[07:00:17] yo dsaez [07:45:53] good morning leila [07:46:25] helloooo dsaez [07:46:55] how are you? [07:47:17] dsaez: I have a favor to ask. Can you schedule a 15-min call between Dario, you and I for discussing historical link graph? And, also move the collaboration grooming to a date in the future when we all can make it? [07:47:27] dsaez: I can't see anyone's calendar. the magics of the free world. ;) [07:48:11] dsaez: doing fine. [07:49:40] leila, sure. But the link graph, it would ve better tomorrow, we have this evening full of meetings [07:49:58] dsaez: actually, try to put the link graph on Monday if possible. [07:51:49] ok [07:55:09] :D [08:01:38] thanks dsaez, [08:11:42] dsaez: do we need the section recommendation today? I'm guessing not? [08:13:16] leila, hmm...as you wish...we could discuss alternatives for labeling/modifying the experiments [08:13:47] dsaez: sure. it's good to have everyone on the same page. agreed. [08:48:53] dsaez: re the section rec meeting, Bob is saying he's in a conference but he can move things around to attend. Should we ask him to move things around? I think he can safely skip, no? [08:49:14] sure [08:49:19] no rushes [08:49:31] press the skip button [08:49:43] let's cancel. [11:43:12] * leila presses cancel [12:15:14] nuria_: hola, sorry I missed your ping [14:03:09] dsaez: ping when you read Pau's email. I'd like to brainstorm with you. [15:56:12] bmansurov: I assume you're on client-side performance discussion re citation usage data, right? [15:56:58] leila: yes, some performance concerns were raised. I'll address them and we'll continue discussion during code review. [15:57:14] bmansurov: great. [18:11:49] Hi bmansurov [18:11:58] bmansurov: Saw your mjessage yesterday ab [18:12:06] about wikidata on hadoop [18:12:15] I've playing with that a bit [18:12:27] o/ joal [18:13:08] bmansurov: I've mostly been willing using teh graph, but the data-loading from dumps step exists [18:13:11] bmansurov: https://wikitech.wikimedia.org/wiki/User:Joal/Wikidata_Graph [18:13:58] joal: awesome, I'll have a look. [18:14:06] bmansurov: The steps are - Put a wikidata json dump onto hdfs, refine it into parquet using my refinery-source patch, then do whatever you like with the parquet data :) [18:14:27] I see [18:14:49] bmansurov: some data already exists in parquet format: hdfs://user/joal/wikidata [18:15:23] not up-to date, but at least predent :) [18:15:27] joal: ok, any plans on making these steps automatic for regular use? [18:15:32] bmansurov: not as of now [18:15:50] ok [18:16:14] bmansurov: I think we're not far from being at a stage where it'll be possible, but it's not yet there [18:16:49] bmansurov: I'll obviously keep your team posted, you'll be one of our main users for those datasets (historical and other dumps_) [18:17:19] joal: thanks, and yes, we'd love to use that data for improving our algoirthms [18:18:03] bmansurov: I do know that - I'm really willing to help - I promise it'll be there, but I need you to be patient :S :) [18:19:55] joal: sure, no rush on our end just yet. [18:20:11] :) [21:29:11] This is your friendly warning to update the research group agenda! [21:29:25] Add your stuff [21:29:29] Note your progress? [21:29:35] ASK SOME QUESTIONS! [21:29:43] Or just show off something cool you found. :D [23:47:08] 10Quarry, 10Epic: [Epic] Make Quarry a robust data exploration tool - https://phabricator.wikimedia.org/T194330#4195825 (10Harej)