| --- Day changed Mon Sep 15 2008 | ||
| kanzure | Heh. In an email I'm writing at the moment, I have this idea going: | 01:23 |
|---|---|---|
| kanzure | I can randomly generate new designs for experiments using WordNet or Cyc relations or something, where we observe mice subjected to a battery of different psychochemical compounds. Then, using WordNet, we can pull out random labels for each of the behaviors, maybe it's a "goal" or maybe it's | 01:23 |
| kanzure | a who knows what that the creature supposedly intrinsically has, and then what? You'd plot the data sets in some multidimensional manner, maybe a State Vector Machine, I'd have to ask some mathematicians, | 01:23 |
| kanzure | and then there's this strong likelihood of statistical irrelevance of assigning these labels to the different phenotypes observed in experiments. These same "phenotypes" are the things of folk psychology as well. The trick is that instead of observing rats, you're observing people. | 01:23 |
| kanzure | http://www.adciv.org/Main_Page manufacturing + post-scarcity wiki | 01:45 |
| kanzure | Joseph correctly understands that all of the 'open manufacturing' and 'post-scarcity' advocates are far too fragmented | 01:51 |
| kanzure | part of the problem is that it's all ungrounded and that's why it's fragmented .. by 'grounded' I mean few of us have the machines, tools and so on to show how this stuff works | 01:51 |
| ybit | joseph, it's not christopher? | 02:22 |
| ybit | grr | 02:22 |
| ybit | charles | 02:22 |
| ybit | http://www.adciv.org/User:CharlesC | 02:22 |
| ybit | http://rand.rice.edu/research.php | 02:23 |
| ybit | ^ quite | 02:23 |
| ybit | ..nice | 02:23 |
| ybit | sounds like the whoever goes through that program would be working with intel | 02:24 |
| ybit | would/will | 02:24 |
| kanzure | Hrm. | 02:46 |
| kanzure | I've been insightful lately. I'm going to forget all of it, so before I do, what should I do? Writing doesn't work, coding doesn't work - all of the data is getting in too slowly for me to do anything productive with it - so what's left? | 02:46 |
| kanzure | Hey spookact. | 02:52 |
| spookact | what's up kanzure | 02:52 |
| kanzure | World domination. | 02:53 |
| kanzure | Hm. I need to go document some progress on stuff. | 02:54 |
| kanzure | http://en.wikipedia.org/wiki/OpenMP | 03:54 |
| kanzure | are preprocessor directives the right way to go about it? | 03:54 |
| kanzure | http://en.wikipedia.org/wiki/Ephemeralization | 05:07 |
| kanzure | 'Ephemeralization is a term coined by R. Buckminster Fuller. It refers to the ability of people to use technological advances to continuously do more with less. ' | 05:07 |
| kanzure | http://www.adciv.org/User:CharlesC | 05:09 |
| bkero | Sweet, hulu added the second season of regenesis yesterday. http://www.hulu.com/regenesis | 05:37 |
| kanzure | "brainblast" | 05:43 |
| kanzure | all of these materials handbooks | 05:43 |
| kanzure | the main thing that they are good for is their graphs of the characteristics of the materials | 05:43 |
| kanzure | there's only a finite number of variables, right? | 05:43 |
| kanzure | page segmentation algorithms to look for graphs on the pages | 05:44 |
| kanzure | then do some curve analysis | 05:44 |
| kanzure | or data point recovery etc. | 05:44 |
| kanzure | Unfortunately, this only works when you're certain of the context of the data plots .. not all graphs give their full context in the caption, but perhaps it can be implied from the subject of the paper or document or page - kinda crappy assumption, but meh | 05:45 |
| kanzure | There's this small, small chance that the excel data format that they provide on matweb.com might be usable | 07:02 |
| kanzure | free registration required | 07:03 |
| kanzure | doesn't work in konqueror | 07:03 |
| kanzure | was figuring I'd code up wget --save-cookies with the login form and then load 'em up for the excel spreadsheet request | 07:03 |
| kanzure | but, sleep | 07:03 |
| kanzure | who are all these peoples? | 16:05 |
| kanzure | http://www.cultureshop.org/details.php?code=COCOREV | 16:06 |
| faceface | no idea | 16:11 |
| faceface | hopefully I'll have some movies for you | 16:12 |
| faceface | lets see if they send them out... | 16:12 |
| kanzure | http://www.ted.com/index.php/talks/paul_stamets_on_6_ways_mushrooms_can_save_the_world.html | 16:28 |
| faceface | what if I don't like mushrooms? | 16:44 |
| kanzure | you expect me to actually view the link? | 16:45 |
| kanzure | I kid | 16:45 |
| * kanzure is hard pressed for time at the moment, will get back to the shrooms later | 16:45 | |
| kanzure | UTCS Corporate Connection/FoCS: Mark Kilgore & Serkan Piantino/Facebook: "What Makes Engineering at Facebook Unique" ACES 2.402, Tuesday, September 23, 2008 4:00 p.m. | 16:57 |
| kanzure | ooh | 16:57 |
| kanzure | http://www.facebook.com/utaustincareerexpo | 16:58 |
| kanzure | maybe I should go be antisocial with them | 16:58 |
| faceface | facebook apparently recommends friends | 17:01 |
| faceface | I have a story about a future like that that I shoudl really write up at some point | 17:02 |
| kanzure | how does the whole social networking thing work with the whole urge to stab them in the face | 17:02 |
| * kanzure questions these 'engineers' | 17:02 | |
| faceface | you want to stab the social network in the face? | 17:10 |
| * faceface invites kanzure to 1001 social networks | 17:10 | |
| kanzure | you mean one of those auto subscriber sites? | 17:15 |
| kanzure | those are interesting, but they're just like all the others because they're a website and not a script I can download | 17:15 |
| kanzure | I want to stab the users of the social nets in the face, that is | 17:15 |
| kanzure | the classic bash.org CRT through which you can stab people | 17:15 |
| kanzure | UTCS Corporate Connection/FoCS: Joseph Sirosh/Amazon.com: "Ahead in the Cloud: The Power of Infrastructure as a Service" ACES 2.402, Thursday, September 24, 2008 4:30 p.m. | 17:16 |
| kanzure | Anybody try out metweb for me? :/ | 20:07 |
| kanzure | http://books.google.com/books?id=gWg-rchM700C&pg=PA501&lpg=PA501&dq=matml&source=web&ots=xGE7qCZzph&sig=Fy1KjJ7-R3a06E9Hwhq7jWteJoc&hl=en&sa=X&oi=book_result&resnum=46&ct=result handbook of materials selection, mentions matml | 20:10 |
| kanzure | CODATA "International Register of Materials Database Managers" for possibly useful index | 20:14 |
| kanzure | "Ericsson (1998) posted his graudation thesis, 'Materials Information on the Internet', containing ~200 site links, well classified and individually annotated." | 20:15 |
| kanzure | lemme see now | 20:15 |
| kanzure | give me 75 graduate degrees, foo | 20:15 |
| kanzure | Ericcson owes me a beer. | 20:15 |
| bkero | Heh | 20:18 |
| bkero | kanzure: Give me your big processings jobs. I want to see how fast I can crunch through them. | 20:18 |
| kanzure | For the automated design lab work? | 20:18 |
| bkero | Sure | 20:18 |
| kanzure | Yeah, wait, how many nodes again can you give me? I'm wanting to do 10,000 to 10,000,000 permutations on designs or something (though not saving that final data set of course, only the good stuff). | 20:19 |
| bkero | 8 cores worth across 3 machines ATM | 20:21 |
| bkero | All ~3GHz 64-bit | 20:21 |
| kanzure | Hrm, I'm not sure the code base is utilizing multiple cores yet. I'll have to fix this. | 20:22 |
| bkero | Thread it out :) | 20:22 |
| bkero | If it's only single core I'll crank that shit through my 4.4GHz Opteron | 20:22 |
| kanzure | You bastard. | 20:23 |
| bkero | Haha | 20:23 |
| bkero | Single core 4.4 < dual core 3 | 20:23 |
| bkero | < quad core 3 | 20:23 |
| kanzure | Why can't *I* get a master's with 200 links? | 20:32 |
| kanzure | what's that, like an hour of bookmarking? | 20:33 |
| kanzure | jesus | 20:33 |
| kanzure | argh | 20:33 |
| * kanzure goes to class before he turns into the hulk | 20:33 | |
| nsh | don't aspire to other people's standards | 20:33 |
| nsh | :-) | 20:33 |
| kanzure | bah! | 20:33 |
| bkero | kanzure: Come on, he had to write a 200 page dissertation on his links. | 20:51 |
| procto | pfft materials research | 20:58 |
| procto | there are only 2 material you should should know | 20:58 |
| procto | graphene | 20:58 |
| procto | and nitinol | 20:58 |
| procto | one for things out of side of squishbodies | 20:58 |
| procto | the other for things inside squishbodies | 20:58 |
| * nsh wants to squishbodies | 21:04 | |
| kanzure | bkero: Bah! Extract the metadata from the links, problem solved. | 22:10 |
| bkero | Heh | 22:10 |
| kanzure | There are far more than 200 sites using the metadata. | 22:10 |
| kanzure | ybit: For fsck sake, I *have* the entire Nature collection. | 23:17 |
| kanzure | What the hell do you need this 'access' thing for ? | 23:17 |
| ybit | A unified vision of the building blocks of life, Jamey D. Marth - p1015 of Nature Cell Biology | 23:56 |
| ybit | kanzure^ | 23:56 |
| kanzure | Lemme see if I have it. | 23:56 |
| ybit | k | 23:56 |
| ybit | september 2008 | 23:57 |
| ybit | latest issue | 23:57 |
| kanzure | No, I don't. | 23:57 |
| kanzure | I'll fetch. | 23:57 |
| ybit | oh, it must be nice | 23:57 |
| kanzure | ? | 23:57 |
| ybit | to have access | 23:57 |
| ybit | instant* access | 23:58 |
| kanzure | I don't understand why Google doesn't do autoindexing with the publishers. | 23:58 |
| kanzure | oh right, because life sucks | 23:58 |
| kanzure | http://nature.com/ncb/ is ironically refusing me. | 23:59 |
| kanzure | If you can get me the exact link to the PDF, I can run it through a once-only proxy link thingy. | 23:59 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!