[00:08:26] (PS1) Ejegg: Use UTC to figure out last point on USD/sec graph [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/179017 [00:22:13] (PS1) Awight: Legal-recommended update to English TY text [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/179021 [00:22:41] (PS1) Awight: other TY updates [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/179022 [00:22:59] awight: thanks re: code review! Also, yeah, very nice garden! [00:25:23] Silly etherpad for DI dumping, https://etherpad.wikimedia.org/p/DonationInterface_refactor [00:26:10] atgo: ejegg: oh, an important thing that should make it into this sprint: RLewis is having a terrible time merging records. [00:27:12] just fwd'd the email. [00:29:31] awight: dang, probably relates to the new fields [00:29:48] will look at error logs [00:30:02] ejegg: iono. I'm suspicious of the Individual vs Organization distinction... [00:30:40] maybe not. I don't see anything touch that distinction in the commit logs... [00:31:10] ejegg: rlewis said contributions would fail to reattach to the newly merged contact, they'd still be linked to the deleted one. [00:34:15] huh, that's definitely not related to contact type [00:35:49] awight it looks like rosie thinks it was user error.. i just talked to her and she's OK [00:37:09] pizzzacat1: That spike K4 saw in the meeting was from 7 hrs ago or so - I made a time zone error deciding where to end the $/sec graph [00:37:18] https://gerrit.wikimedia.org/r/179017 [00:38:03] ohhh [00:38:09] atgo_: she had this happen last week, and the contribution was not reattached. It doesn't sound like user error... [00:38:44] awight: i literally just talked to her IRL [00:38:47] she said she unclicked something [00:42:23] awight: definitely feel free to follow up, but she sent an email to the same effect [00:42:34] i'm going to pack up and head home. be back online in a bit [00:43:12] ejegg it's running fine for me but I don't have the most up to date data [00:43:46] maybe I should just go ahead and do an updated dump [00:43:51] atgo_: np. I asked her if she thought it was the same thing as last week. [00:44:32] pizzzacat: take a look at the last graphed point on the usd/sec graph [00:44:47] it will be the day & hour local time, not UTC [00:45:02] until that tiny fix [00:45:38] ah good point [00:47:43] (CR) Ssmith: [C: 2] Use UTC to figure out last point on USD/sec graph [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/179017 (owner: Ejegg) [00:47:53] (Merged) jenkins-bot: Use UTC to figure out last point on USD/sec graph [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/179017 (owner: Ejegg) [00:52:52] Aww, I was mean and the K4-713 went away. [00:54:39] (CR) Pcoombe: [C: 2] Legal-recommended update to English TY text [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/179021 (owner: Awight) [01:01:15] (PS1) Awight: Merge remote-tracking branch 'origin/master' into HEAD [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/179034 [01:01:38] (PS2) Awight: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/179034 [01:01:52] (CR) Awight: [C: 2] Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/179034 (owner: Awight) [01:01:56] (Merged) jenkins-bot: Merge master into deployment [wikimedia/fundraising/crm] (deployment) - https://gerrit.wikimedia.org/r/179034 (owner: Awight) [01:02:27] !log update crm from 3d657972029ea221b321470102c99ad74027b6f7 to 28b68e23b670fe52a401659bde800b64d05e25bf [01:02:32] Logged the message, Master [01:15:10] Happy Freudian slip: "diagnore" [01:15:18] it's fixed! :) [01:18:50] (PS2) Ssmith: Updated chart for basic design. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/178995 [01:19:04] (CR) jenkins-bot: [V: -1] Updated chart for basic design. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/178995 (owner: Ssmith) [01:19:09] (PS1) Ejegg: Update minfraud API to 1.60 [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/179039 [01:40:21] K4-713: fyi, I was wrong about the payments_fraud_breakdown thing. [01:40:27] foiled by identical column names [01:40:39] Aha! [01:40:50] Glad to hear I left a bear trap in there after all. [01:41:09] nah I think I just failed to scroll right or something inexcusable like that. [01:43:30] ejegg: K4-713: hey, where is #1957 at? I see some conversation started in CR but then it looks like u all went IRL or something. [01:43:53] (CR) Krinkle: "Haven't verified locally, but assuming this works, this looks good." [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/178974 (owner: AndyRussG) [01:43:59] awight: Ah, that. [01:44:31] So, this conversation went all over the place. [01:44:48] Here's where I think we ended up: [01:45:13] We currently have a test url for the .css, which is exactly where the real .css used to live. [01:45:18] On... foundation wiki. [01:45:38] If you change the css there, you should be able to test it if you point your dev box to the GC sandbox. [01:45:57] yep [01:45:59] Additionally: You moved the prod .css to payments. [01:46:03] y [01:46:18] and I rebased on top of that, so it's ready to go [01:46:35] So, once the test thingy looks good, we can copy it to prod css and squish that in with the form changes. [01:46:42] the CR-1 comment sounds ominous, maybe just update that? [01:47:06] Oh, well, that may actually be real. [01:47:17] :) I will stay out of it for today [01:47:27] The .css in the iframe is not the same .css they were injecting into the forms via the appeal. [01:47:27] I'm being rapidly familyjacked n e way [01:47:41] awight: enjoy! [01:47:41] yikes! [01:47:49] That's what I said. [01:47:56] It just looks like "-1". [01:48:06] And then I went "You're doing what with the what?" [01:48:20] K4-713: the css for the the iframe in the commit is the stuff from the mingle card [01:48:49] Right, and that's cool. But we need to update the form .css on payments at the same time. [01:49:04] that commit also adds the input to the real form css [01:49:20] (PS3) Ssmith: Updated chart for basic design. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/178995 [01:49:25] (CR) jenkins-bot: [V: -1] Updated chart for basic design. [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/178995 (owner: Ssmith) [01:49:27] I guess we just need to see if that interacts with the stuff the appeal pulls in. I'll export the appeal [01:49:40] It should be identical with what the appeal pulls in. [01:49:40] and take a look at the whole shebang locally [01:49:55] ah, let me see what that is then [01:50:18] the stuff in the commit is just copied from the mingle card's iframe stuff with a couple selectors changed [01:50:39] Yeah, that jives with what I thought was where we had left it. [01:50:50] I'd test locally, but I'm all hosed. [01:51:12] Oh, fooey. [01:51:19] * K4-713 kicks GC audit job [01:52:00] Guess we actually started timing out. [01:52:23] It's almost like we've been under unusually high load for a few days or something. [01:58:05] (CR) Awight: [C: -1] "still not ready for master AFAICT?" [wikimedia/fundraising/dash] - https://gerrit.wikimedia.org/r/178994 (owner: Ssmith) [02:00:02] kbye! [02:08:09] (PS3) Ejegg: Update input and select styling for GlobalCollect [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/174617 [02:08:56] That just sticks the css from the appeal at the bottom of the gc sheet. I'll do another change to clean up overridden old stuff [02:25:26] (PS1) Ejegg: Clean up GlobalCollect style sheet [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/179052 [02:43:09] ejegg: yt? [02:56:20] oh hai [02:56:37] I just sent an email. Not serious, but... [02:56:56] ...I figured out what was breaking DI forms on my local. [02:57:09] Oh? [02:57:19] not seeing the email yet. Subject? [02:57:29] oh, htere it is [02:57:34] The problem went away completely, when I disabled CentralNotice. [02:57:40] whoa! [02:57:49] It was like... wait. [02:57:53] I was certainly using the same wiki for both at one point [02:57:55] Did you ever play old Sierra games? [02:58:17] which did they put out? [02:58:19] Space Quest, King's Quest... LSL? [02:58:39] This probably also applies to LucasArts' Monkey Island, and things of that nature. [02:58:41] naw, guess I missed those [02:58:51] oh, monkey island was fun! [02:58:57] That kind of a puzzle, huh? [02:59:01] Well, there's this mental state in which you start trying complete nonsense, because you're out of ideas. [02:59:17] That was me about 45 minutes ago. [02:59:23] Only, with LocalSettings. [02:59:48] I remember something about the JSON shim being disabled at least in the version we were running till last month [03:00:04] let me see if I get the same issues [03:00:39] It's not really a big deal at the moment, because as I said, I may be the only person anywhere trying to have them both on the same wiki. [03:00:44] But. [03:00:54] well, I definitely have cn commented out in my payments LocalSettings [03:01:00] My point is roughly /headdesk [03:01:05] That is all. :) [03:01:29] heh. [03:01:45] cn really ought to play well with other extensions! [03:02:03] The cool thing about the sierra games is that until very late, you had to type commands. [03:02:15] So, things like "Eat Desk" would be okay to try. [03:02:30] Eew! I can't eat that! [03:02:48] That's basically what the game would say. [03:02:52] "You can't eat that" [03:03:34] Anyway, it's dark outside, and I am increasingly likely to be eaten by a Grue if I don't leave the office soon. [03:03:41] like Zork, but with a sense of humor [03:03:44] heh [03:03:49] :D [03:03:51] beat me to the reference [03:03:55] have a good night [03:14:52] K4-713: totally the json dependency [03:17:47] We should make a card for that, then. [03:18:19] or upgrade payments before we run CN there! [03:18:24] baha [03:18:36] the json RL module just doesn't exist in the version we're running [03:18:44] Fundraising banners on donation interface... [03:18:47] Who in their right mind would run those both on the same... :) [03:19:09] So you're saying, most people at the WMF would do it...? [03:19:13] "While you're here, would you like to give us more money?" [03:19:18] heee [03:19:49] Okay, well, then I'm definitely over it. After a beer. [03:20:16] How about an algorithm that detects rebellious OCD and in those cases shows a banner saying "don't donate again! that was enough! don't do send more! really!" [03:20:53] If you have an algorithm to detect OCD... would that thing ever stop detecting? [03:20:59] Probably not. Not for me, anyway. [03:21:17] Probably not for most wikipedians [03:21:28] Maybe we should run it against new employees. [03:21:45] "Sorry, but it just... wouldn't work. Trust me, you'd be happier elsewhere." [03:21:58] filter out the non-weirdos [03:22:07] something something culture fit. [03:22:08] Detection algorithm: do you have a WP account? Severity measured by # of <10 char edits [03:22:33] Hi Krinkle how's it going? :) Thanks for the CR! awight mentioned earlier that you might have done some preliminary work towards separating out different tests in bannerController.lib.tests.js...? Just thought I'd ask quickly before diving in... thanks! [03:22:56] Hmm [03:23:16] Hey: I have an idea. We should all stop doing things for a few hours. [03:23:23] no way [03:23:35] Well, things that show up in public repos, anyway. [03:23:45] Ah hmmm [03:23:46] K4-713: ejegg|away: I'm just remembering a skit from a comedy show decades ago with this crazy guy holding a salt shaker [03:24:02] I want to know more. [03:24:04] So he keeps saying to himself, "Don't put salt in our eye! Don't put salit in your eye" [03:24:09] Again and again, faster and faster [03:24:29] And eventually it morphs into "Put salt in your eye!" and he does it [03:24:39] And screams "aaargh" and starts over again [03:25:06] I have never seen that skit. But the premise is hauntingly familiar. [03:25:09] (u sure u wanted to know?) [03:27:34] K4-713: found it... https://www.youtube.com/watch?v=_83MEuLoz9Y [03:27:44] Not exactly as I remembered it [03:28:06] Actually funnier [03:28:09] Kids in the Hall! [03:29:28] See, _that's_ the Canadian culture I brought my kids here to learn about [03:29:56] * K4-713 giggles at the chart gag [03:30:13] Quick, add a link to the dash! [03:30:20] :D [04:01:47] (CR) AndyRussG: "Thanks, @Krinkle!" [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/175732 (owner: AndyRussG) [04:04:28] ejegg|away: now I'm getting the non-WP infobox :( https://www.google.ca/search?q=jon+stewart&ie=utf-8&oe=utf-8&rls=org.mozilla:en-US:unofficial&client=iceweasel-a&channel=sb&gws_rd=cr&ei=tBeJVMS0OoSzyQT89oGIDA [17:10:41] * AndyRussG waves [17:13:18] hi AndyRussG [17:13:35] He y :) [17:15:14] This morning I lost what I thought was a medium-sized bag of kitchen garbage [17:15:30] I was like, where did I put that? I thought I was loosing my mind 8p [17:15:57] After like 20 minutes i realized it'd gotten squished down and lost when I put it in the recycling bin [17:16:21] One of my most original wastes of time ever I guess [17:17:13] heh [17:17:56] I have this new pasttime of finishing my cat's discarded hunting... [17:18:28] When I hear him chase a rat into the house, I then have to tear the place apart and find the poor wounded animal's final resting place [17:18:31] before it decomposes :( [17:18:41] Usually it's somewhere horrific under our bed [17:19:18] aww that's sweet [17:19:25] our cat only barfs under our bed [17:19:32] Yes he's very generous with his spare meat [17:20:13] "Have a half-mouse-foot...! Lemme hide it so you can have some fun catching it, too." [17:21:04] Our cat doesn't even want to go outside now... [17:21:20] And... now it's raining on top of the snow... [17:21:27] BTW how's the wind and rain there? [17:22:18] I think the rats come to the surface when it rains... my cat is very interested in getting rained on then running inside. [17:22:56] Hmmm [17:24:53] At least he's got a plan, it sounds like [17:29:17] jumping someone else's train... [18:13:16] (CR) AndyRussG: "Sweet! Successfully smoke tested..." (3 comments) [extensions/CentralNotice] - https://gerrit.wikimedia.org/r/177475 (owner: Awight) [18:14:23] (PS1) Ejegg: Merge branch 'master' into deployment [wikimedia/fundraising/dash] (deployment) - https://gerrit.wikimedia.org/r/179163 [18:14:39] (CR) Ejegg: [C: 2] Merge branch 'master' into deployment [wikimedia/fundraising/dash] (deployment) - https://gerrit.wikimedia.org/r/179163 (owner: Ejegg) [18:14:52] (Merged) jenkins-bot: Merge branch 'master' into deployment [wikimedia/fundraising/dash] (deployment) - https://gerrit.wikimedia.org/r/179163 (owner: Ejegg) [18:16:15] !log updated dash from 08b078acf904d563030ff7a37b2af8df88387e29 to 6631a97e5e3e688bc0f4d2a1f6f5d97744dba0f4 [18:16:19] Logged the message, Master [18:22:48] K4-713: did you come in or wfh? [18:22:56] atgo: I am in the office. [18:23:11] Phab tech talk: http://www.youtube.com/watch?v=_yr5z9Ix2f8 [18:23:14] so... did you walk from embarcadero? [18:23:18] yeah i'm watching it AndyRussG :) [18:23:23] cool! [18:23:28] atgo: Yep. [18:23:42] And I've got to say: I am seriously underwhelmed by this drizzle. [18:23:42] K4-713: you are hard core [18:23:52] yeah i think it was worse earlier and also farther north [18:24:00] I hope so. [18:24:05] i forgot to bring in some chairs from my deck and they are all over the place [18:24:10] Because that was embarassing. [18:24:29] they predicted 40 mph winds around 6am [18:24:34] not sure what actually happened [18:26:07] There were a couple dead branches in my front yard, but... [18:26:14] ...yeah, I don't know either. [18:26:47] I am pretty happy I pruned around the data line going in to my house, though. [18:26:53] It sort of goes through the middle of a willow. [18:26:54] that was smart [18:27:07] I'll let a lot of things slide... [18:27:11] ...but not the data. [18:27:12] :) [18:27:44] Sometimes you just need to get a pole saw. [18:27:53] ...for the data. [18:29:07] atgo: So, how are we doing otherwise? Anything good? [18:29:27] Or, bad and new? [18:29:32] ... same? [18:29:40] Okay, great. [18:29:52] I feel like I've got a pretty good handle on old bad news. [18:30:00] one thing i heard this morning, apparently reports in civi are taking so long that caitlin v is giving up. ejegg do you have any ideas on this? [18:30:16] * K4-713 mumbles something about dedupe [18:30:24] yes. that i want. [18:31:19] There was a terrible idea we batted around for a while, about trying to beat a read-only database connection in to civi core. [18:31:43] But, this turned out to be prohibitively weird. [18:32:54] Also, I just had a pretty good typo I can't quite get over: "brohibitively". [18:33:12] * Jeff_Green makes a mental note of this ^^^ [18:33:36] brohibitively... [18:33:42] so many good uses [18:34:35] Many events in my past have just been described with one word. [18:35:00] and it leads to brohibition too [18:35:10] good. [18:35:15] :D [18:35:15] :P [18:35:59] re. db, it's concievable we've suddenly crossed some kind of threshold where we were once making use of memory efficiently, and are now spinning disks to do queries [18:36:56] atgo, and I suppose ccogdill: When you say queries are so long you're giving up, is it actually you that's giving up, or is the system timing out? [18:37:04] caitlin v :) [18:37:04] Or, both? [18:37:07] Ah. [18:37:10] Er... [18:37:22] 5-10 minutes said cv and then she gave up [18:37:31] i told her to let it run and see if it ever made it [18:37:31] Also, are other people trying to run reports at the same time? [18:37:33] But, it was still spinning? [18:37:37] ejegg: i'm not sure [18:37:39] Yeah, okay. [18:38:01] i just pinged cv to join the convo [18:38:03] Right now, it's pretty easy to guess that some data shovel is running all the time. [18:38:20] So, we are probably... you know, Lugging down there too. [18:38:36] word [18:38:48] hmmm. dash is running off the replicated copy, so shouldn't be contributing to the problem [18:38:52] I’m not having that problem, for what it’s worth :) [18:38:54] We have had issues in the past, in which two people trying to grab audit data at the same time have timed eachother out. [18:39:06] Like, without Big EN being in full swing. [18:39:11] mhmm [18:39:18] yeah it's the same report they've been pulling daily [18:39:28] I don't know what that is. :) [18:39:29] well, i don't know that that's true, actually [18:39:33] but they have been pulling the same one daily [18:39:36] lybunt [18:39:42] this may have been something different [18:39:42] Jeff_Green: we found a search results cache table that doesn't need replicating, in case that would help anything. It gets a huge amount of inserts and updates. civicrm_prevnext_cache [18:39:58] Lybunt, like the totally new thing this year? [18:40:25] CaitlinVirtue: Hi! [18:40:26] lybunt is a built in civi feature that i believe was fixed for this year.. tbh i don' tknow :) [18:40:47] We were just talking about how much civi is ruining your process. [18:40:49] K4-713: theoretically, the new lybunt should be more efficient than the old one [18:40:51] @K4-713 hello! [18:40:54] ejegg: it probably wouldn't hurt, but I've never heard of queries backlogging due to what the replication thread is up to [18:41:09] but I don't think we proved that in any way [18:41:21] ejegg: More efficient, in the sense that it's not totally broken? :p [18:41:30] Jeff_Green: OK, so probably wouldn't solve our problem [18:41:57] K4-713: and also uses precalculated is_YYYY_donor columns instead of multiple joins to the contributions table [18:42:13] though it still joins at least once to contrib to get a previous year total [18:42:25] CaitlinVirtue: Can you tell us a little bit about what you were doing that stopped working? And, if there's a usual time of day that you do it? [18:42:45] sure, i run the custom lybunt each morning [18:43:01] ejegg: Yeah, I was sort of thinking that we might have to start thinking about caching query results. [18:43:03] ususally it takes about 2-4 minutes to run and report rsults [18:43:14] and then another 2-4 to export the results to csv [18:43:23] is that a query we could do on demand, i.e. while we're watching the db? [18:43:33] CaitlinVirtue: Okay. What's morning? I want to make sure we don't have some automated job hogging resources right then... [18:43:50] Jeff_Green: I think we just press a button somewhere. [18:43:57] it wasn't outputting any results the day that adam reset civi, but it was great after the reset [18:43:58] But, wait. That's like my entire job. [18:44:12] K4-713: reboot! [18:44:34] between 7 and 10 am depending on how anxious I am to see the data [18:44:47] Okay. Let me look at something really quick. [18:44:48] this morning i can't get any results at all. [18:45:00] CaitlinVirtue: Would you mind starting one while we're looking right at it? [18:45:14] sure. should I kill the one I have going too? [18:45:29] CaitlinVirtue: Have you tried running the old lybunt lately? If one runs and not the other, that could help us pinpoint the issuee [18:45:39] i'll try that right now [18:45:40] Oh, you have one going? [18:45:46] Hang on a second... [18:46:21] @ejegg wait, which olf lybunt? the orginal from civi? or LYBUNT MG 2014? [18:46:23] the query is copying to temp table [18:46:51] @K4-713 yes, that one's been running since 10:02 [18:47:16] Jeff_Green: Sounds right. [18:47:46] CaitlinVirtue: LYBUNT MG 2014 is an instance of the old one I think, so ifif you've run that and it worked I'd be interested to know [18:48:05] @ejegg ok trying now [18:48:18] oh, please wait till the new one is done! [18:48:33] ok, stopping [18:48:48] the new one has now been running for over 40 minutes [18:49:01] wat [18:49:05] Dang. [18:49:16] Yeah, that's... out of control. [18:49:29] it's doing temporary table and filesort on civicrm_contribution [18:49:31] yeah... [18:50:08] ejegg: Is there some of that we can avoid recalculating every time? [18:50:19] and that table is 4GB [18:50:23] Or, rather, more of that. [18:51:29] K4-713: I'll look at the remaining joins to civicrm_contribution and see if I can get rid of the one that rolls up last year's contributions [18:51:55] the query is like a page long, I can't begin to wrap my head around it [18:52:04] How are you defining "last year"? Calendar year? [18:52:12] fiscal [18:52:18] hurm. [18:52:30] If only we knew how long it took for donations to settle down... [18:52:31] Jeff_Green: mind dumping that query someplace? Would save me some time [18:53:15] @K4-713 just FYI it's fiscal for this report, but there's other instances where i would like to use calendar year [18:53:36] CaitlinVirtue: Yeah, the actual output shouldn't care. [18:53:56] K4-713: if we assume we only want to run lybunt for the present FY, adding one more column to WMF_donor could get that [18:53:57] I'm trying to identify data that we can... kind of lock down. [18:54:14] ejegg: yeah, sec [18:54:55] CaitlinVirtue: the LYBUNT part of the report, i.e., the years in which people donated, is fixed to fiscal years for the custom version [18:54:57] ejegg: It sounds like the most hammered-on use case is going to use that a lot. :) [18:55:38] Like, constantly. [18:56:06] We're going to have to do something stupid for the year rollover, though. :/ [18:56:14] (CR) Pcoombe: [C: 2] "Looks good, thanks!" [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/179022 (owner: Awight) [18:56:23] (Merged) jenkins-bot: other TY updates [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/179022 (owner: Awight) [18:56:28] @ejegg if I ran the report from the template, I could use the calendar year feature though, right? [18:57:19] CaitlinVirtue: only the built in LYBUNT, not the WMF LYBUNT. The WMF LYBUNT uses the a new table which has the donor years pre-calculated. [18:58:05] CaitlinVirtue: unfortunately we didn't take the calendar/fiscal selector out of the UI yet [18:58:19] hmmm, ok. That's fine for the immediate future (dec 2014) [18:59:31] @ejegg cy lybunt is important to me too, bc most donors don't care about our fiscal year. the calendar year is what impacts their taxes [18:59:58] ejegg: see lutetium:/tmp/big_ugly_query [19:00:07] Jeff_Green: thanks! [19:00:11] np [19:00:55] @ejegg would this also be the time to mention that I think something is wrong with past years custom lybunts? [19:01:05] CaitlinVirtue: OK, that will take a different approach, or a whole lot more custom fields. Guess we'll worry about that later [19:01:16] @ejegg yes, later is fine [19:01:18] CaitlinVirtue: what seems to be wrong with past year ones? [19:01:36] the numbers for cy and fy are the same for the 2013 report: 526 [19:02:08] if i'm interpreting that correctly, it means that 0 prevous donors gave again in the 2nd half of the fiscal year 2013-2014 [19:02:17] which seems highly unlikely [19:02:31] but maybe that's bc of the cy issue you just mentioned? [19:02:34] if it's the custom lybunt, it just means we're misleading you by letting you choose calendar year [19:02:40] ah [19:02:45] yeah, it's still going by fy [19:03:05] ok, i'll re-run and sort the data from the csv to get an accurate figure then [19:04:17] Jeff_Green: well, there's only one join on civicrm_contribution [19:04:44] or, wait, that's not a lybunt report [19:05:01] that's just a search over contributions [19:05:05] i don't know what that is, it's just the query that was running [19:05:19] That could certainly get in the way of running a lybunt though [19:05:37] @K4-713 @ejegg can I run additional instances of lybunt right now? the other one is still just spinning and going on 60 minutes now [19:06:18] ejegg: seems bad that a simple search of contributions copies the entire 4GB table to a temp table [19:06:27] CaitlinVirtue: sounds like the old one died in the db somehow, as the big query Jeff_Green saw running was just a searchh [19:06:32] Jeff_Green: yup [19:06:58] @ejegg @Jeff_Green so that's a yes then? [19:07:01] if you want to fire off the lybunt query I can fetch [19:07:33] CaitlinVirtue: it's more of "an error has occured" [19:07:37] :-) [19:08:20] @Jeff_Green sorry no comprende. I should kill the exising query and can try running something else? [19:08:31] there's no query running at the moment [19:08:40] not from the db's perspective, anyway [19:09:03] @Jeff_Green ok thanks [19:09:07] if you can rerun it, I'll take another look on the db while it's going [19:09:55] @Jeff_Green ok, running from the template now [19:10:06] started at 11:10 am [19:11:12] it's doing a lot of smaller queries [19:11:43] the one we want should have some joins to wmf_donor [19:14:31] ccogdill: hey, I wanted to check in about https://wikimedia.mingle.thoughtworks.com/projects/online_fundraiser/cards/2103 [19:14:54] uhh okay [19:14:55] After a false start... I realized that the error is indeed in the query, or at least with the interpretation of the query results. [19:15:03] What r u doing with the output? [19:15:28] awight lives! [19:15:33] I found that there is a "risk_score" column in both tables, and the payments_fraud_breakdown.risk_score is the one u were looking for [19:15:36] atgo: hi! [19:15:36] well, I wanted to use it to be able to identify why a donor was getting caught in the fraud filter [19:15:37] you doused? [19:15:43] not undead yet [19:15:49] ccogdill: yes [19:15:51] if it was a high-ish value donation, it would be worth my time to check [19:15:58] ccogdill: you OoO btw? [19:16:04] yep [19:16:17] ccogdill: ok. well, run that query again and we can go over the results... [19:16:35] I think you'll see the issue much faster than I did, though. [19:16:45] ejegg: anything I can do to help with the LYBUNT? [19:17:10]