[00:00:04] :P [00:00:22] actually, i just did that, yea [00:00:28] spagewmf: oh cool! re hatjitsu [00:00:31] that told me it's a way to tell how hard it is [00:00:54] so it's a number , right [00:01:06] right, and it has been refined over the years to be more useful than easy/medium/hard [00:01:27] in that context "each task is a story point" means it has a value of 1? [00:01:35] and is added to the global score of that story? [00:01:53] mutante: since we can't edit story points in phab right now, I guess the default is 1 [00:02:44] but, each task would have a score of 1, 2, 3, 5, 8, 13, 21 (fibonacci) [00:02:56] higher number means takes longer [00:03:18] over time your team starts to figure out what those numbers mean for them (no two teams are necessarily the same) [00:03:19] if i imagine having to do that for each ticket in RT.. that would be a ton of guessing [00:03:56] its so you can plan "ok, our team can probably get through these 6 feature requests/bug fixes in the next two weeks" [00:04:00] and almost always wrong in the end (as in , what seems easy turns out to be hard , but nobody expected it) [00:04:10] you get better at it as you do it :) [00:04:30] every team starts out with wild swings in the number of points completed in the first few sprints [00:04:36] i guess as long as the time spent on 'meta' stuff is not getting too long [00:04:45] is that also measured? [00:04:48] well, you dont' have to do every single thing in your backlog [00:05:13] I think any attempt to estimate the difficulty of a software dev/operational task, is almost always doomed from the start. it's very difficult to estimate [00:05:17] mutante: "grooming" is the word for that, and there's two parts to that [00:05:47] 1) after each sprint there is a retrospective, where you/your team say "this worked, this didn't, let's do this instead" and part of that is "boy did we underestimate doing X" [00:05:59] is there one person doing that for all tasks ? [00:06:01] 2) the product owner's job is to groom the backlog, including prioritization [00:06:23] 3) the team, as a team, assigns points to the tasks the product owner wants int he current sprint [00:06:37] (the product owner can't do that) [00:06:46] twentyafterfour, meet https://en.wikipedia.org/wiki/25_or_6_to_4 :) [00:07:44] mutante: also, the backlog should be groomed, but there's no point in making it perfect "all the way down" to the last item if you aren't going to get to that last item any time soon. [00:07:53] mutante: ie: you groom it just enough just in time [00:09:45] greg-g: were you a scrum master, or on an Agile team? [00:10:11] spagewmf: no [00:10:19] I want to say "not yet" but I don't want to scare twentyafterfour ;) [00:11:59] spagewmf: Dan G and I went to a "Certified Scrum Product Owner" training together [00:12:42] Flow team was trained by Tomas and Arthur Richards themselves. [00:12:58] lols scrum this scram that [00:13:19] greg-g: I'm not scared. we did this stuff at deviantART...to an extent anyway [00:14:08] whew [10:04:10] oh, the on-wiki "tracked" template is working, nice: https://www.mediawiki.org/w/index.php?title=Talk:Phabricator&oldid=1186684#Why_Wikitech_accounts.3F [10:04:49] :) [11:52:54] Does anyone here have wikibugs-l list administrative access? I need to subscribe phabricator to the mailing list but I don't know how to do that since I won't receive the subscribe-confirm email [11:56:31] The email should be from noreply@phabricator.wikimedia.org [12:15:15] twentyafterfour: bottom of https://lists.wikimedia.org/mailman/listinfo/wikibugs-l says that ^d and Brion have [12:26:14] Why the heck does https://www.mediawiki.org/wiki/Talk:Phabricator/Help use Flow? Are we a Betafeature? And https://www.mediawiki.org/wiki/Talk:Phabricator does not. And https://www.mediawiki.org/wiki/Talk:Phabricator/FAQ redirects to the latter. [12:42:35] andre__, because I requested a Flow page there. Why not? [12:42:55] Phabricator is for forward-looking people :) [12:43:08] So the Help subpage has Flow? The main page not? [12:43:17] I just don't get all this inconsistency sometimes I guess. [12:43:52] Help discussion page is for asking questions, Phabricator discussion page is for whatever people come up with [12:45:32] inconsistency is the seed of evolution [12:46:14] hehe [12:46:17] well said :) [14:48:33] https://bugzilla.wikimedia.org/show_bug.cgi?id=71270 is something we should probably get chasemp_ to work on :) [14:49:53] or any ops folks [14:50:19] I was meaning as he's the current point person for phab [14:50:28] I'm just asking around about how we assign ipv6 addresses [14:50:33] doing the change shouldn't be too difficult [15:37:32] I need a phabricator.wikimedia.org admin and the link on mw.o won't let me see who I can ask. Anyone here? [15:40:47] What do you need? I think it may only be andre/chase/daniel [15:41:39] oh, and twentyafterfour [15:41:52] hexmode: whatya trying to do? [15:42:13] Reedy: I'm looking for an account like it says to do at [15:42:18] https://www.mediawiki.org/wiki/Phabricator#Access_to_phabricator.wikimedia.org [15:42:26] hexmode, got your LDAP credentials ready? [15:42:26] ah [15:42:28] ^ [15:42:29] :) [15:42:43] andre__: are these the same ones as Gerrit? [15:42:47] hexmode, yes [15:42:49] /wikitech [15:43:01] then yes [15:43:08] hexmode, then please register now and tell me when you're in. And do not upload files. [15:43:17] heh [15:43:31] * andre__ speaks with a serious, deep voice [15:45:01] You really make me want to upload something :P [15:45:19] How dare you?! [15:45:21] verifying email [15:46:15] GO GO GO! :P [15:46:18] andre__: awesome. I'm in [15:46:22] I can't wait for some lurker in here to register while it's open for someone else :) [15:46:32] hexmode, Welcome to the future! [15:46:35] * Reedy blocks greg-g [15:46:47] greg-g, I check the accounts list every 5 minutes. [15:46:59] andre__: when do you sleep?! [15:47:02] it is so glorious! [15:47:02] I've set up a Nagios/Ganglia notification if there's a new user. [15:47:07] andre__: wow [15:47:07] greg-g, sleep is so 80es [15:47:15] andre__: I'm honestly impressed ;) [15:47:28] greg-g, it's easy to impress you by lies. I should remember that. [15:47:40] figured [15:47:43] jerk [15:48:12] This "future" isn't as glorious as I had hoped :( [15:48:26] oh well [15:48:48] actionable critique or gtfo ;) [15:48:57] :P [15:49:03] hexmode broke it already [15:49:11] andre__: no you don't have a notification setup :P [15:49:12] hexmode: how's life? [15:49:15] I was just saying it looks the same... [15:49:23] greg-g: life continues apace [15:49:30] good enough I suppose [15:49:43] That's all I can hope for :) [15:49:43] YuviPanda, wait, how did you get access to my computer? I thought I fixed that Bash issue? :P [15:49:51] andre__: :D [15:50:14] There's another ;D [15:50:27] hmm, it *would* be nice to setup notifications of some form for 'languishing patches' and 'languishing bugs' [15:50:29] or something similar [15:58:04] that's what that one app is for, with those pretty graphs that quim used in their quarterly review [15:58:18] well, to graph it, not alert on it (except manually) [16:03:04] http://korma.wmflabs.org/browser/bugzilla_response_time.html [16:03:11] I promise it will all break with Phab. [16:36:41] hah [18:08:13] andre__: beautiful! [18:18:58] yay :) [21:11:50] could someone open phabricator for palexis? [21:11:58] Reedy: andre__ ^^ [21:12:12] but who is palexis? :) [21:12:21] andre__: Alexis [21:12:25] and where? [21:12:37] hexmode: does she have a wikitech/ldap account? [21:12:42] the woman helping Markus and I with MarkYMarkus [21:12:45] ah! [21:12:48] didn't know, sorry :) [21:12:50] greg-g: just set one up [21:12:53] * greg-g nods [21:12:53] andre__: np [21:13:30] hexmode: has Alexis already sworn to not upload files? [21:13:37] andre__: truth be told, I happen to be married to her, as well [21:13:45] hexmode, uh! I see! :) [21:13:49] * andre__ only knows a male Alexis [21:13:56] andre__: she knows [21:14:12] hexmode, she may hurry up and register. Go go go! :) [21:14:14] now you know a female [21:14:25] Nice! [21:14:28] I've never met a male one. [21:14:36] greg-g, you have not been to France it seems. [21:14:40] nope [21:14:43] :) [21:15:03] hexmode, as usual, tell me when Alexis is in [21:15:44] andre__: done [21:15:53] alright [21:16:09] she chose "Alexis" because of the meaning [21:16:11] * greg-g imagines andre__ pulling up a draw-bridge [21:16:24] didn't realize everyone would harp on her [21:16:30] hexmode, I should have studied either Latin or Greek I guess... [21:16:39] instead I ended up with French as a 3rd language. [21:17:04] Alexis is russian, fwiw [21:17:09] uh? [21:17:23] A lot of russian men with the name [21:17:39] * andre__ googles Алексис [21:18:41] greg-g: that would be a high-speed draw-bridge - if somebody was on it it'd be a trebuchet for them :P [21:18:50] hahaha [21:18:54] (lame deployment system jokes) [21:52:23] andre__: still online? [21:52:55] greg-g: yeah. I told you I don't sleep! [21:53:09] * andre__ has to finish a paper til Tuesday hence seeing his computer screen 24/7 [21:55:27] andre__: oh, sorry! but.... [21:55:35] but, but, but, eh? :P [21:56:05] if you were asked, would you be able to answer how many bug reports are filed... actually, this doesn't make sense the way I'm asking it. I'll tell you what I'm really trying to do. [21:56:30] So, I want to know if the time from Thursday afternoon through to Tuesday Morning is useful in finding issues on the test-wikis and mw.org [21:57:02] in otherwords, if I switch the schedule around to only leave one day between testwikis and non-wikipedias, am I going to potentially make things worse [21:57:12] I thought you might have some thoughts on that [21:57:50] hmm, I think we once discussed this via email already [21:58:06] yeah.... [21:58:11] so you could query how many bugs are filed per day [21:58:30] but I'm really not sure if that is a causal data source. [21:59:07] because it's hard to compare with how many reports you'd receive anyway. I'm afraid we have nothing really to "compare" [21:59:25] yeah [21:59:37] like "this is the basic level of reports that we get anyway by people who only use en.wp or such" plus "this was by people running into issues first on mw.org" or such [21:59:48] and volunteers naturally work more over weekends, potentially skewing the results (more feature requests than actual issues, for instance) [22:00:23] hmm, that could be an interesting data question, but that would require correct data in Bugzilla, e.g. people marking enhancement requests as such [22:00:31] and people suck in this. [22:00:43] yeah [22:00:45] :/ [22:01:12] I don't know specifically about WM BZ, but I have papers stating that 50-60% of users do not find or care to set that "Severity" field to enhancement. [22:01:22] compared to about 10% in JIRA which has a better interface for that. :P [22:01:34] heh [22:01:34] * andre__ tries to look like a scientist [22:01:57] the funny thing when you read scientific papers is how many researchers blindly trust the data and expect it to be set correctly. [22:02:12] really? that's sad [22:02:22] I mean, maaaaybe in well groomed bug databases [22:02:34] but I don't know how many of those exist [22:02:39] I've read several (otherwise good) papers which make that mistake, coming to wrong conclusions about bug management in Bugzilla [22:02:46] * greg-g nods [22:03:09] yeah, how to judge the quality of triaging? :) You're only left with manual assessment looking at them as a researcher. [22:03:20] s/them/tickets/ [22:03:36] I'm very curious how such things might change or not in Phabricator. [22:04:07] it'll be easier to have "the big ole giant mess o' everything" and "a well groomed-ish smaller subset that we're working on" [22:04:12] at least [22:04:45] yeah, hope so [22:04:51] (I'm slightly concerned if upstream development will have an open ear on experience in "established" projects having used bugtrackers for more than 10y and all the noise and problems arising, [22:04:52] and if they are willing to implement things I consider necessary in the vey long run - Phab is a fresh project, no huge half a million entries DBs yet to tackle) [22:10:11] * greg-g nods [22:10:21] well, thanks anywho, andre__ [22:10:26] heh, sure :)