[15:40:47] * Nettrom goes to chase down the ICWSM deadline
[17:10:37] <jonas_agx>	 Helder, hey -- I won't be able to attend our meeting today. This [Brazilian] afternoon I'll work somethings out, will you stay here around 17h [UTC-03]?
[17:10:57] <Helder>	 I'll probably be here jonas_agx 
[17:11:04] <jonas_agx>	 Great
[18:03:34] <halfak>	 Hey Ironholds 
[18:03:36] <halfak>	 We're moving to a new room. 
[18:03:39] <halfak>	 Got kicked out by events. 
[18:03:40] <halfak>	 We'll be on the call soon.  
[18:06:04] <guillom>	 Does anyone know Ed Chi or people at PARC personally? Apparently their WikiDashboard tool ( https://blogs.parc.com/blog/2010/09/revamping-wikidashboard-3/ ) relied on the (late) Toolserver, and since we moved to Labs the dashboard has been broken.
[18:21:28] <Ironholds>	 halfak, poke me when you have a free 5 mins for revert talkage :)
[18:25:38] <halfak>	 Hey Ironholds.  
[18:25:42] <halfak>	 Just getting plugged in. 
[18:25:48] <Ironholds>	 cool!
[18:25:59] <halfak>	 In the meantime, see https://meta.wikimedia.org/wiki/Research:Revert
[18:26:59] <Ironholds>	 ta
[18:27:29] <halfak>	 So, in practice, I use a 15 revision "window" 
[18:27:44] <halfak>	 And when I'm looking for "damage" I use a 48 hour "sunset"
[18:28:02] <halfak>	 "window" is the maximum distance in revisions that a revert can cross. 
[18:28:25] <halfak>	 "sunset" is the time since the "reverted" revision that the "reverting" revision can take place. 
[18:29:37] <halfak>	 ... -[]-[]-[]-[reverted_to]-[reverted]-[reverted]-[reverting]-[]- ...
[18:29:41] <Ironholds>	 hmn
[18:30:24] <Ironholds>	 so the logic would look like: when there's a match, if match2 - match1 <= 15, for each intervening revision, (if it's <=48 hours after match1), is_revert = TRUE?
[18:31:34] <halfak>	 revdiff(reverting, reverted_to) <= 15 -> "revert"
[18:31:55] <Ironholds>	 gotcha
[18:32:02] <halfak>	 timediff(reverted, reverting) <= 48hours -> "possibly a revert for damage)
[18:32:10] <Ironholds>	 aha!
[18:32:13] <Ironholds>	 so they're distinct tests?
[18:32:19] <Ironholds>	 not nested tests?
[18:32:22] <halfak>	 Well, it's rather, one is a subset of the other. 
[18:32:29] <Ironholds>	 oh, gotcha
[18:33:27] <Ironholds>	 so a reverted edit is any edit in a sequence between two matching edits, where the matching edits are within 48 hours of each other?
[18:33:38] <Ironholds>	 and the matching edits are within 15 revisions of each other?
[18:34:34] <halfak>	 Hmm... Still not quite.  is_revert only considers the revdiff between the matching sha1s
[18:35:12] <Ironholds>	 gotcha
[18:35:16] <halfak>	 is_revert_for_vandalism includes the timediff between reverted and reverting edits (only checked in enwiki)
[18:35:21] <Ironholds>	 aha!
[18:35:23] <halfak>	 And only on namespace )
[18:35:31] <halfak>	 * namespace == 0 
[18:35:33] <Ironholds>	 okay, so the subsetting is not a subsetting of tests for a single "reverted" metric
[18:35:46] <halfak>	 +1 
[18:35:50] <Ironholds>	 it's: you get reverts, and AV reverts. One has a nested test.
[18:36:05] <halfak>	 yes. 
[18:36:23] <Ironholds>	 reverts are edits between two matching sha1s with a 15-edit span: AV reverts additionally have to have those matching sha1s within 48 hours, in line with your work with RStu
[18:36:30] <Ironholds>	 *thumbs up*
[18:36:34] <halfak>	 +1
[18:36:35] <Ironholds>	 got it. Implementing should be a doddle.
[18:36:44] * halfak googles doddle
[18:37:25] <halfak>	 Cool! :) 
[18:39:57] * AndyRussG waves :)
[18:40:28] <AndyRussG>	 halfak: this morning I finally got around to watching the video of your October talk on the techno-social, great stuff!
[18:40:53] <AndyRussG>	 Ironholds: did you come in for all-staff, or have I just not seen you yet?
[18:41:19] <Ironholds>	 that's not a mutually exclusive pair of questions
[18:41:28] <halfak>	 AndyRussG, Would love to chat about it more if you are interested.  :) 
[18:41:41] <AndyRussG>	 halfak: me to! Was just about to say the same
[18:42:00] <AndyRussG>	 Ironholds: it's a trick question, but the real question is, are you around?
[18:42:11] <Ironholds>	 nope.
[18:42:23] <AndyRussG>	 Ironholds: awwww
[18:43:18] <AndyRussG>	 Ironholds: I guess you're not coming for tech days either? If not on this trip, hope to see u around another time, then... ;)
[18:43:42] <halfak>	 AndyRussG, any plans for lunch yet?
[18:44:14] <Ironholds>	 AndyRussG, nope. I'm not visiting the office for the all-staff.
[18:44:45] <AndyRussG>	 no plans yet! :) Only FR-tech standup at 1 pm
[18:45:16] <halfak>	 Bah.  That's my opening.  I'll be around this weekend too.  I'll probably be working from the office. 
[18:45:23] <AndyRussG>	 Standup is only 10 minutes
[18:45:35] <halfak>	 Oh cool. So, 1:10 
[18:45:37] <halfak>	 :) 
[18:45:44] <AndyRussG>	 Yeah! that'd be great
[18:46:01] <halfak>	 Cool.  Looking forward to it. 
[18:46:06] <AndyRussG>	 Yeah! me too
[18:46:31] <AndyRussG>	 Let's say 1:15? Standup officially starts at 1:03 pm (don't ask)
[18:47:55] <AndyRussG>	 I'm at a desk right behind the big long table on the North side of the 3rd floor
[18:49:12] <AndyRussG>	 Ironholds: oh well...
[18:50:46] <halfak>	 AndyRussG, sounds good. I think I'm directly behind you. 
[18:51:04] <AndyRussG>	 halfak: ah OK I see you now :)
[19:06:04] <Ironholds>	 so, meeting? ;p
[19:09:26] <Ironholds>	 leila, ?
[19:09:50] <leila>	 geez, Ironholds. sorry. lemme tell them
[19:09:57] <leila>	 they're standing in the hall-way talking
[19:09:59] <leila>	 brb
[19:11:10] <leila>	 Ironholds, moving to collab space. sorry. :-(
[20:30:05] <Helder>	 halfak, do you know if your presentation from last week is recorded somewhere on youtube?
[20:31:11] <halfak>	 Yes.  Sec. 
[20:31:17] <Helder>	 halfak, I found it! 
[20:31:20] <Helder>	 never mind
[20:31:21] <Helder>	 https://www.youtube.com/watch?v=53bG9mYMYE8#t=11m20s
[20:31:32] <Helder>	 (it was in your report =/ )
[20:44:56] <halfak>	 :P