43:56] lost you ejegg [19:02:02] XenoRyet: are you busy? want to try getting https set up locally? [19:02:35] signing off fr-tech, have a good one! [19:02:44] I think I need to redo my local https so it's supported by chrome [19:06:09] ejegg: Sure. What's involved? [19:06:36] https fer what? [19:06:54] * cwd minds others' business [19:07:11] For testing Amazon locally [19:07:15] cwd even for dev, you need https to run the amazon payments widget [19:07:45] does it require a trusted cert or can you self sign? [19:08:19] cwd you can self-sign, but I think new chrome doesn't trust the cert i generated a while back [19:08:25] some kinda new requirements [19:08:54] XenoRyet: are you running paymentswiki via vagrant, or on the real box? [19:09:01] ssl is hell [19:09:06] Vagrant [19:09:16] vagrant is also hell [19:09:19] * cwd salutes [19:09:38] hmm, ok, lemme try starting that up here. this apt might just have a good connection [19:10:03] ok, so there IS an 'https' role [19:10:16] Yes, but it doesn't provision correctly for me. [19:10:30] what's the error? [19:10:43] anyway, it probably doesn't apply https to the payments wiki [19:10:53] Error: Puppet::Parser::AST::Resource failed with error ArgumentError: Invalid resource type nginx::site at /vagrant/puppet/modules/role/manifests/https.pp:16 on node vagrant.mediawiki-vagrant.dev [19:11:21] oh hey, yep, that's the same thing I get. [19:11:46] guessing you've already updated all the git? [19:12:30] It's a fresh box as of like three or four days ago. [19:12:40] well... [19:13:23] i'd start with the nginx module [19:13:30] maybe there's a syntax error? [19:13:42] ok, I don't see an nginx role [19:14:01] would that be some kind of vagrant extension? [19:14:30] from the looks of it i think it would be searching for modules/nginx/manifests/site.pp [19:14:51] it could be different depending on how it's defined [19:14:56] but probably in there somewhere [19:15:31] this puppet tree and the prod frack one share a common ancestor but i think they are pretty divergent by now [19:15:37] huh, puppet/modules/nginx is empty [19:15:57] ohhh, it's a submodule now! [19:16:05] ah ha [19:16:06] XenoRyet: git submodule update --init! [19:17:23] and now it's doing things rather than just crashing [19:17:30] \o/ [19:17:35] Yea, looks like same here [19:18:12] looks like clean provision [19:18:40] heh, guess this apt's connection is still no match for california [19:18:48] Heh [19:19:07] I'm a little out in the boonies today too [19:20:38] ah, drat, I got some errors [19:21:00] lemme see, I should at least pull down jgleeson's patch and test that [19:21:29] That might be the secret sauce. It worked for me last time. [19:21:34] XenoRyet: want to start looking at what it'll take to extend that https to the payments site? [19:21:51] Sure [19:22:43] Point me in the right direction to get started? [19:22:55] heh, I guess looking at that https role? [19:23:12] Good a place as any, I'll start taking a look. [19:23:20] you can check out frack puppet too [19:23:37] so maybe the payments server should depend on the https role [19:23:51] XenoRyet: I'm pretty sure the real sites are still running on apache [19:23:59] actually i think vw vagrant has a different setup [19:24:02] are you using php-fpm? [19:24:20] oh? [19:24:31] it's not just using nginx as an ssl termination? [19:24:39] we do on prod, yeah [19:24:41] and mod-php [19:24:46] though i think this can all change [19:24:52] 'cause it sounded like XenoRyet was running non-ssl just fine without the nginx submodule [19:25:00] to just nginx/fpm [19:25:34] although nginx is open core bullshit [19:25:46] open... core? [19:26:09] is that some obscure music genre? [19:26:16] hahaha [19:26:19] nginx could totally be a band name [19:26:35] it absolutely sounds like a metal band of some sort [19:26:55] but nginx is also an obnoxious corporation [19:27:04] ahh, got it [19:27:13] you have to pay for "premium" nginx features [19:27:14] open source at the core, but tons of proprietary extras [19:27:18] yep [19:28:41] damn, I was going to say at least that means a bunch of money will go to paying devs [19:28:50] but i do think fpm is a far superior way to run php [19:28:51] but that's only until oracle decides to buy it [19:28:57] heh yeah [19:29:59] i have no problem with making money of floss [19:30:05] it's what wmf does [19:30:19] but there's also like opensuse where they make money off support contracts [19:30:26] but they open source everything still [19:32:03] but taking open source and closing it up is whack [19:32:23] i'm looking at you apple [19:32:25] yeah, that's some nastiness [19:40:12] Fundraising Sprint Winter Wanderland, Fundraising-Backlog, MediaWiki-Vagrant: Replace upstart with systemd unit in crm::dash - https://phabricator.wikimedia.org/T154264#3855208 (Ejegg) a:cwdent>jgleeson Fixed by I305845cab26b1bee [19:41:47] Fundraising Dash, Fundraising-Backlog, MediaWiki-Vagrant: Fundraising dashboard vagrant role is broken - https://phabricator.wikimedia.org/T162644#3855214 (Ejegg) [19:41:52] Fundraising Sprint Winter Wanderland, Fundraising-Backlog, MediaWiki-Vagrant: Replace upstart with systemd unit in crm::dash - https://phabricator.wikimedia.org/T154264#2905818 (Ejegg) [19:42:59] Oh, food's here. I'll be back in a while. [19:46:37] Fundraising-Backlog: Update vagrant main page test links - https://phabricator.wikimedia.org/T136379#3855228 (Ejegg) Open>Resolved a:Ejegg Fixed in f481568099904dc7 [19:55:49] well hello lowercase andyrussg [19:56:22] ejegg: hello ever-lowercase ejegg :) [19:56:37] just finally re-activated my vps [19:56:41] :) [20:16:46] fundraising-tech-ops: reimage tellurium to jessie - https://phabricator.wikimedia.org/T165391#3855310 (cwdent) [20:39:48] ok, seems like there's an issue with the https passthrough when the forwarded port is not 80 [20:40:01] that nginx conf file needs to be templated [21:27:37] XenoRyet: I was trying this, but it doesn't work: https://gerrit.wikimedia.org/r/399706 [21:28:03] XenoRyet: are you able to see anything on https? [21:28:24] i just get a bad gateway error from nginx [21:28:53] because it's using :80 for upstream and apache's only listening on 8007 [21:29:18] lemme see, maybe just easier to fix my port config? [21:30:02] Na, I haven't gotten anywhere with that either [21:42:10] XenoRyet: do you remember where the forwarded ports are configured for vagrant? [21:44:47] ah, there's something in .settings.yaml [21:45:11] sorry, my kids are going nuts over here, I'm a mite distracted [21:45:44] fr-tech this seems like the wrong time to decide this, but in upgrading my civi patch i think i did it wrong [21:45:52] i think it does need to be just one activity record [21:46:03] i'm going to chat with eileen after the holidays [21:46:35] XenoRyet: lmk when you're back [21:47:15] mepps sounds reasonable to me. what changed your mind? [21:47:31] just in debugging my code i'm having to look at the sample data [21:47:40] and i just think it's messy for the remaining contact to have two activity records [21:47:54] and the records are very similar [21:48:04] so it just seems like extra, useless data [21:48:50] i originally had it this way but i think after i found the issue where they originally implemented this i thought there was some rationale for two records [21:48:56] but in talking it out with eileen now i'm not so sure [21:49:21] not now, but as i worked on this issue [21:49:21] right, the issue with one record linked to both was how exactly to link the second, right? [21:49:40] In the case of a manual merge, you'd want the person who did the merge to be linked too [21:49:45] yeah but i think what makes sense there is if the dleeted contact is the assignee [21:50:02] because neither record was using assignee previously [21:50:10] and it does capture the person who does the merge already [21:50:29] ok, so 'asignee' is a free slot, but it's a little funny semantically [21:50:30] running to the store, will have phone and computer [21:50:42] yeah that is the downside [21:50:47] wouldn't asignee usually be someone who has a civi account? [21:51:08] i'm not sure, but we're just using it the other way hwo it's currently patched [21:51:26] as in, currently the assignee is the newly merged contact record [21:51:33] I still think there should be something more fundamental than the activity linking the old + new [21:51:49] like, a whole new merged_contacts table [21:51:54] i wonder if i can re-find the issue where they decided this [21:52:36] i mean to have two activities after merge [21:52:39] where the old->new pointer actually gets updated every time the new contact is re-merged [21:53:15] https://issues.civicrm.org/jira/browse/CRM-18106 [21:53:29] so you can have an O(1) lookup to make the link from the deleted contact to whatever active contact currently holds its stuff [21:53:42] and show that link right on the main contact summary [21:53:59] and it looks like eileen considered the options around the assignee too [21:54:47] i'm curious if it's something they've considered before... [21:55:04] hmm, 'after discussing on irc' says the ticket [21:55:09] was that this channel? [21:55:22] so it looks that's in json ^^ [21:55:33] prob civi irc [21:56:11] sorry, what's in json? [21:57:06] Eileen's comment: After discussing on IRC the plan is [21:57:06] 1) 2 activities, source is the logged in user, target is kept contact for 'Contact Merged' activity and trashed contact for 'Contact Deleted by Merge activity'. parent_id field links the 2 [21:57:06] 2) Store a json of moved ids in the activity (along the lines of {'contribution': [21:57:06] {'id' :1, 'id:2} [21:57:11] though i'm not seeing them in the db records [21:57:26] did she mean on the contact? [21:57:53] nope not seeing it there [21:59:08] wild animals I tell you [21:59:21] hmm i'm not sure that got implemented [22:00:23] ejegg: Anyway, I just fixed an issue I had with my test creds, and the Amazon widget is working fine for me now. Didn't have to do anything other than enable that role. [22:00:26] gotta run ejegg but might be around later [22:00:28] hmm, ok [22:00:36] XenoRyet: ah, nice! [22:01:23] XenoRyet: OK, so you see the comment on line 322 of the amazon adapter? [22:01:51] In the sandbox, you request specific error conditions using the seller_authorization_note [22:02:20] and the 'InvalidPaymentMethod' one ought to just show an error above the payment form [22:02:48] Ok [22:02:50] Try uncommenting it right now to see the current [object Object] behavior [22:03:44] then https://gerrit.wikimedia.org/r/399659 should show an actual message [22:04:00] Cool, I'll take a look. [22:05:38] thanks! [22:16:07] argh, i was blissfully unaware of the xmas music here for a whole hour [22:16:12] now i can't un-notice it [22:21:23] ejegg: Spoke a little too soon on the widget. Got to the login portion. Do we have a test user for the sandbox lying around somewhere? [23:19:32] going to get some dinner