[00:03:06] <Ironholds>	 yup! although you might also poke Roan?
[00:03:13] <Ironholds>	 he's been messing around in core for donkey's years.
[00:03:52] <leila>	 ok, cool! thanks!
[00:03:56] <legoktm>	 werdna is the one who did the user_properties table refactoring and stuff
[00:04:15] <legoktm>	 it used to just be a blob in the user table :P
[00:04:42] * Ironholds shudders
[00:04:44] <Ironholds>	 ah, the blob days
[00:04:56] <Ironholds>	 when user_rights rampaged across the landscape, sowing devastation in its wake
[03:00:45] <halfak>	 Hey Ironholds.  Still around?
[03:43:14] <halfak>	 Ironholds, app page views fit beautifully. 
[03:43:19] <halfak>	 :) 
[03:58:40] <Ironholds>	 halfak, I'm back
[03:58:41] <Ironholds>	 yay!
[03:58:48] <Ironholds>	 check mobile web? That's de-searched and de-spidered.
[16:03:27] <Ironholds>	 morning halfak :)
[16:03:38] <Nettrom>	 o/ everyone
[16:06:35] <halfak>	 Hey Ironholds 
[16:06:38] <halfak>	 Yo Nettrom 
[16:06:56] <halfak>	 I'm in your lab, sciencing your phenomena.
[16:07:00] <Ironholds>	 nice!
[16:07:06] <Ironholds>	 any luck sciencing the mobile web data? ;)
[16:07:35] <halfak>	 Yes.  Still a huge peak at 18 min. 
[16:07:46] <halfak>	 But it fits a bit better.  I'll demo during our meeting today. 
[16:07:49] <halfak>	 *present 
[16:08:42] <Ironholds>	 cool!
[16:09:00] <Ironholds>	 then I shall launch a query retrieving uri_path so we can see wtf the peak is ;p
[16:09:15] <Ironholds>	 did we ever work out if the 18m events are from a particular group of people?
[16:09:43] <halfak>	 I haven't had time yet.  I'm still just catching up on new datasets. 
[16:10:09] <halfak>	 We have AOL, Wikipedia, League of Legends, Stack overflow and Movielens. :) 
[16:10:17] <Ironholds>	 wooo!
[16:11:08] <Ironholds>	 SO? I can write a paper on how long it takes Brian Ripley to reply to R questions with dickishness
[16:13:28] <halfak>	 Ironholds, you changed timestamp formats on me. :P 
[16:13:54] <halfak>	 "%Y-%m-%d %H:%M:%S" --> "%Y-%m-%dT%H:%M:%S"
[16:14:06] <halfak>	 In a way, it is helping me keep track of which datasets have been updated. 
[16:16:18] <halfak>	 brb reboot
[16:23:00] <Ironholds>	 did I? Dammit, I thought I'd sanitised them.That's the requestlog format.
[18:42:40] <Ironholds>	 halfak, your wifi is bad indeed ;p
[18:45:32] <Ironholds>	 okay, I have another meeting now. Womp womp :(
[18:47:50] <halfak>	 Back in call
[18:47:54] <halfak>	 OK. 
[18:47:58] <halfak>	 Will post summafry afterwards
[20:47:07] <Ironholds>	 halfak, you're back!
[20:47:15] <Ironholds>	 you want the good news or the really good news?
[20:47:15] <halfak>	 Yeah.  Bad WIFI
[20:47:29] <Ironholds>	 the good news: I validated my worries about hashing!
[20:47:45] <Ironholds>	 the really good news: you should recheck mobile_web_events.tsv and see if you still find the same oddball spikes.
[20:47:49] <Ironholds>	 I may have solved it.
[20:48:27] <Ironholds>	 basically, we have 5-6 uniques making the vast majority of requests at precisely 1200 and 1201 seconds.
[20:48:32] <Ironholds>	 I think this is the source of the problem.
[20:54:21] <Ironholds>	 halfak, ^
[20:54:29] <halfak>	 in meeting brb
[20:54:33] <Ironholds>	 kk!
[22:06:53] <Ironholds>	 this is like, the third meeting today that's been in some weird place. grumblegrumble fifth-floor-buildout already.