[16:50:30] ccogdill: Are you in the office today? [16:50:43] I’m WFH in the morning, k4-713 [16:50:48] then office in the afternoon [16:50:57] Ah, okay. [16:51:12] wondering about the standup? [16:51:16] ccogdill: I'm trying to figure out if it's worth trying to have the chat standup, yeah. [16:51:31] So many people just bailed and I don't know who's left. [16:51:36] maybe we should just write ours on the thread too? [16:51:37] Want to just go email today? [16:51:40] heh [16:51:42] yeah I think it’s just the two of us [16:51:44] haha same page :) [16:52:07] Ppena is till at a conference.... what about the-wub? [16:54:08] hes out for the week [16:54:34] ccogdill: Well, shoot. [16:54:39] :) [16:54:47] so email? :) [16:55:07] Seems like the right thing to do at this point, yes. Heh. [16:55:11] haha okay [16:55:34] Also... I think I'm going to try to go over our regular meetings and see if there's something available on 3, if we usually get a room on 6. [16:55:39] ...unless you want to do that. [16:55:51] Standup said R67 this morning, and I didn't really have another plan. [16:57:23] yeah I’ve sort of given up on conference rooms for these 2 weeks but it’s worth a shot [16:58:39] Let me see what I can do. [16:58:59] Hopefully some people are bailing out and planning on WFH for the duration because "ARGH NOISE" or whatever. [16:59:38] pizzzacat: Are you here? [17:00:03] I'm here online [17:00:09] or did you mean IRL? [17:00:12] because no [17:00:20] Heh, either way. [17:00:25] What are you up to today [17:00:26] ? [17:01:06] right now I'm reading an article about how people say they want simplicity in features design but actually they want all the features :p [17:01:20] but in general I'm doing WorldPay, slides, and prototyping today [17:01:21] Heh. [17:02:16] pizzzacat: Oh yeah: It seems those slides will not be wasted. :) [17:02:29] orly? [17:02:36] I'm not sure how much time we'll get yet, though. More on that later, probably. [17:03:17] Are you still interested in doing at least part of the presentation? [17:03:25] ah I see. yeah I'm more just viz'ing the graphs we want, which are also reusable, and some can go into the public report hopefully [17:03:52] if we don't use them all it's still interesting hopefully on meta or public wikis somewhere [17:03:59] K4-713: "When you are writing automated tests, including unit tests, make super sure that your tests are not contacting the "outside world" (other sites, Tool Labs tools, financial services, non-Wikimedia APIs, and so on) unless you specifically mean to." - is that about right? [17:04:44] sumanah: I'd go so far as to say "Don't ever contact third parties in automated testing." [17:05:18] Got it! Thank you K4-713 [17:05:20] sumanah: But, it's been implied in the past that I'm unusually twitchy about that. :) [17:05:34] Probably that incident with the bicycle couriers. Heh. [17:05:47] Heh! Thanks K4-713 [17:06:01] No problem! Let me know if you could use anything else. [17:06:09] * pizzzacat gets the reference [17:06:34] It's now up K4-713 https://www.mediawiki.org/wiki/Talk:Security_for_developers/Architecture#Idea_re_unit_tests [17:06:47] I'm sure my old colleague would be *thrilled* I'm using his momentary lapse as a go-to example. Heh. [17:06:50] Next: I add the "good example" re ripping out the "talk to the outside world" functionality [17:06:59] He had to send out *so many* apology fruit baskets. [17:07:20] "sorry we went bananas, enjoy this fruit basket!" [17:07:49] sumanah: Yeah... so this is one of those things where it might be good to imply that all third party communication should use one function that's easily overrideable. [17:08:04] People run into issues if they don't do that. [17:08:30] K4-713: did you need me for something or were you just checkin' up? [17:08:52] pizzzacat: Er, I needed you for the check up. That's it. :) [17:09:37] ok cool. [17:11:54] sumanah: I guess what I'm getting at is "wrap anything that you will need to override for testing". In this case, cURL functionality. Does that make sense? [17:12:50] Or even... "wrap anything that will make testing tricky." [17:12:59] ...then you can just not go there. [17:30:37] (Abandoned) Awight: Partially revert "additional TY translations ready to publish" [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140134 (owner: Awight) [17:30:57] K4-713: I added your thoughts to https://www.mediawiki.org/wiki/Talk:Security_for_developers/Architecture#Idea_re_unit_tests so people more knowledgable than I can use or comment. Thank you [17:31:34] woot. [17:32:05] sumanah: Awesome. I think this is the second time I have publicly had an opinion. Exciting. :p [17:32:13] :) [17:35:12] baaahahaha sumanah: you know about our secret IRC channel, right :p [17:35:29] awight: uh no? what? [17:35:37] LOL oh just kidding [17:36:01] K4-713: https://www.mediawiki.org/wiki/Talk:Security_for_developers/Architecture#Good_example_of_simplicity - on simplicity [17:36:05] @Khorn (WMF): suggested an example from Fundraising. The DonationInterface extension "provides fundraising mechanisms for collecting and tracking payments through various payment gateways." [17:36:10] Thus the team wrote unit tests that ensure DonationInterface correctly reads from and writes to those gateways. Instead of writing dozens of individual mocks that might fail and thus incur charges, add noise to logs, cause inaccurate financial reports, or worse, the team: [17:36:10] subclassed each unit test from a class that managed all interaction with gateways [17:36:11] for the tests, overrode the "actually talking to the outside world" functionality and silenced it [17:36:11] Thus it's simpler to add a new gateway and write a new test, and we're less likely to accidentally break security rules with a bad test. [17:36:53] K4-713: so lemme know if I inaccurately wrote down what you told me, and please also gimme a link to the relevant code [17:37:28] sumanah: The more I think about it, the more using DonationInterface as an example of simplicity makes me giggle. [17:37:46] I hear ya. But I think *this one example* totally makes sense [17:37:48] ^^ +1 [17:38:17] "There is exactly one simple thing in DonationInterface, and here it is." [17:38:22] heh [17:38:54] "Khorn has assured all of us that it was, in fact, deliberate." [17:39:03] <_< [17:39:04] >_> [17:40:44] sumanah: Yeah, there are things that should be changed in there. [17:41:00] For clarification, we do mock gateway communication. It's not really an "instead". [17:41:21] We just started by damaging the default, for increased safety. [17:42:23] ...where in this case, the default was actually talking to other systems. [17:52:38] K4-713: ok. anything else? and should I be looking in DonationInterface/gateway_common/gateway.adapter.php ? [17:53:07] * K4-713 looks conflicted [17:53:26] So, as it turns out, the concept is much simpler than the actual code. [17:53:30] :/ [17:54:15] I think I came down with selective amnesia about the actual code, because it's... brutish. All my fault, too. [17:54:51] The problem is that php doesn't do multiple inheritance.... [17:54:53] * sumanah looks for the "explode" function [17:54:58] * K4-713 looks foolish [17:55:00] not sure where it is defined [17:55:15] explode? That's a php thing. [17:55:28] Unless you were not being literal. :) [17:55:29] that's not relevant actually, nm, I was just curious [17:55:29] the name caught my eye [17:55:40] It's among my favorites. [17:56:13] I did not know that's built into PHP! [17:56:13] * sumanah is enlightened [17:56:21] Just takes a string and creates array elements out of it, separated by some specified substring... [17:57:04] So, if you explode a string that is actually a human-readable statement in English, and explode it on spaces, all the words become array elements. [17:57:44] ...with awkward punctuation hanging around. [17:59:52] no prob. got it. [17:59:52] thanks for the explanation [17:59:53] now, back to DI ..... I can just point to that file if that's where this lives mostly [18:03:16] sumanah: Yeah, speaking of explosions... [18:03:45] ...that one file, is sort of ridiculous. Let me see if I can explain: [18:04:15] Every 3rd party gets its own "gateway" class on our end, in the actual code. [18:04:30] These classes are all descended from a common parent class. [18:04:50] For the tests, we have to do one further descendant for each gateway class... [18:05:05] ...and each one overrides the same parent communication function. [18:05:33] So, it looks like a totally insane amount of code duplication. Because you can't inherit from multiple classes. [18:06:11] nod [18:06:20] But, given the choice between ugly production code and ugly tests, I decided to shovel all the insanity into the test code. [18:06:28] Until we think of something better. [18:07:05] The relevant file that creates all these descendants for testing, is... [18:07:11] * K4-713 looks at own checkout [18:07:49] ejegg: u can try #wikimedia-staff now [18:08:10] pizzzacat: on that note, you have access to ^^ ? [18:08:21] awight: Wow. Who did you have to ambush? [18:08:53] woohoo! thanks for getting me in [18:09:06] yes I do awight [18:09:21] pizzzacat: okay you are just ignoring the gossip train then [18:09:25] wise move [18:09:38] yeah I can't figure out how to send messages in there, there's no dropdowns [18:09:49] pizzzacat: baaahahaha [18:09:54] :p [18:09:59] didyu try pasting a graph? [18:10:19] sumanah: Sorry: Distracted by RL. Check out DonationInterface/tests/includes/test_gateway/test.adapter.php [18:10:19] * sumanah listens to K4-713 and is in suspense :-) [18:10:23] Thank you! [18:10:24] I turned on my video and mouthed the words I wanted to chat [18:10:35] but it still didn't work [18:10:47] sumanah: I am partially mortified to be sending anybody to that file. [18:10:49] :) [18:10:50] * awight waves at keyboard [18:11:09] I then tried talking with my eyes like Tyra Banks [18:11:17] still....nothing [18:11:20] pizzzacat: ... [18:11:20] sumanah: I swear we do know about multiple inheritance. We just can't have nice things until php 5.4 [18:11:22] BAHAHA [18:12:16] awight: I can't wait. To collapse that file to 1/5th of its current size. [18:12:54] sumanah: Maybe you could reclassify this as an example of desperation in code. [18:13:03] heh [18:13:04] K4-713 oh we're just being silly because I just had a majorly derp moment with Bugzilla interface [18:15:24] there is cute on my desktop if anyone needs a fix [18:15:50] awight is not making it up. [18:16:38] awight: Now come over here and check out this stupid diagram. [18:16:46] This is not a fair trade. [18:20:11] what is this cute [18:20:18] is it another baby [18:28:25] https://www.mediawiki.org/wiki/Talk:Security_for_developers/Architecture#Good_example_of_simplicity is better. thank you K4-713 - will improve further [18:31:52] (PS1) Awight: Translations of Lila's TY letter [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 [18:32:50] mwalker: K4-713: ez and fun ^^ [18:34:48] sumanah: haha, nice. I'll think about this on and off as the day goes on... maybe there's something more useful I can point at, too. [18:35:34] thx! [18:50:19] (PS1) Awight: Delete all old TY translations [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140437 [18:58:20] (PS1) Awight: Slam the door to old TY generators [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140440 [19:57:31] K4|fooding: https://meta.wikimedia.org/wiki/Fundraising/Translation/Thank_you_email_20140606 [20:00:45] Jeff_Green: we need that http_fails thing in Drupal! [20:00:58] ? [20:01:01] Everyone's first request to Civi times out or takes 5 min it's terrible [20:01:09] oh [20:01:20] isn't it a knob you can adjust and blast? [20:01:53] $conf['drupal_http_request_fails'] = TRUE; [20:01:56] iirc drupal still uses unpuppetized config files, manually edited at boron:/srv/ and sync_blasted [20:02:01] oh! [20:02:02] thx [20:03:01] Jeff_Green: ah well so I can edit the sites/default/settings manually, but eventually I want that to be in the /etc/fr file [20:03:06] which can wait [20:03:37] we just need to coordinate on it I guess? [20:04:46] ARgh, I guess I set it to FALSE. [20:05:17] Jeff_Green: I don't have perms it seems [20:06:08] Jeff_Green: oops wrong machine, ^^ that was probably a lie. [20:06:25] no I was accidentally correct [20:08:16] ha [20:08:19] looking [20:09:17] boron:/srv/org.wikimedia.civicrm/drupal/sites/default/civicrm.settings.php ? [20:09:33] or [20:09:38] boron:/srv/org.wikimedia.civicrm/drupal/sites/default/settings.php ? [20:12:00] the latter [20:12:12] Jeff_Green: I think we should try flipping FALSE to TRUE [20:12:32] ok [20:13:04] blastered [21:37:54] Uhm. More ridiculous documentation. First they want me to disintegrate captured mandates. Now, there's a "Boarding procedure". [21:38:06] What am I boarding? [21:38:08] I don't know. [21:38:14] But I'm taking all of you with me. [21:38:56] * awight is whisked down the moving sidewalk [21:39:08] PACK YOUR BAGS or something. [21:39:22] or they will be disintegrated [21:39:30] Only if I capture them. [21:40:57] AGH! All it says in the "Boarding procedure" section, is "There are no additional boarding requirements for merchants". [21:41:05] That's... comforting? [21:43:18] I wonder if this another example of weird airline requirements that throw back to version one of so many random APIs that haven't been particular to the airline industry since... they paid for round 1. [21:43:50] I mean, I hope so. Otherwise, I have literally no clue what they're talking about. [21:45:36] the bags will be dematerialized [21:46:54] atgo: I need one of these to be SEPA compliant. http://www.onepagewonder.com/tron_laser.jpg [21:47:44] Sorry about the late notice, but I don't make the rules. [21:47:46] ... [21:47:47] :) [21:48:24] K4-713: u on that CR? [21:48:31] or shall I bug others? [21:48:46] Agh. I can do that. [21:48:55] * awight dangles a carrot [21:49:11] Which commit, specifically? [21:49:20] I seem to have lost my mind. [21:49:39] https://gerrit.wikimedia.org/r/140432 https://gerrit.wikimedia.org/r/140437 https://gerrit.wikimedia.org/r/140440 [21:49:55] That's three commits. [21:50:06] u got it [21:50:11] interest. [21:50:20] Also I don't speak those languages. [21:50:35] * K4-713 shrugs [21:51:07] awight: Did you change anything substantial about the content... areas, I guess? [21:52:07] K4-713: The first patch should have pulled new translations from meta.wmo [21:52:14] the second just deletes stale translations [21:52:26] the third is a safety thing for the future [21:52:39] awight: What's up with the new style var reference in first name here: https://gerrit.wikimedia.org/r/#/c/140432/1/sites/all/modules/thank_you/templates/html/thank_you.cs.html,unified [21:52:58] hmm maybe we can ge that in the budget [21:53:23] awight rosie is suggesting waiting until caitlin v is back to do the case demo [21:53:31] atgo: that works [21:53:54] (CR) Katie Horn: [C: -1] "I found a thing. Inline. Probably more places." (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 (owner: Awight) [21:54:10] she's back june 30 [21:54:50] atgo: well... I think multiple iterations is a good idea here, unless you think it's offensive / a buy-in issue [21:55:20] i could go either way. just given the lck of rooms... [21:55:31] apparently sylvia would like to do it sooner than later [21:55:34] awight: Do you want me to find more of these things? There's another one in nb. [21:56:06] Confusingly, it's not everything, or even most of them. [21:57:22] awight that's fine. i'd like to be there to some extent, but if tomorrow's the day that's fine [21:57:32] we can do a followup with virtue when she's back [21:58:11] (CR) Katie Horn: "Okay, only one more place." (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 (owner: Awight) [21:58:39] awight diderot is open at 3 tomorrow [21:58:43] (PS2) Awight: Translations of Lila's TY letter [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 [21:59:27] and friday isidore is open 1-2 [21:59:46] next week is less terrible [21:59:53] awight: So... deleting nearly 50 files is okay because then they'll default to en? [22:00:04] yah i hope so [22:00:10] ...was that for me? [22:00:16] K4-713: yes [22:00:21] hehe, okay. Just checking. [22:00:48] K4-713: although it's definitely one of those things I should... test. [22:01:02] * K4-713 squints with increasing suspicion [22:02:09] ...can you do that now? :) [22:02:12] code withers correspondingly-- oh hey it's all gone! [22:02:22] lol if I got off my lazy feet [22:02:34] hey pizzzacat do you want to call in for standup in 30? [22:02:55] I was hoping to just email it? [22:03:03] sure! sounds good [22:03:09] :) [22:03:20] <:o) [22:04:13] pizzzacat do you think you'll be able to look at the RTL language issues before the conference? just checking to set expectations for prod team [22:04:17] card #16809 [22:04:20] 1680 [22:04:27] dear god. that would be a lot of cards. [22:04:36] yes :) [22:05:00] thanks! [22:09:13] (CR) Katie Horn: "Okay, just one more inline issue." (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 (owner: Awight) [22:10:56] (CR) Awight: Translations of Lila's TY letter (1 comment) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 (owner: Awight) [22:11:22] (PS3) Awight: Translations of Lila's TY letter [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 [22:13:30] K4-713 i'm thinking about making anew business goal that would encapsulate all the requests we get for data and the like from production team [22:13:34] objections? [22:13:57] Not at all. Please do. [22:14:16] I'd like to get a handle on the volume there, anyway. [22:14:37] yeah... i'm thinking we can split it out into capabilities about 1- data requests 2- feature changes 3- translation pushes, etc. [22:15:30] (CR) Katie Horn: [C: 2] Translations of Lila's TY letter [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140432 (owner: Awight) [22:16:33] atgo: Is this also going to encompass test result digging=type things? [22:16:45] K4-713 i think it should [22:16:50] * K4-713 nods [22:16:51] Cool [22:17:09] awight: Any verdict on the english fallback through deletion? [22:17:40] oooh. /me tests [22:17:46] * K4-713 twiddles [22:17:54] u can also test :p [22:18:03] Hmmmmmmmmmm. [22:18:18] http://CRM/admin/config/thank_you/test [22:18:26] Is this one of those things I can do through the... yes, that. [22:18:29] okay then. [22:18:40] whitescreen... [22:18:52] oh it's just me [22:19:54] K4-713 i kind of also want to sort out a way to track "this donation is doing something weird" [22:21:15] Oh... like "The currency spontaneously changed from euros to korean won, sometime between the point we requested the payment and when it showed up in the nightly reconciliation file?" [22:21:49] .... yeah [22:21:54] I like that. Because then we can make a pie chart of gateways that transmogrify donations. [22:22:00] oh hah the test page does not allow u to choose nonexistent languages :-/ [22:22:03] ...and present them with a little award. [22:22:55] K4-713: k, tested. [22:23:06] How did you... do that, if they weren't presented? [22:23:11] I hacked the form [22:23:21] I guess... we need that. [22:24:37] awight: Buh. What... happened to the dependency here? https://gerrit.wikimedia.org/r/#/c/140437/ [22:24:48] (PS2) Awight: Delete all old TY translations [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140437 [22:24:52] (PS2) Awight: Slam the door to old TY generators [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140440 [22:25:14] Okay, now it's not mad anymore. [22:25:53] (CR) Katie Horn: [C: 2] "DELETE." [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140437 (owner: Awight) [22:26:32] (CR) Katie Horn: [C: 2] Slam the door to old TY generators [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/140440 (owner: Awight) [22:26:49] awight: Okay, done. Can I go back to my incomprehensible documentation now? [22:46:09] mwalker: https://meta.wikimedia.org/w/api.php?action=query&format=json&meta=messagegroups&mgformat=flat&mgprop=workflowstates&mgroot=Centralnotice-tgroup [23:07:34] (PS1) Awight: Translations of Lila's TY letter [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140595 [23:07:36] (PS1) Awight: Delete all old TY translations [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140596 [23:07:38] (PS1) Awight: Slam the door to old TY generators [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140597 [23:08:05] (CR) Awight: [C: 2 V: 2] Translations of Lila's TY letter [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140595 (owner: Awight) [23:08:12] (CR) Awight: [C: 2 V: 2] Delete all old TY translations [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140596 (owner: Awight) [23:08:19] (CR) Awight: [C: 2 V: 2] Slam the door to old TY generators [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/140597 (owner: Awight) [23:09:04] !log update crm from 26460d6eaec26861661322df8e9f07a8b0519677 to f3389daa94e9ad924175bdf0d5bc09c4a26aeb8c [23:09:09] Logged the message, Master [23:19:58] atgo: Uhm. Do you know if we need/want to use the SEPA logo somewhere? [23:20:22] K4-713 hadn't thought about it... does it say anywhere that we have to? [23:20:31] what's our typical practice around such things? [23:20:49] No... but they leave a conspicuously large version in one of their docs. [23:21:15] I'm not sure there is a typical practice yet. [23:21:30] I mean, in one sense, this could be sort of like a cc logo. [23:22:38] I mean, maybe we should test two forms and see if people... trust more if it's there. [23:23:03] It might make us more credible. Or something. [23:23:38] http://www.agion.nl/assets/sepa-logo.jpg [23:24:06] ...I like how it isn't a vector graphic for some reason. [23:28:35] yeah that's probably a good idea [23:28:43] no, but it's a HUGE jpg [23:28:52] single euro payments area. that's where we are. [23:29:02] pizzzacat: Can you just drop that image on a form? No resizing necessary. [23:29:03] :p [23:29:32] * pizzzacat reads [23:30:05] K4-713 question... do you know if it's possible to set a transition such that when a triage item is turned into a tech task it automatically sets "tech status" to backlog? [23:30:05] wow that is BEAUTIFUL [23:30:16] SINGLE EURO PAYMENTS AREA [23:30:25] AREA [23:30:38] because it's really frustrating that it doesn't do that. [23:30:53] atgo: As a matter of fact... [23:30:59] Do you want me to just do it, or show you? [23:31:06] I'd rather show you. [23:31:08] show me! i'll come over [23:31:11] woot. [23:31:31] do we srsly want this on the forms? I can play with it [23:31:43] It would look better if it were just a little bigger [23:32:12] also is this a Thing right now? I thought the process of becoming compliant was going to take a bit? [23:32:22] or maybe I'm crossing things [23:32:25] ha [23:32:43] pizzzacat it's not yet a thing, but is starting to be a thing [23:33:27] gotcha. [23:33:45] seems like there's always some new kind of money deal going on over there [23:35:03] yeah... so we probably have to deal with the forms for it in like.. a month [23:35:04] or so [23:35:22] Or sooner. [23:36:28] I think there are going to be new required fields with this. But, I mean, most of it I'll just add badly to the UI and then look uncomfortable until pizzzacat has time to make it pretty. [23:39:00] pizzzacat another question - for your WP form iteration 3, will that include the identical to GC form? i know it's in the card, but if you'd rather it be separated we can do that [23:39:31] Bah. [23:40:27] It seems like that one should be separated into its own card [23:40:49] because ideally I will just do the refactoring that K4-713 and I touched on a bit ago [23:40:57] and have all our forms working as one [23:41:39] when I talked to Peter recently, one of the things he mentioned as wishing to fix/get into our flow was the ability to do testing, similar to how we do banner testing, for forms [23:41:56] Oooh, which version of that thing we talked about? Because I got a bit carried away and started making things up. [23:42:07] ha [23:42:43] Is this the Grand Unifying Form theory? [23:42:47] welllll we just talked about kind of DRYing it up and having a more centralized way to serve forms [23:42:49] yes [23:42:52] Ah. [23:42:57] is that no longer a thing we want? [23:43:05] No, we absolutely want that. [23:43:16] It's just... probably several iterations down the road. [23:43:26] oh, ok. [23:43:35] I'm afraid if we try to do that all in one shot, we're going to go crazy. [23:44:01] so then is it not worth it to try to make GC forms look the same as WP forms? [23:44:02] So, whatever you do should definitely bring us into a position to make that easier. [23:44:09] Er. [23:44:21] Megan *really* wants to test as close as we can get. [23:44:36] I tried to talk her out of it. [23:44:39] I really did. [23:44:41] I guess it's just copy-pasting css [23:44:50] mostly [23:44:51] (PS1) Mwalker: WIP: Python Silverpop Export [wikimedia/fundraising/tools] - https://gerrit.wikimedia.org/r/140613 [23:44:52] arghblargle. [23:44:56] yeah. [23:45:35] I'm having a blazing argument with myself right now. [23:46:10] ...about whether or not we should try to make the old-style form work in the new system, or just cheap out heavily until we know it's worth it. [23:47:04] So, let's see... if we test them and the old style loses, we'll probably just reimplement all the forms in the new style. Slowly, as we can test them. [23:47:11] yeah either way takes time.. but one is somewhat more wasted-feeling [23:47:15] If, however, we test them and the old style wins... [23:47:27] ...then they tell you to kill the new style. [23:47:42] the cool thing is how I hate both styles. [23:47:46] bahaha [23:48:29] From my perspective, I'd much rather the old style loses. Then you get time to refactor everything nicely. [23:48:41] It's not clear what happens if the old style wins. [23:48:42] heh yeah. [23:48:48] how do we plan to test these things? [23:49:00] They probably say "why are you putting so much work into something we already have"? [23:49:06] * K4-713 sighs [23:49:06] I guess I'd just add in the old crapp…uh the old style css [23:49:11] heh [23:49:28] Well, I mean, the restructuring of all the old forms. [23:49:38] It's easier to sell up if we actually have new forms. [23:49:41] :/ [23:49:46] truth [23:50:16] As far as how we test these things: the-wub just adds the ffname to the parameters passed in from the banner. [23:50:46] So, one banner sends to the old form, and an identical-seeming banner sends them to the new one. [23:51:46] ah easy. [23:51:51] yep. :) [23:52:02] I actually really like how the form chooser ends up working. [23:52:07] The config suuuuuuucks. [23:52:11] But something has to, I guess. [23:52:14] :/ [23:52:50] I really wish I could make my dream forms [23:53:03] Never let go of your dream. [23:53:08] pahaha [23:53:24] the problem is I have too many dreams so I need to like, finish the dreams of today [23:53:33] before having the dreams of tomorrow [23:53:50] My dreams are largely structural. [23:54:58] mine are largely superficial. [23:55:13] We should be friends. [23:55:26] we should like, work together on something sometime [23:55:34] :p [23:55:48] :) [23:56:18] Back to the Thing: I'm hoping to arrange the work that whatever wins, it's not a step backward structurally. [23:56:59] ok. I would guess that the new forms would win over the old.. but if they don't I'd like to be able to drill into why [23:57:11] Another thing to keep in mind is that they're probably always going to make you prove that new things are better than old things via donor testing. [23:57:31] The trick is to set it all up so DonationInterface wins no matter what happens. [23:58:20] haha well if the old ones win I'm RLY interested in why. [23:59:09] awight, are you still around?