[00:47] SketchCow: not quite AT-related, but netflix just killed their API: http://developer.netflix.com/blog/read/Changes_to_the_Public_API_Program [00:48] see: "The changes, outlined below, are designed to allow us to focus our API efforts on supporting the products and features used most by our members. They are also designed to allow us to continue to offer the public API program in a way that aligns with our goals." ...then ... "We will no longer issue new public API developer keys." ... "We will no longer accept new API affiliates." [00:59] that sucks [01:19] These companies think they are helping themselves [01:19] someone is probably already working on a way to exploit netflix not having this feature anymore into a money making possibility [01:20] paid public api's! [01:20] that too [01:20] there are many I would pay to have brought back [01:22] why don't they think people would pay [01:22] because "pirates!!!!" [01:22] or something [01:22] :) [01:22] you know [01:22] it's not like the music industry is recovering or anything [01:23] I know, a bunch of bullshit so a fucking manager can act all powerful [01:23] It's not like linux users pay more for humble bundles than anyone else... [01:23] My word is LAW!!!1! [01:23] omf_: ah yes [01:23] like my office [01:23] "you must be in the office". [01:23] ..... even when we are doing a full reneveation and there isn't actually any room to sit down? [01:23] "Yes!" [01:23] ok, I'll come in, and not get any work done all day. Cool, thanks. [01:24] Managers who do that are worthless. They think that is how to do a good job [01:24] oh and we might ask you to come in at the weekend to help with moving stuff [01:24] Smiley, I had to come in at night a few times and help install servers and shit [01:24] me: heh, ok good luck with that [01:24] concidering I do what is basically 13hour shifts due to trains etc because you _won't_ let me work from home sometimes.... lets see how well that goes down. [01:24] omf_: occasionally is fine [01:24] we had a sys admin staff of 5 and yet programmers still had to come in and help [01:25] but not every single time [01:25] Could you of done those servers some other time if someone had let you? [01:26] nope. Scheduled maintainence windows were at night only [01:27] Hardcore bureaucracy [01:30] fun times. [01:30] we just don't do that [01:30] small enough to just laugh most of the time [04:08] so i have to redownload the video pages again [04:08] did something thing stupid [04:09] tryed to open a text file when selecting a everything [04:09] including the warc.gz [04:10] so if you guys don't get this sorry [06:54] Just saw the newest Star Trek trailer. I still have no clue what the plot is for that movie [10:09] so i'm near the end of the images ids of g4 [10:10] only about 40k ids to check to go [10:33] looks like my item for commodore 64 training tape is hot right now [10:34] its number 6 download in computersandtechvideos collection [10:41] I just uploaded a grab of the twit cleaner site [11:50] So an opinion/experience question [11:51] How much working space do you set aside for data projects? [11:51] I am currently setting aside 2tb and it is not enough [11:55] I just stacked up a few hundred gigs of archiveteam stuff and it takes forever to upload [11:58] you can get 1TB servers from ovh at your local kimsufi for <20€ / month [12:01] I am looking at their site now [12:40] http://googleresearch.blogspot.com/2013/03/learning-from-big-data-40-million.html [13:25] ersi, thanks for that article [13:29] aaaawww they don't talk about how they tie this information into freebase [13:29] It was mentioned in a freebase talk from 2 years ago, they were trying it out then [13:30] disambiguation is hard, but the work google has done benefits everyone [13:30] the data is in freebase which is CC-BY so everyone has it now. It is also cross referenced against the universal rdf ids [13:31] so more complex relations can be formed and not worry about stale data sources as the project continues [13:34] hell the schema system for freebase is a master class in taxonomy [13:34] RIP MetaWeb [13:35] everything they did except graphd is still open and updated by google. I think metaweb is resting nicely [13:35] I always tell people about metaweb so they realize google didn't start this, they bought it [13:35] Indeed [13:36] And people pissed on metaweb's head until google bought them [13:36] the common sentiment "It cannot be done" [13:36] Heh, Pitbull - Back In Time is a fitting tune to listen to while chatting in AT channels [13:36] yeah [13:36] which is quite unfortunate in itself [13:36] People never recognize greatness [13:37] semantic meaning on the web is now possible because of the decade of work metaweb did [13:37] 10 fucking years but it is worth it [13:38] I could take freebase, tie it against the wayback machine, do a sample set of learning and then be able to do facebook style graph searches against the past of the web [13:39] I never understood why google doesn't factor time into search [13:39] for a large majority of their searchable content they have reliable date information. [13:40] They could make billions just selling analytics off that [13:40] how a product or brand changes over time and map out the growth over the internet [13:41] then factor in they have incoming and outgoing link data on these sites. They could rank these people and then compare against a human tested corpus and bam new information [13:42] For us mere mortals the major problem is the fact that a server cluster to run this costs a small fortune [13:43] I'd like to know how much data they keep. Especially web history. [13:44] The n-gram data google released even as large as it is, is still manageable on a single server [14:34] http://www.osnews.com/story/26849/Google_called_the_MPEG-LA_s_bluff_and_won [18:28] so i have save a few missing clips [18:28] found there file names cause alot of images use the video name for image files [18:40] so i may try to grab the tech guy audio archive [18:59] i have uploaded over 15k videos to g4video-web collection [19:03] PresserTestH261_G4750.swf: http://archive.org/details/g4tv.com-video35810 [19:04] its a ces presser test video [19:04] also PresserTestH261_G4750.swf is the desc