[15:55:12] (CR) Ejegg: "Hi Mhutti1," [extensions/DonationInterface] - https://gerrit.wikimedia.org/r/263438 (owner: Mhutti1) [16:20:23] Fundraising-Backlog: Estimate the proportion of people who send us complaints vs are affected by an issue - https://phabricator.wikimedia.org/T120722#1986515 (MBeat33) For starters here are two batches of GC100 Auth Fail errors (usually a bank doing maintenance causes this). These errors occur after donors s... [16:50:34] Fundraising Sprint Bloodletting 2016, Fundraising-Backlog, MediaWiki-extensions-DonationInterface, Unplanned-Sprint-Work, and 2 others: CI: test DonationInterface against fundraising/REL1_25 - https://phabricator.wikimedia.org/T124105#1986619 (cwdent) @hashar I just pushed up my first guess at wha... [16:54:08] (CR) Ejegg: "It would be great if we could apply the 'chargeback' and 'refund' financial_type entries to the negative transaction, but I guess that dep" (3 comments) [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/261757 (owner: Eileen) [17:00:07] i can't believe phab doesn't have a setting that is just "email me when something relevant to me happens" -- e.g. i am mentioned by @name, i am assigned to/removed from a task, or there is activity on a task i am assigned to [17:00:12] those are the only times i want email [17:05:54] Fundraising Sprint Asbestos Removal 2016, Fundraising Sprint Bloodletting 2016, Fundraising-Backlog, FR-Ingenico: Refund not matching original amount--GC original may have been imported incorrectly - https://phabricator.wikimedia.org/T123404#1986678 (awight) a:awight>None [17:06:34] Fundraising Sprint Asbestos Removal 2016, Fundraising Sprint Bloodletting 2016, Fundraising Sprint X-Ray Spex, Fundraising Sprint Yo La Tengo, and 4 others: Do not show donation form error message: "No processors available". Fix UI and plug holes. - https://phabricator.wikimedia.org/T117872#1986682 (... [18:26:13] awight, ejegg morning! GC dropped those missing audit files off, i see them and they look to contain the info. i was thinking of punching the job, any objections? [18:26:18] (PS1) Ejegg: Ensure 'Gateway' option value exists [wikimedia/fundraising/crm] - https://gerrit.wikimedia.org/r/267707 [18:26:21] hi cwd! [18:26:29] sounds fine to me [18:26:48] groovy, here goes [18:28:23] yay! [18:30:10] ASUS haters will enjoy this anecdote. I attempted to back up my laptop this weekend, which failed due to stupid. But the strain of actually reading the entire SSD contents was enuf to cause complete SATA controller meltdown, looks like that laptop is not coming back to life. [18:30:49] Wow, that's pretty bad. [18:31:52] yowza! [18:32:15] oh man, speaking of software operations brickin: https://github.com/systemd/systemd/issues/2402 [18:32:34] POETTERING!!!! [18:32:49] * cwd shakes fist [18:36:33] hahaha. I was just scratching my head, as installing npm caused rEFInd to be installed [18:36:42] that can only be bad [18:36:46] whoa [18:37:13] javascript that writes to your hardware? [18:37:57] and of course, I don't see any dependencies pointing to it in the deb database [18:38:30] Setting up refind (0.10.0-1) ... [18:38:30] efibootmgr: EFI variables are not supported on this system. [18:38:30] ShimSource is none [18:38:30] Installing rEFInd on Linux.... [18:38:30] The ESP doesn't seem to be mounted! Trying to find it.... [18:38:32] ///boot doesn't seem to be on a VFAT filesystem. The ESP must be [18:38:35] mounted at //boot or //boot/efi and it must be VFAT! Aborting! [18:38:37] Sooo scary [18:39:10] i'm so glad we decided to ditch bios boot [18:39:42] Malware is way more flexible now [18:43:48] wmf should buy linux dorks the xps13. it's a better computer in every way, and half the price [18:44:11] the webcam is stupid, but if everyone had it no one would notice [18:44:26] also it motivated me to cover it with a sticker and carry a usb one [18:45:23] which i am glad i did in general. there is clearly a browser api webcams. zero day would be a terrible day. i like keeping it physically disconnected when not in use. [18:46:33] *for webcams [18:47:50] yeah, I'm not sure how much standardized equipment actually benefits OIT--it's not like they're changing inner tubes. [19:23:24] awight: the more I look at the new financial stuff, the more confused I get. I [19:23:41] seeriously [19:23:42] think financial_type may not actually be the right place for 'refund' and 'chargeback' [19:23:59] it's only used in the line_item table [19:24:00] eileen had an opinion on that... [19:24:26] ahh, someplace in this channel's logs? [19:29:28] Are you CR'ing the refund patch? I think that captures her recent thingking [19:29:54] Yeah, that's what i'm diving into this stuff for [19:30:33] https://gerrit.wikimedia.org/r/#/c/261757/ [19:30:34] k [19:30:42] so, this doesn't seem to be happening quite: https://wiki.civicrm.org/confluence/display/CRM/CiviAccounts+Data+Flow#CiviAccountsDataFlow-RefundORCancelaCompletedcontribution(changestatustoRefundedorCancelled) [19:30:58] I'm not getting any negative entries [19:31:18] though that may be a change between 4.6 and 4.7 [19:31:33] hrm I was noticing that I would only get one financial trxn entry using the stock Civi refund functionality [19:31:36] Have you compared that? [19:31:43] u just edit a contribution and set the status to Refunded. [19:31:51] yeah, I'm just looking at the stock functionality now [19:32:02] and it doesn't create any negative rows [19:32:05] gasp [19:32:10] it worked for me [19:32:37] hrm, lemme see... [19:37:12] awight: ooh, wrong 'Refunded'. Gotta figure out why that status fix update isn't running [19:38:02] if only drush cooperated with xdebug... seems to just hang forever [19:39:35] lol i'm an idiot. combing through audit files for an hour. forgot to run the download job. [19:39:43] hehe [19:40:00] strings turned up in the old, wrong file that's always been there [19:40:17] but my memory had gotten stale [19:46:08] careful overwriting stale memory with new microcode [19:46:26] I know a kid who got stuck like that. [19:47:10] no new files [19:47:13] * cwd long sigh [19:47:18] i gotta go shovel [21:09:10] ejegg: I just logged this https://issues.civicrm.org/jira/browse/CRM-17951 [21:14:20] heh, I was in the middle of creating something very similar! [21:19:48] Fundraising Sprint William Shatner, Fundraising Sprint X-Ray Spex, Fundraising Sprint Yo La Tengo, Fundraising Sprint Zapp, and 3 others: [EPIC] Build Frig: tools to help update our deployment branches - https://phabricator.wikimedia.org/T116820#1987667 (cwdent) [21:19:50] Fundraising Sprint Asbestos Removal 2016, Fundraising Sprint Bloodletting 2016, Fundraising Sprint William Shatner, Fundraising Sprint X-Ray Spex, and 5 others: Automate deploy branch/submodule pointer updates - https://phabricator.wikimedia.org/T123838#1987665 (cwdent) Open>Resolved I rolled... [21:30:55] call dropped but I guess we were at the end [21:32:41] wow - that was 2 factor authentication just kicked me off - it's a month since I entered a code [21:34:36] eileen1: so, I only see financial_type_id in line_item, and Civi doesn't seem to create a second line_item for the refund [21:34:39] that's actually kind of impressive [21:34:43] (2fa) [21:34:51] Fundraising Sprint Bloodletting 2016, Wikimedia-Fundraising-CiviCRM: Please refresh staging DB - https://phabricator.wikimedia.org/T125445#1987733 (Eileenmcnaughton) [21:34:53] Fundraising Sprint Asbestos Removal 2016, Fundraising Sprint Bloodletting 2016, Fundraising Sprint X-Ray Spex, Fundraising Sprint Yo La Tengo, and 4 others: [Epic] CiviCRM upgrade: Adapt refund processing & reporting to reflect changes since the upgra... - https://phabricator.wikimedia.org/T116317#1987732 [21:35:18] ejegg: it creates a row in civicrm_financial_trxn for the refund [21:35:46] (line items are like the line items on an invoice, civicrm_financial_trxn is where it stores payments against that invoice … more or less) [21:36:34] huh, so a refund doesn't get a new line item [21:37:26] so we'd have to indirectly distinguish refunds from chargebacks by which account they were associated with? [21:39:43] ejegg: yeah ! [21:40:34] We could also use/abuse the column civicrm_financial_trxn.trxn_result [21:44:17] I'm not sure there would be a downside - that column is basically for storing information from the payment processor [21:44:17] ah, I think there's something else we're now storing in wmf_contribution_extra that would make more sense in trxn_result [21:44:17] gateway_status_raw [21:44:18] yeah - sounds about right [21:44:23] ejegg: the Financial Accounts are actually lablelled 'Refund' 'Chargeback' [21:44:28] ah, cool [21:44:52] yeah - so we could just use that rather than go through the hoops [21:45:22] (I find the financial account vs type confusing - but mostly I find I can ignore it :-) [21:46:01] OK, so do we need to have all that ironed out before we deploy and run updatedb? [21:50:41] Fundraising Sprint Bloodletting 2016, Fundraising-Backlog, FR-Adyen, FR-Smashpig: Process refunds in Adyen audit file - https://phabricator.wikimedia.org/T121129#1987794 (Ejegg) a:Ejegg [21:51:19] ejegg: well I think the stuff that is tidying up historical entries is probably OK - but it's held up on the staging refresh anyway [21:52:10] I think we are back to trying to record financial_account_id to be Refunded or Chargeback as we see fit? Historically I think that's OK because the parent contribution was showing that so the fixes still stand [21:55:08] do we need to stop creating the old-style refunds when we run the update script? [21:55:35] yeah that's what I keep trying to get my head around! [21:56:18] this ticket here was what was needed to update ones done through mark_refund https://gerrit.wikimedia.org/r/#/c/266469/ [21:57:12] and actually what you can see is that some extra financial_trxn lines were being created - so I guess we do need to fix the process before running that [21:57:20] ah, right, locally I wasn't getting any negative trxns till I ran that [21:57:34] hmm [21:57:43] oh yeah - that's right [21:57:50] because the problem is the 2 * REfund status [21:58:07] yah, that messed with the refund logic in a few places [21:58:25] so perhaps if we run that & get rid of the other then we are getting new transactions that more or less reflect how we want the entries to look [21:59:06] when done via the ui, yep! [22:00:03] hmm perhaps we should get refund working without separate handling for chargebacks & then follow up with that - I can see only 500 over 2 years [22:00:25] nah [22:00:26] oh yeah? Well, I guess it's not a big data loss issue then [22:00:53] yeah - hmm - well I'll work on getting the chargebacks working today - since waiting on Jeff anyway [22:01:45] so, if the refund type is indicated by the to_financial_account, would that throw off the totals for the main account? [22:02:26] or... it would be like a fee, right? [22:03:14] which is recorded as a positive trxn from the main account to a fee account [22:04:12] ejegg: if you look at https://issues.civicrm.org/jira/browse/CRM-17890 there is a table of how a fee looks [22:05:20] it's like a transfer from the main account (which would be 'Donation') to the other account - in our case 'ChargeBack' or 'Refunded' [22:05:34] it's arguable whether we want the transfer to refunded or not [22:06:18] ok, so we would record it as a positive total_amount, from Donation to Refunded [22:06:54] yeah - that sounds right [22:07:02] got it [22:07:32] I thought I'd found something but… https://wiki.civicrm.org/confluence/display/CRM/Admin+refunds+and+chargebacks [22:08:24] ah yeah, i came across that too, plus the patch he submitted [22:08:54] ejegg: so, if we add another financial_trxn to move the data to a chargeback/ refund account [22:09:03] the issue is how we report on them [22:09:26] one option would be to add a field to the contribution - has chargeback - for really easy searching [22:09:40] another is just to make sure there is a report that will find them [22:10:06] well, we'd still put the original into 'refunded' status, right? [22:10:40] and we'd deal with the refund transfer the same way we deal with the fee, I guess [22:11:34] the contribution would have a refunded status [22:12:02] and linked to it, through entity_financia_trxn->financial_trxn would be a row moving $$ to chargeback [22:12:22] right [22:13:13] it might be a bit hard to get to from the UI - but we could a) add to advanced search the ability to search on financial type of transaction (upstream that) or b) add a custom field on the contribution to denote it or c) ask people to rely on a report that we think exists ... [22:15:16] just searching on the status = refunded wouldn't be enough? [22:18:42] oh huh, we already added a 'refund' account, but it's listed as 'Revenue', not 'Expenses' [22:19:31] ejegg: yep! [22:19:46] the status refunded won't give us chargebacks [22:20:27] oh hang on - the financial type of the main contribution is a bit of a 'have' [22:20:57] because a contribution can have multiple line items & multiple payments which can have multiple types [22:26:07] hrm, not sure I understand. [22:29:43] still seems like putting 'refund' and 'chargeback' in the financial_type table isn't quite aligned with core, if refunds don't get their own line items [22:33:10] ejegg: normally people wouldn't put refund & chargeback as types [22:33:16] financial types [22:34:24] arguably you might edit the Expense financial account associated with Donation to be 'Refunded' Financial Account [22:34:28] yah, just trying to figure out how we make it work the same for us whether we refund via UI or via back end [22:35:00] but really the model probably should have a Financial Account for refunds & a financial account for chargebacks for each financial type [22:35:10] do we do chargebacks from the front end? [22:36:03] I dunno - maybe for gateways where the audit parser can't handle them? [22:38:32] I wonder if we should get some advice from our actual finance people [22:38:50] ejegg: So, I've actually added a bunch of other things to the refund form - date, refund_trxn_id. I feel like we are saying we want to also be able to specify the account we are refunding out of [22:39:19] there is a guy who I assigned that ticket to (Joe Murray) who will need to approve any handling that we upstream [22:39:38] I might go on skype & see if I can track him down :-) [22:39:41] hmm, account we're refunding out of, or into? [22:40:12] well from a user perspective it's out of [22:40:34] Refund against account x [22:40:45] that wouldn't always be the account the original trxn went into? [22:44:27] hmm, maybe we need new options for 'financial account relationship' [22:45:24] so expense account is 'Banking fees', refund account is 'Donation Refunds' and chargeback account is 'Donation Chargebacks' [22:46:05] that seems 'most correct' [23:05:10] ejegg: I'm going to look more into the payment api (which is added to 4.7 but still not quite complete). I guess getting it right & tested upstream & porting that will pay off in the long run [23:10:34] right on [23:13:47] Fundraising Sprint Bloodletting 2016, Fundraising-Backlog, FR-Adyen, FR-Smashpig: Process refunds in Adyen audit file - https://phabricator.wikimedia.org/T121129#1988076 (Ejegg) Looks like we need to subscribe to a second report to see chargebacks. There's no type for them on the Payments Accountin... [23:26:15] (CR) Ejegg: [C: -1] "Oops, the change to ProcessCaptureRequestJob::factory breaks the audit processor." [wikimedia/fundraising/SmashPig] - https://gerrit.wikimedia.org/r/266434 (https://phabricator.wikimedia.org/T122244) (owner: Ejegg) [23:37:17] Fundraising Sprint Bloodletting 2016, Wikimedia-Fundraising-CiviCRM: Please refresh staging DB - https://phabricator.wikimedia.org/T125445#1988130 (Jgreen) I just finished dumping the database onto lutetium, will load it up tomorrow! jg [23:40:53] Fundraising-Analysis, Research management, Research-and-Data-Archive: Big English data analysis support - https://phabricator.wikimedia.org/T118054#1988144 (Deskana) @dartar I was going to close this as resolved, but it wasn't clear to me from looking at the #research-and-data-archive board whether or... [23:56:18] Fundraising-Backlog: 'continue' button on the iDeal donation form not working - https://phabricator.wikimedia.org/T125470#1988189 (MBeat33) NEW [23:57:17] Fundraising-Backlog, Wikimedia-Fundraising: 'continue' button on the iDeal donation form not working - https://phabricator.wikimedia.org/T125470#1988201 (atgo) Thanks @mbeat33. We'll get this one in the hopper in one of the next couple of sprints. [23:58:11] ty atgo [23:58:13] Fundraising-Backlog, Wikimedia-Fundraising: 'continue' button on the iDeal donation form not working - https://phabricator.wikimedia.org/T125470#1988189 (atgo) Adding @jrobell, @ccogdill_wmf and @pcoombe for awareness