[12:09:04] Team-Practices, Engineering-Community, Human-Resources, Wikimania-Hackathon-2015: Code of conduct/Phabricator etiquette/Friendly Spaces/etc breach consequences and escalation paths - https://phabricator.wikimedia.org/T90908#1070884 (Qgil) [14:54:56] fwiw I'm working from home this morning. will be in SF in the afternoon [14:55:40] also fwiw, i hope to change my irc nick today (when I set up a bouncer) [16:30:17] This is Kevin. Can someone here help me with a quick irc/bouncer test? All you have to do is tell me how my status changes as I leave and come back. [16:30:26] sure thing meeple27 [16:31:56] awjr: thanks. I'll leave, and if you can say something while I'm gone, that would be great. I'll come back in a couple minutes. [16:32:52] something while i'm gone [16:33:12] something else [16:34:19] did it work meeple27? [16:34:30] awjr: hm. that test didn't test one thing I wanted, but what did my status look like? [16:34:51] meeple27: it looked like you signed off/left the room [16:35:23] hm. ok. more testing needed later. that's all for now. thanks. [16:35:30] np :D [16:35:43] meeple27: what are you using for your bouncer? [16:35:54] awjr: znc [16:36:02] nice that is what i use too [16:36:19] currently running on my laptop itself, and once i get it configured properly, i'll host it on a personal server i have [16:37:58] there's been talk on and off over the last few years of someone setting up one common bouncer for anyone on staff to use [16:38:03] like, a wmf-hosted bouncer [16:38:24] which would be *really* nice and help make the process less difficult, particularly for less technically inclined folks [16:44:26] yes, an official bouncer would really help [18:06:05] awjr: the official bouncer thing is a tar pit. I got stuck in it at one point as did ^d and Yuvi. ITOPS said if they made one they would hide it behind the VPN which to me was not very useful either. [18:06:24] It all boils down to credential management [18:06:46] ergh [18:07:02] the bouncer needs to be "you" when it talks to freeenode which means it needs access to your freenode password [18:07:41] since some of us are in "sensitive" channels this causes security concerns [18:08:37] That's why we can host a communal bouncer in labs [18:09:36] ^d and Yuvi were going to get around that by setting something up on Linode and expensing the $20/mo for the VPS [18:10:03] but then they got too fancy for their own good with wanting a web frontend and other magic [18:10:21] and lost interest [18:11:12] theoretically my bouncer is working now, and I'm about to head into the office, so someone please let me know what I looked like when I was gone [18:11:26] was = future tense [18:11:45] heh [18:12:18] * bd808 goes back tot he future to see how things turn out for those crazy kids [19:25:22] well, i'm back, i guess. [19:32:29] https://docs.google.com/a/wikimedia.org/spreadsheets/d/1BXG3W-0FHRRHjICbf1tKmsxaKjssaJuhjCW_r5BXj08/edit#gid=617494673 [19:33:43] https://etherpad.wikimedia.org/p/relengq3health [20:18:43] awjr: when I was away, did I get a cool _away marker, or did I exit and rejoin? [20:19:28] meeple27: neither [20:19:39] it doesn't look like you signed off/on and your nick didng change [20:20:13] last time it looks like you quit was ~1036 and rejoined ~10:48am [20:20:21] interesting. did anything happen between my "was = future tense" and my "well, i'm back"? Nothing got replayed, although it did in a different channel [20:20:55] meeple27: bd808 said 2 things, that's it [20:21:13] argh, so that didn't work. wonder why it worked elsewhere but not here [20:21:50] :( [20:22:16] ok, well i'll watch it next week and twek as needed. thanks. [20:22:20] meeple27: your whois didn't change [20:22:24] tweak, not twerk [20:22:35] Did you tell znc to buffer this channel? [20:23:01] bd808: znc slightly exceeded my knowledge and understanding, so I can't answer that [20:23:07] heh [20:23:12] I didn't specifically tell it to buffer this channel, but i did set up buffering in general [20:23:35] maybe a conflict between auto join and chansaver? [20:24:03] * bd808 goes to find his config [20:24:38] bd808: Maybe we could email each other some config files [20:24:58] meeple27: znc, right? [20:26:10] meeple27: here's my znc conf with password stuff redacted: http://pastie.org/private/hle28rzanbp3qgkxg5nmkg [20:31:17] meeple27: i think most of that was autognerated with only small changes from me [20:32:44] mine looks pretty similar -- https://dpaste.de/hZza [20:32:55] awjr: znc, yes. Thanks. I'll compare yours to mine when I get home tonight. (I don't have the server private key here) [20:32:59] bd808: Thanks too [20:34:35] y'all need moar channels. eg. #wikimedia-kawaii >.> [20:34:58] https://upload.wikimedia.org/wikipedia/commons/6/6d/Wikimedia-kawaii-lolcat.png [20:35:12] lulz [21:01:25] awjr: I'd love some help on https://www.mediawiki.org/wiki/Wikimedia_Engineering/2014-15_Goals#MediaWiki_Core if you've got a few minutes to brainstorm [21:01:48] I'm searching for a key result of making OAuth less sucky [21:02:06] and not coming up with anything "hard" to commit to [21:02:18] sure bd808 - one sec, i just need a quick break after the meeting i just wrapped up, back in a jiff [21:02:24] np [21:08:34] bd808: how do you know OAuth is sucky right now? [21:08:45] Erik is yelling at me :) [21:08:53] has he filled a Task ? :D [21:08:58] oh yes [21:09:12] key result: erik stops yelling at me [21:09:12] The complains are UI/UX mostly [21:09:18] hehe [21:09:27] kristenlans: thanks a lot to have lead the RelEng health check survey. I really appreciate those quaterly team reviews [21:09:50] That's the KPI for everything right? yells/day by Erik [21:10:15] hashar: you're welcome, I enjoyed it! [21:10:28] Chris suggested either more users authorizing per unit time or more active apps [21:10:44] <-- Antoine there [21:11:00] neither are really directly traceable to code improvements, but process improvements would be good too [21:12:28] A really process centric goal would be a mean time to approval of X for a new application request [21:13:03] bd808: i think it's OK if the KR doesn't necessarily directly trace back to code improvements; and i like the notion of making the KR user-centric [21:13:20] bd808: is that a human process? [21:13:29] yeah [21:13:51] and the improvement would come from handing the process off to stewards and teaching them how to do it [21:14:02] i guess it comes down to what problem are you trying to solve - if the problem you're fudnamentally trying to solve is time-to-approval, then i think that makes sense [21:14:39] The asks are a mixed bag but heavily biased to end-user experience [21:15:02] making the screens less confusing for users who hit an OAuth app [21:15:25] so app approvals doesn't align closely with that [21:15:26] same thing applies there though - what's the end goal of making the screens less confusing? [21:15:52] "increasing adoption" I suppose [21:16:15] which ties into Erik's next FY goal of more edits via OAuth tools [21:16:48] well, his next FY is more "things from Labs" but that implies OAuth [21:17:37] so it sounds like mostly what you're trying to do overall is increase adoption (presumably measurable by users authorizing?) [21:18:16] Yeah. I think that sounds right. [21:18:57] so the objective might be something like 'increase adoption of oauth', key result '% increase in authorization requests' or whatever? [21:18:58] authorizations per unit time is a crappy metric though [21:19:07] wht about total auths? [21:19:17] because one task is to decrease the need to auth as often [21:19:22] well, i guess you still need to time unit... [21:19:25] ah [21:19:34] then how about auths by unique users/apps? [21:19:45] yeah that might be a good combo [21:19:56] or # of users/apps making auths [21:20:47] I think this gets me on a good track. I need to poke at that database to see what was can know about this [21:21:12] cool man [21:21:15] :D [21:24:33] bd808: could you say something like "top 5 usability problems will be fixed", and then as the product owner you would choose and document those top 5 in phab? [21:25:20] soo gameable :) [21:26:23] I hate this kind of planning by the way. I want to be on the downstream side of this working on the tactical plan to meet the stratedgy [21:27:23] The OKR page says "An objectively measurable metric" so I'm guessing top N problems is not objectively measurable. [21:27:53] since we don't have voting for issues selecting the top N is subjective [21:46:54] truly though, everything is subjective [22:28:34] bd808: With a true product owner, I would trust them to be able to pick the top 5, although there would be some question of how big/small each one was [22:28:49] like "add an exclamation point" vs. "improve speed by 100x" [22:29:36] the PO is entrusted with funneling all the competing voices into a single voice, however they choose to do that