--- Day changed Fri Jul 17 2009 00:04 -!- ybit [n=ybit@unaffiliated/ybit] has quit [Read error: 113 (No route to host)] 00:04 -!- ybit [n=ybit@unaffiliated/ybit] has joined #hplusroadmap 00:08 < kanzure> yaml.constructor.ConstructorError: expected a scalar node, but found mapping 00:08 < kanzure> in "", line 6, column 12: 00:08 < kanzure> blockhole: !part 00:08 < kanzure> genehacker: multiple proteins from a solution 00:09 < kanzure> based on mass and/or density 00:09 < kanzure> (not sure which) 00:10 < CIA-73> skdb: kanzure master * r5d1467c / pymates/pymates.py : ugh, errors with pymates - http://bit.ly/6fuYm 00:26 < genehacker> multiple proteins seperated? 00:26 < genehacker> if so let's build one 00:27 < genehacker> also check out Precision Monolithic Casting in Fluidics Quarterly 9 in the Eng Library 00:27 < genehacker> will give scans later 00:34 < genehacker> how about around lunch tomorrow? 00:35 < kanzure> maybe. 00:35 < kanzure> i need to do something in the morning, but I will probably be on campus by noon 00:37 < genehacker> ok cool 00:38 < kanzure> I'm not digging the MPTS crosslinks for the monoliths in the CD protein purification device 00:38 < genehacker> yes me either 00:38 < genehacker> sigma aldric 00:38 < kanzure> *crosslinker 00:39 < genehacker> you mean what they used for the colum 00:39 < genehacker> n 00:39 < kanzure> yes 00:39 < kanzure> I'd like to see a micrograph of the column. 00:39 < kanzure> oh wait they had one in the paper 00:40 < genehacker> it's some weird silicon compound 00:40 < genehacker> maybe DRT could make it 00:40 < genehacker> you have >3 ochem books 00:40 < genehacker> can I borrow one 00:40 < kanzure> which one 00:40 < genehacker> anyone 00:40 < kanzure> um. sure. I recommend the PDFs though. 00:40 < kanzure> you sure you want a physical book? 00:41 < kanzure> dead tree format 00:41 < genehacker> PDFs? 00:41 < genehacker> where? 00:41 < genehacker> fork em over! 00:41 < kanzure> http://heybryan.org/books/Chemistry/ 00:41 < kanzure> i think the principle of their MTPS-crosslinked column may be similar to the logarithm involved in gel electrophoresis 00:42 < genehacker> what? 00:42 < kanzure> it doesn't look like hte proteins have any particular chemical interaction with the column 00:42 < genehacker> oh 00:42 < kanzure> do you remember the experiment in physics class where you would have this bed board of nails, 00:42 < genehacker> well duh 00:42 < kanzure> you would put it on an elevation 00:42 < genehacker> yes 00:42 < kanzure> and then roll marbles down and track where they would land with a probability distribution 00:43 < kanzure> I suspect that's what's going on here 00:43 < genehacker> ummmm 00:43 < genehacker> more like gel electrophoresis 00:44 < genehacker> where heavier stuff goes slower 00:44 < genehacker> or do you not get the principle of gel electrophoresis 00:45 < genehacker> or do I not understand what you're talking about? 00:45 < kanzure> gel electrophoresis can be described by a logarithm 00:46 < genehacker> how so? 00:47 < genehacker> give me the formula 00:47 < kanzure> small fragments travel much faster than large fragments 00:48 < kanzure> consider the ferguson plot 00:48 < kanzure> http://www.his.com/~djt/ferguson.gif 00:48 < genehacker> heybryan is down 00:49 < genehacker> and that shows? 00:49 -!- wrldpc2 [n=benny@72.72.22.132] has joined #hplusroadmap 00:49 < kanzure> it should be back up 00:49 < kanzure> it's the log of mobility versus gel concentration 00:49 < genehacker> yup 00:49 < kanzure> guess that's not as helpfulexplaining it 00:50 < genehacker> mobility goes WAY DOWN with gel concentration? 00:50 < genehacker> got it 00:50 < kanzure> http://www.aesociety.org/areas/ferguson.php 00:50 -!- wrldpc2 [n=benny@72.72.22.132] has quit [Client Quit] 00:50 < kanzure> with infinite gel concentration, mobility hits zero 00:50 < genehacker> FUNDAMENTALS OF NUCLEATION AND GROWTH 00:50 < genehacker> FUCK YEAH 00:50 < genehacker> that's on my homework 00:52 < genehacker> oh wait maybe not 01:02 < CIA-73> skdb: kanzure master * r257b63a / pymates/monday.py : it would be nice if this worked by monday - http://bit.ly/ifxf1 01:08 < CIA-73> djangit: Meredith L. Patterson master * r0047918 / docs/mlp-workingnotes : 01:08 < CIA-73> djangit: Committer: Meredith L. Patterson 01:08 < CIA-73> djangit: On branch master 01:08 < CIA-73> djangit: Changes to be committed: 01:08 < CIA-73> djangit: (use "git reset HEAD ..." to unstage) 01:08 < CIA-73> djangit: new file: docs/mlp-workingnotes - http://bit.ly/5ndE2 01:08 < CIA-73> djangit: kanzure master * re684bc8 / wiki/views.py : merged maradydd's notes - http://bit.ly/btAfY 01:13 < genehacker> ??? 03:26 < genehacker> where can I access the chat logs 03:29 < genehacker> heh 03:29 < genehacker> eznurak 05:07 < fenn> why does cia say "changes to be committed:" 05:09 < fenn> i wish one could add comments to comments 05:09 < fenn> it'd probably just turn into some sort of torah-splosion 06:27 -!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has joined #hplusroadmap 06:34 < fenn> i must admit i dont see the point of this http://gitorious.org/free-manufacturing 06:34 < fenn> nor do i understand smari's reply, which seems non-sequitur 06:35 < Smari> fenn, It was an attempt to guide him onto the right track without outright saying that he was barking up the wrong tree. 06:35 < Smari> besides, he isn't. 06:39 < Smari> What he's done so far is fairly good 06:39 < fenn> is there a human-readable view of the latex somewhere? 06:40 < fenn> i dont actually know how to render latex 06:40 < Smari> hang on, I'll do it. 06:41 < Smari> By all means learn it. It's a good typesetting system. 06:41 < fenn> bah 06:41 < Smari> it's on its way... 06:41 < fenn> asciidoc does everything i want 06:42 < Smari> that must mean that you do not want normal people to read your documents. Nor have pretty equations. Nor have beautiful diagrams or good margins or awesome PDF files. 06:42 < fenn> you know i really hate "pretty equations" with a passion 06:42 < Smari> haha 06:42 < Smari> why? 06:43 < fenn> and latex always seems to make this bobbly up-down text which nobody but me ever mentions 06:43 < fenn> Smari: they assume you know wtf they mean in the first place 06:43 < Smari> fenn, don't confuse pretty equations with arrogant mathematicians. 06:43 < fenn> how am i supposed to know what "d" stands for, or that funny squiggle, or your particular arrangement of glyphs 06:44 < fenn> there's no sense to it 06:44 < fenn> its 100% pure arbitrary convention 06:44 < Smari> So is writing in general, isn't it? 06:44 < fenn> at least i can look up words 06:44 < Smari> You can look up notation too. 06:44 < fenn> how do you look up glyphs positioned with respect o another? 06:44 < Smari> mathematical notation has been very much improved over the last 100 years, but it still suffers from its users being very arrogant and not wanting to spend time explaining what the hell they mean. 06:45 < Smari> Juxtaposition bears context, which, if you understand the functionality encoded in a glyph, you will also understand. 06:45 < fenn> how do you look up notation? 06:48 < Smari> How do you look up symbols in general? 06:48 < Smari> When encountered with, say, the greek letter xi, how do you know what it is? 06:48 < Smari> http://detexify.kirelabs.org/classify.html <-- this helps. 06:49 < Smari> http://smari.yaxic.org/fmas.pdf 06:51 < fenn> doesnt detexify require me to have the latex source in the first place? 06:51 < Smari> if you don't like equations, how do you signify things like, oh, say, Maxwell's equations? 06:51 < fenn> that is to say, not a "pretty equation" 06:51 < fenn> i just dont like the hieroglyphics 06:51 < fenn> math is cool 06:52 < Smari> So you're okay with integral symbols and 'dx' and division lines and subscripts and powers and all that? 06:52 < fenn> but most equations would generate a "error: foo is undefined" if a computer tried to read it 06:52 < fenn> (and the surrounding text) 06:53 < fenn> Smari: NO that's the crap i'm against 06:53 < Smari> That's because computers trying to parse that typically don't have the ability to realize abstract relationships. 06:53 < Smari> They expect every variable to have a value. 06:53 < fenn> there's nothing abstract about it 06:53 < Smari> Are you saying that y = a+bx is not abstract? 06:54 < Smari> That that particular formula is completely useless without the granted context of the symbols having values? 06:54 < fenn> the problem with that is you dont know which symbols are constants 06:54 < Smari> Then assume none to be. 06:54 < fenn> you have to 'just know' that x is a variable and b is a constant 06:54 < Smari> x need not be a variable, b need not be a constant. 06:54 < fenn> fine 06:55 < Smari> You could plot that function for varying values of b and a.. 06:55 < fenn> but 99% of the time, an arrogant mathematician will expect you to know that you're parameterizing x 06:55 < Smari> yeah, I agree that that is a very bad thing. 06:55 < Smari> I generally don't disagree with you on this.. it is very arbitrary. 06:56 < fenn> see "zen of python" etc etc 06:56 < Smari> but I'm trying to convey that in this case the price of criticism is a viable alternative. 06:56 -!- wrldpc2 [n=benny@pool-173-48-253-182.bstnma.fios.verizon.net] has quit [Read error: 113 (No route to host)] 06:56 < fenn> er, please reformulate that last sentence 06:57 < Smari> Show me a sensible contextual syntax with which I can describe, to grab an model at random from the book beside me, Padé approximants. 06:57 < fenn> well, lisp is one example.. 06:57 < fenn> not that i'm saying everyone should switch to lisp overnight 06:58 < Smari> Where everything looks the same; very horrible to visually parse. 06:58 < Smari> The thing about math notation is that it's whiteboard friendly. 06:58 < fenn> (and only whiteboard friendly) 06:58 < Smari> And for people like mathematicians who do all of their thinking on whiteboards, that's fine. 06:58 < Smari> As long as they realize that it's only one of many possible notations. 06:59 < fenn> i just wish it didnt creep into everything else 06:59 < Smari> One of my professors once ran through a proof in symbolic logic that you could, logically and without contradiction, use the same symbol repeatedly in a logical statement with different meanings. 06:59 < fenn> dammit there ought to be download/executable source for every engineering paper 07:00 < Smari> but yes, most notation does suck. I don't know how often I've been confused by notation of something relatively simple... and I'm with Lockhart on the issue that teaching notation should be secondary to teaching mathematics. 07:01 < Smari> A lot of schools teach notation as if it were mathematics. 07:01 < Smari> Teach kids to manipulate symbols in a notational grammar, without understanding the underlying abstraction. 07:01 < fenn> 93 pages in category Mathematical Notation 07:02 < fenn> i'm glad computers can't parse hieroglyphics 07:02 < Smari> you hit a kid over the head with (a+b)² = a²+2ab+b² often enough and he'll remember it and possibly be able to use it, but it won't do him a bit of good because he won't understand it unless he starts playing around with it and explores the implications. 07:02 < Smari> Actually.. have you seen Fortress? 07:03 < Smari> http://research.sun.com/projects/plrg/PLDITutorialSlides9Jun2006.pdf 07:03 < Smari> Check it out. 07:04 < fenn> wasnt that sort of the point of haskell? but without all the icky symbols 07:05 < fenn> "program structure should reflect the science" 07:09 < Smari> Dude. You can't even open a file in Haskell without resorting to Voodoo. 07:09 < Smari> Haskell doesn't reflect any science except computer science. 07:10 < Smari> I would love to see somebody try to write something like haplotype association analysis software in Haskell. 07:10 < fenn> actually that sounds like something it'd be good at 07:11 < fenn> factoring large sets 07:11 < Smari> Perhaps. 07:11 < Smari> Still sounds like a very far-fetched thing.. 07:12 < Smari> Lunchtime. Let's see if I can find some food. 07:12 < fenn> "maybe you should have to ask for sequential execution" (from fortress) 07:12 < Smari> isn't it like, 7 in the morning for you? 07:12 < fenn> yes, i got up a couple hours ago 07:19 < Smari> Damnit, my falafel isn't hot through 07:19 < Smari> rushing too much 07:19 < fenn> burn it! burn it all! 07:23 -!- genehacker [n=noko@67.9.157.136] has quit [Read error: 110 (Connection timed out)] 07:25 -!- genehacker [i=genehack@wireless-128-62-158-110.public.utexas.edu] has joined #hplusroadmap 07:43 < kanzure> Smari: you can't generally google an equation. 07:45 < Smari> kanzure, agreed. 07:45 < fenn> paul hasnt been writing to OM, maybe that's why it seems much more topical than usual 07:46 < Smari> hahaha 07:48 < Smari> I don't know which is funnier.. that you should make that connection, or that you should consider it to be currently topical. 07:49 < kanzure> it's just sad that paul is so far off track 07:49 < kanzure> so that can't be it:) 07:50 < fenn> Smari: well i'm quite excited about that cable based dremel thingy 07:51 < fenn> maybe i've just been seduced by the video production 07:52 < Smari> fenn, me too. I set it as "featured video" on the fablab.is/wiki 07:52 < Smari> Before that there was no such thing as a "featured video" on the wiki ;) 07:52 < fenn> heh 07:52 < kanzure> huh didn't know fablab.is had a wiki 07:53 < kanzure> it would be nice plugin for djangit to have git-pull requests for documentation from multiple wikis or something for a common fablab namespace 07:53 < kanzure> just a pipe dream 07:53 < fenn> that's what samrose is on about "flows" 07:54 < fenn> git seems to work well enough for the linux kernel 07:54 < Smari> kanzure, the wiki is the only thing I touch on fablab.is; the rest is managed by an ass-backwards web management system that only works in Internet Explorer. 07:54 < Smari> Not my choice. 07:54 < Smari> Company policy. 07:54 < Smari> Another reason for why I'm quitting. 07:55 < fenn> news, drama! 07:55 < fenn> what could be more exciting than running a fab-lab in the middle of the ocean? 07:55 < kanzure> on this episode of hplusroadmap, smari meets the league of doom and faces a tough decision 07:56 < fenn> Smari: did Kokompe ever go anywhere? 08:04 < kanzure> fenn, do you know what this error is really saying? 08:04 < kanzure> yaml.constructor.ConstructorError: expected a scalar node, but found mapping 08:04 < kanzure> in "", line 6, column 12: 08:04 < kanzure> blockhole: !part 08:05 < fenn> it's probably the line before it 08:06 < fenn> is that code up somewhere? 08:06 < kanzure> yes, it's in pymates/models/blockhole.yaml 08:06 < kanzure> and the error is generated when running pymates/tests.py 08:06 < fenn> where are you pushing to lately, github? 08:06 < kanzure> for some reason when I push to /var/www/skdb.git/ my git wants to push to /var/www/skdb/ 08:06 < kanzure> it's up to date on hte server 08:07 < kanzure> *the 08:08 < fenn> urls: 08:08 < fenn> - 'http://heybryan.org/' 08:08 < kanzure> ? 08:08 < fenn> uh wait nevermind 08:12 < Smari> fenn, not really. 08:12 < kanzure> it might be because of the word scalar in my pymates.py code 08:13 < Smari> fenn, The project died because most of the people involved had never worked on a distributed free software project before and therefore didn't know how to use mailing lists and chatrooms effectively. :P 08:13 < Smari> fenn, there is a fairly neat bunch of code though.. 08:13 < Smari> A pretty powerful CSG engine that is just begging to be used for something real. 08:14 < kanzure> CSG? really? 08:14 < kanzure> can you link to that? 08:14 < fenn> is this code available for mere mortals to share/modify? 08:14 < fenn> cam.py had some weird license 08:14 < Smari> 3cm/minute, 30 cm, ... 90 minutes. 08:14 < Smari> fenn, MIT license, not very weird. 08:15 < Smari> I don't know where the git currently lives... 08:15 < kanzure> Smari: that whole "don't know how to collaborate" issue has come up with the diybio folks too :/ 08:15 < kanzure> Smari: you should wrangle your code more effectively 08:15 < Smari> another cause of it dying is that the repo kept getting moved around without anybody being notified. 08:15 < fenn> in particular cam.py said "Permission granted for experimental and personal use;" 08:16 < fenn> whatever that means 08:16 < Smari> we originally hosted it on the MIT Media Lab cluster, but that crashed (!) and it got pushed to davidad.net, and from there somewhere else. 08:16 < kanzure> davidad.net is david dalrymple 08:16 < kanzure> hah 08:17 < fenn> i thought the point of universities was that they hosted stuff for you 08:17 < Smari> kanzure, yup 08:17 < kanzure> i stalked david last year and met him 08:17 < Smari> haha 08:17 < Smari> I met him for the first time in Norway. 08:17 < kanzure> bachelors degree at age 12, CBA by age 15, taught by kurzweil and gershenfeld-- I hated him immediately 08:18 < kanzure> (I kid, I kid) 08:18 * kanzure has been jealous 08:18 < Smari> Second time in Chicago, third time in Boston. It seems that I never meet Fab Lab folks twice in the same country. 08:18 < Smari> (Boston is in Massachusetts, Chicago is in Illinois. I don't consider federations to be countries) 08:19 < kanzure> wouldn't have guessed by anarchism.is 08:19 < Smari> ;) 08:19 < Smari> laser etching marble is coming out at a loss. 08:19 < kanzure> could you find the repo? that would be hot 08:19 < Smari> yeah 08:19 < Smari> or 08:19 < Smari> rather 08:20 < Smari> I'll make a tarball and share it 08:20 < fenn> repo would be better eventually 08:20 < Smari> true 08:20 < Smari> but I don't have one at the moment ;) 08:20 < Smari> Hmm. 08:20 * fenn has to go pretend to do stuff 08:20 < Smari> Are you guys familiar with Forrest Green perchance? 08:21 < Smari> He's from Austin methinks. 08:22 < Smari> uploading the tarball now.. perhaps I should have "make clean" in it first, but I didn't, so there. 08:23 < kanzure> people like you who don't 'make clean' are the reason why .. somethings omething something 08:23 < kanzure> I'm sure you can come up with something sufficiently insulting there 08:24 < Smari> Woah 08:24 < Smari> There's some very neat stuff in there. 08:24 < kanzure> oh? 08:24 < Smari> I had forgotten about a lot of it. 08:29 -!- genehacker [i=genehack@wireless-128-62-158-110.public.utexas.edu] has quit [Read error: 110 (Connection timed out)] 08:30 < Smari> okay, uploaded with objects :P 08:30 < Smari> http://smari.yaxic.org/kokompe.tar.gz 08:31 < Smari> a lot of the beauty resides in engine/ 08:31 < Smari> there's some UI attempts in other directories, mostly failed 08:32 < Smari> http://i32.tinypic.com/beiu10.jpg 08:34 -!- samrose [n=samrose@c-24-11-214-181.hsd1.mi.comcast.net] has joined #hplusroadmap 08:41 < kanzure> 37MB? 08:41 < Smari> *shrug* 08:42 < kanzure> I am sometimes surprised about how large sources can get 08:42 < kanzure> for instance, I recall firefox being something like 150 MB 08:43 < Smari> Uncompressed it contains 29 MB of testcases. 08:43 < kanzure> O.o 08:43 < kanzure> huh, david dalrymple actually contributed code? 08:43 < samrose> hey kanzure, how are the grammars we discussed yesterday generally generated? 08:43 < kanzure> oh, I thought this was the silly desktop environment thingy for semantic webs 08:43 < Smari> kanzure, yes? 08:43 < kanzure> samrose: you should read graphsynth.com I guess. but allow me to explain. 08:43 < samrose> I am reading now 08:44 < samrose> looks like they enter in 08:44 < Smari> kanzure, so more than half of the archive is just .stl files, etc. 08:44 < samrose> to tables 08:44 < kanzure> samrose: in graph grammars, there are sets of rules. a "rule" is basically a search-and-replace rule that finds something like on the left and then replaces what it finds with the right 08:44 < kanzure> so yes, a domain expert has to come up with the grammar rules 08:44 < kanzure> however, recently one of the programs that Tanin has been working on in the lab 08:44 < kanzure> is a way to automatically extract rules from a graph 08:44 < samrose> based on Chomsky's theories, looks like 08:44 < kanzure> given something like "what the graph generically solves" and "the actual solution" 08:44 < kanzure> yes 08:45 < kanzure> (one of the boxes in the lab is named chomsky) 08:45 < kanzure> so there's this tree of possibilities that is generated 08:45 < samrose> does the automatic extraction use Natural Language Processing by chance? 08:45 < kanzure> by successive application of different recognized left-hand sides 08:45 < samrose> ah 08:45 < kanzure> samrose: I don't know what that means 08:45 < kanzure> in the case of rule extraction, there are two graphs that are being worked with 08:46 < samrose> I have been working with natural language processing as a way to auto tag content 08:46 < samrose> and looking at graphsynth, it made me think that it could apply to these grammars 08:46 < kanzure> (1) a "Function Structure Graph" which shows nodes and edges connecting together to symbolically represent something like "import human energy, convert to mechanical energy" 08:46 < kanzure> (2) the "Component Structure Graph" (CSG) 08:46 < kanzure> er, CFG. Component Flow Graph or Configuration Flow Graph 08:46 < kanzure> anyway, the CFG is where you see parts connected to one another that solve something in the Function Structure Graph 08:47 < kanzure> so in automatic rule extraction it's something like taking a look at the tags in the CFG and look at what functions they claim to be solving 08:47 < kanzure> and from there you can extract possible new rules to try out on your data set 08:47 < samrose> I see 08:47 < samrose> that is really awesome 08:47 < kanzure> fenn and I think it's kind of lame 08:47 < kanzure> it works though 08:48 < kanzure> but not for actual engineering 08:48 < samrose> Well, the concept is interesting 08:48 < samrose> if you are interested in networks, stuff like that 08:48 < kanzure> right 08:49 < samrose> the evolutionary design that I have been keeping track of works with different data 08:49 < samrose> but the process is somewhat similar 08:51 < kanzure> Smari: looks like you have some algebraic classes in kokompe for some reason. did you consider sympy? 08:52 < Smari> kanzure, no we didn't. The algebra in kokompe is a real interval algebra that nobody seems to have devised in software before... at least not freely available. 08:52 < kanzure> literature references? 08:52 < Smari> There's a file there somewhere, OpenCAD.pdf, which is a slideshow from a talk I gave in Chicago that explains the concept. 08:53 < kanzure> okay, thanks 08:53 < samrose> I wonder if this rule extraction could work with natural language processing. So that you could take something like skdb data, and process it first, then sort based on names, verbs etc etc, then train a program to use this to build rules. 08:53 < kanzure> samrose: did you see monday.py? 08:53 < samrose> nat lang processing outputs data like this http://socialsynergyweb.org/culturing/blog/stanford-parser 08:53 < kanzure> http://github.com/kanzure/skdb/blob/257b63a5f99b33582dec43c7fc39af179eb2dc48/pymates/monday.py 08:54 < Smari> kanzure, that said, it was made in OpenOffice, so it might exist as .odp or even (gasp!) .ppt 08:54 < kanzure> so, compatibility is a method that I'm supposedly writing 08:54 < kanzure> which will figure out whether or not two parts can be combined 08:54 < kanzure> this is not the same thing as a grammar rule really 08:54 < kanzure> one idea for "automated design" that I've been meaning to flesh out more thoroughly goes something like this.. 08:55 < kanzure> there are all sorts of protocols in biology and other industries 08:55 < kanzure> even in cooking, and each requires different instruments at different calibrations 08:55 < kanzure> and different parameterizations 08:55 < kanzure> so it would be hot if you could somehow figure out a chain of parts or tools that effectively implement the same solution 08:55 < kanzure> even if they look different 08:56 < kanzure> so, for instance, there's this peanut-butter-jelly functionCAD file running around here somewhere on mst.edu (somewhere)- it shows the process or recipe of making a PBJ sandwhich. but you could implement the machine with a giant toaster-thingy, or you could have someone in a kitchen carry it out 08:56 < kanzure> so it might not be rule-building, but the end result- a sort of 'automated' design- is kind of the same 08:57 < samrose> yes, this seems dependent on how "solution" is represented 08:57 < samrose> what I mean: 08:57 < samrose> how a machine can compare outcomes 08:58 < kanzure> if you're interested in the comparison of outcomes, 08:58 < kanzure> there's a paper that I have been writing on that 08:58 < samrose> and know a similar solution has been acheived 08:58 < kanzure> I turned it in last friday 08:58 < kanzure> see the "clustering" header here: 08:58 < kanzure> http://adl.serveftp.org/dokuwiki/scripts 08:58 < kanzure> we did a comparison of designs via something called a DSM 08:58 < kanzure> an adjacency matrix of components from the database 08:58 < kanzure> and then took the euclidean norm of the matrix difference 08:58 < kanzure> and called this the "difference" 08:58 < samrose> I am familiar with clustering 08:59 < samrose> interesting 08:59 < kanzure> there are also some other distance-metrics, like principle-component analysis 08:59 < kanzure> which should definitely be incorporated 08:59 < kanzure> (I threw together k-means clustering of the euclidean distance metrics. Nothing spectacularly interesting.) 09:00 < Smari> I've got a paper to write later on on the ferry: "An infrastructure-centric view of the Icelandic financial collapse"... I'm not sure whether to write it in English or Icelandic though. 09:00 < Smari> Yay, I can't wait for today's ferry ride! 09:00 < kanzure> ferry ride off or on to the island? 09:00 < Smari> off 09:00 < Smari> Going home for the weekend. 09:00 < samrose> where is "home"? 09:00 < Smari> 3 hours out at sea, just me, my laptop, loud music, no Internet. :) 09:01 < Smari> Reykjavík. 09:01 < Smari> but I work in Vestmannaeyjar until the 7th of August. 09:01 < Smari> It's about 4 hours travel between the two places... 3 hours by ferry and an hour's drive... I go home when I can, which is mostly just weekends. 09:02 < samrose> heh, this is pretty innovative if you ask me http://adl.serveftp.org/dokuwiki/scripts 09:02 < samrose> what you outline there 09:02 < samrose> Smari maybe will be a calm evening too 09:03 < samrose> weatherwise 09:03 < Smari> samrose, yeah. It's been calm all week. I was hoping to go out ribbing this week but things didn't pan out.. 09:05 < samrose> got this book from K ̄rlis Repsons via email 09:05 < samrose> heh, not sure what happened there 09:06 < samrose> "Free Manufacturing" 09:06 < Smari> Yeah 09:06 < Smari> did you get the .pdf? 09:07 < samrose> yes 09:07 < samrose> thanks 09:08 < samrose> he seems to be talking about what we are talking about 09:12 < samrose> there are some "rules" that he is asserting that I don't agree with... 09:12 < samrose> but, it is interesting never the less 09:16 < Smari> yeah 09:17 < Smari> I don't think he's barking up the wrong tree, I think he's just making too many false assumptions.. 09:20 < Smari> He's also trying to apply terms like "axiom" to situations where they don't apply. 09:32 < samrose> Yes, I think you could point to construction by bees, ants, other insects to prove some of his rules wrong 09:32 < samrose> for instance 09:37 < Smari> yup 09:40 < fenn> woah 8MB/s from yaxic.org 09:55 -!- kardan_ [n=kardan@p54BE704D.dip.t-dialin.net] has joined #hplusroadmap 09:56 -!- kardan| [n=kardan@p54BE6E89.dip.t-dialin.net] has quit [Read error: 60 (Operation timed out)] 10:21 -!- genehacker [i=genehack@wireless-128-62-88-74.public.utexas.edu] has joined #hplusroadmap 10:21 < Smari> fenn, I have good tubes. 10:41 < drazak> kanzure: any luck on any of those papers? 10:42 < fenn> drazak plz post a link to the urls again 10:43 < drazak> Http://drazak.net/~drazak/pubmedlist.txt 10:43 < drazak> my uncle found me Discher, Putra, and Malchesky 10:44 < kanzure> I don't mean to be "that guy", but, where the hell did my mailbox go 10:44 < fenn> bulldozed 10:45 < fenn> there's some letter on the counter about it 10:45 < fenn> drazak: how do i get full text from that page? 10:45 < drazak> fenn: you're looking at teh wrong guy 10:45 < kanzure> pubmed does not give you full text 10:45 < drazak> no idea 10:45 < kanzure> pubmedcentral does 10:45 < kanzure> pubmed just gives you abstracts 10:45 < kanzure> if you gave a list of pubmed abstract links, fuck you 10:45 < kanzure> you deserve to not get the papers :p 10:46 < kanzure> but you can try pubget.com 10:46 < kanzure> http://pubget.com/ 10:46 < kanzure> it largely only works for pubmedcentral 10:46 < fenn> i figured it out 10:47 < kanzure> ? 10:47 < fenn> you click on the logo of the journal 10:47 < drazak> eh 10:47 < drazak> I don't think you guys will have much luck, not that I think about it 10:47 < drazak> my uncles harvard journal collection thing only found 3 of them 10:48 < drazak> I wanted teh other ones, but I don't think I'm gonna be able to get them 10:48 < drazak> s/not/now/ 10:50 -!- Smari [n=spm@nmi-gw.eyjar.is] has quit [Read error: 113 (No route to host)] 10:54 < kanzure> fenn: any ideas about my bug 10:54 < fenn> drazak i got 1 2 3 4 6 7 9 10:56 < fenn> http://adl.serveftp.org/papers/drazak.zip 10:57 < CIA-73> skdb: kanzure master * ra401efc / pymates/monday.py : want some overloads eventually - http://bit.ly/9Atkx 10:58 * kanzure needs to figure out how to add some hooks 10:59 < drazak> fenn: thanks dude 11:05 < fenn> gosh kanzure how did you manage to write such broken code in so few lines 11:05 < kanzure> what's so broken? 11:13 < kanzure> weird I looked up my error message and got to here: 11:13 < kanzure> http://pyyaml.org/ticket/79 11:13 < kanzure> which was submitted by clay@shirky.org 11:14 < kanzure> i think I'm supposed to know that name 11:24 < kanzure> yay got it to suck slightly less 11:24 < fenn> kanzure: pull please 11:25 < kanzure> merge conflict in pymates/pymates.py. /me investigates 11:26 < kanzure> ah, looks like we both made Part.__init__ slightly better 11:26 < kanzure> heh I like how we made the same damn changes 11:26 < fenn> yay python 11:26 < kanzure> what does loader.construct_mapping() do that loader.construct_scalar() doesn't? 11:27 < fenn> constructs a mapping instead of a scalar 11:27 < kanzure> gee thanks 11:27 * fenn wonders how to explain this 11:27 < kanzure> why does python care about the difference though 11:28 < kanzure> gah our timestamps are off 11:28 < kanzure> you committed one minute into the future of when I committed the fixed merge 11:29 < CIA-73> skdb: kanzure master * rbb9c88a / (pymates/monday.py pymates/pymates.py): make it suck slightly less - http://bit.ly/1Jn8e 11:29 < CIA-73> skdb: fenn master * r0b6ccdd / pymates/pymates.py : construct_scalar only works for scalar values.. who woulda thought - http://bit.ly/zG92M 11:29 < CIA-73> skdb: kanzure master * rf897824 / pymates/pymates.py : fixed merge issues. gee. - http://bit.ly/69uye 11:36 < kanzure> fenn: there's a way that you can add cia.vc to oyour post-commit hook 11:36 < kanzure> *your 11:37 < fenn> i'm sure there is 11:37 < fenn> would you rather have cia reposrt status of adl than github? 11:38 < kanzure> yeah 11:38 < kanzure> although the github links are niece 11:38 < kanzure> *nice 11:38 < fenn> what does the hook look like? 11:38 < kanzure> it was a clicky button pressy interface on github 11:38 < fenn> oh 11:38 < kanzure> apparently you have to download a script and call the script from your commit hook 11:39 * fenn looks 11:39 < kanzure> registration for skdb and cia is open to anyone though, no passwords on http://cia.vc/ by default. 11:40 < fenn> um really? it's asking me for user/pass 11:40 < kanzure> you still have to log in anyway 11:40 < kanzure> it's so that they can yell at you if you disrupt an actual project 11:41 < fenn> i'm thoroughly confused 11:41 < fenn> why do you keep turning wondershaper off 11:41 < kanzure> when you log in, you add a project to yourself 11:41 < kanzure> wondershaper isn't off 11:42 < fenn> then why are you getting 1000ms ping times 11:42 < kanzure> and if it was, I just turned it back on 11:42 < kanzure> I'm getting 500 ms ping times to google 11:44 < fenn> it looked like you were messing with it and now it's back to ~1500ms 11:45 < kanzure> why does it suck so much 11:46 < fenn> because you're overloading it, for the fiftieth time 11:47 < kanzure> what am I overloading it with 11:47 < kanzure> ooh scarry irc text 11:47 < kanzure> scary 11:47 < fenn> heybryan.org why do you think i keep turning it off 11:48 < kanzure> what's the point of telling me to use wondershaper if that doesn't fix it 11:48 < fenn> because you arent using wondershaper? 11:48 < kanzure> but I am using wondershaper 11:48 < kanzure> wondershaper eth1 400 400 is what we found "worked" 11:48 < kanzure> but apparently it's not working anymore? 11:49 < fenn> but you aren't, because when you say you turn it on, it works 11:49 < fenn> can you read a ping result? 11:49 < kanzure> the time= part? 11:49 < fenn> here's the procedure 11:49 < fenn> ping google.com, turn wondershaper upload down until ping time is reasonable (<250ms) 11:50 < fenn> then add that to your local bootup script 11:50 < kanzure> average roundtrip is 189.47 ms 11:50 < kanzure> isn't that reasonable? 11:50 < fenn> do ping from a different computer 11:51 < fenn> wondershaper favors its own short packets 11:51 < kanzure> okay 11:51 < fenn> whatever it's at now is good 11:52 < kanzure> wondershaper eth1 200 200 11:52 < kanzure> (fyi) 11:52 < fenn> try 400 just out of curiosity 11:53 < kanzure> it was at 400 previously 11:53 < kanzure> ok. set to 400. 11:53 < kanzure> ok reset to 200. 11:54 < fenn> wow i thought 400 would work 11:54 < kanzure> so apparently 50 kbps breaks the router :p 11:54 < kanzure> but 25 kbps doesn't 11:55 < fenn> yeah typically they have an upload of about 40kB/s 11:58 < genehacker> http://www.technologyreview.com/blog/arxiv/ 11:58 < genehacker> electron beam on chip 11:58 < genehacker> kanzure how do I access the chat log? 12:01 < kanzure> do you keep your own logs? 12:02 < kanzure> chatzilla can do logging for you 12:02 < kanzure> so if so you just look on your hard drive 12:02 < kanzure> if you do not do your own logging, then you have to pray to ybit 12:11 < genehacker> shoot 12:12 < genehacker> I posted something a while back about a bacteria that produces toluene 12:12 < genehacker> or did it? 12:12 < genehacker> wikipedia article has no info on it 12:13 < kanzure> fenn: when I load something from yaml, all of the attributes go into the first named parameter of my object. why? 12:14 < fenn> mapping is wrong? 12:15 < genehacker> heh 12:15 < kanzure> fenn: what mapping is wrong? 12:15 < genehacker> that AB matter has an ultimate tensile strength of 160 wekanewtons per square meter 12:16 < kanzure> i sure hope that's not an si unit 12:16 < genehacker> it is 12:16 < genehacker> weka is 10^30 12:16 < genehacker> of something 12:18 < genehacker> a cable of 1 square meter of AB matter would weigh ALOT 12:19 < genehacker> but one needs AB matter to produce it industrially 12:20 < CIA-73> skdb: * r3656f597af8d /change.txt: yet another test of cia script 12:20 < fenn> about fucking time 12:21 < genehacker> unless there's some super neutron mirror 12:23 < fenn> now how did that take an hour i wonder 12:25 < genehacker> huh I'm a bit suspicious now 12:26 < genehacker> calculations of neutron star matter UTS is about 4.4*10^18 newtons/m^2 12:27 < fenn> yep 12:27 < genehacker> they say 10 billion times the strongest steel 12:28 < genehacker> steel has tensile strength of 440 MN/m^2 12:28 < genehacker> still that's pretty strong 12:29 < genehacker> now where's the closest neutron star... 12:29 < fenn> read "dragon's egg" to get a feel for how things work in neutron stars 12:29 < fenn> keep in mind you'll need a mini black hole to keep the neutronium from evaporating 12:34 < genehacker> been wanting to read that 12:34 < genehacker> heh 12:36 < fenn> http://adl.serveftp.org/papers/Robert%20L.%20Forward%20-%20Dragon's Egg.pdf 12:37 < fenn> oops forgot one 12:37 < fenn> http://adl.serveftp.org/papers/Robert%20L.%20Forward%20-%20Dragon's%20Egg.pdf 12:37 < genehacker> bolonkin has some calculations in there about making a version of a human out of AB matter 12:38 < genehacker> IE replace every atom in your body with AB matter counterparts 12:38 < genehacker> and 200 of you could fit in a microbe 12:38 < fenn> http://adl.serveftp.org/papers/Robert%20L.%20Forward%20-%20Starquake.pdf 12:38 < fenn> not quite that small 12:39 < fenn> based on the density of real neutronium it'd be more like 0.1mm 12:39 < fenn> er, 1mm, sorry 12:41 < genehacker> how do you get that? 12:42 < genehacker> is that hot neutronium? 12:42 < fenn> well lets calculate it 12:42 < fenn> google tells me " The density of neutronium is 4 * 10 ^ 17 kg/m^3" 12:44 < genehacker> bolonkin say 3 instead of 4 12:44 < genehacker> oh wait he says it varies 12:45 < fenn> units '(4 * 10 ^ 17 kg/m^3)' '70kg/(0.006mm)^3' = about 1 12:45 < fenn> so i was off by a couple orders of magnitude 12:45 < genehacker> so that's about right 12:45 < genehacker> are you calculating a human by weight? 12:46 < fenn> equivalent mass 12:46 < genehacker> bolonkin does it by molecules 12:46 < fenn> what is a 'molecule' of neutronium 12:47 < genehacker> like a proton or neutron 12:47 < fenn> a proton is a molecule? 12:47 < genehacker> no 12:47 < genehacker> it isn't 12:47 < genehacker> he proposes the tiny human as a thought experiment 12:47 < fenn> i think i've heard enough already 12:48 < genehacker> I think you have 13:18 -!- genehacker [i=genehack@wireless-128-62-88-74.public.utexas.edu] has quit [Read error: 110 (Connection timed out)] 13:31 < kanzure> fenn: I am very confused. 13:31 < kanzure> def my_method(first=1, second=2, third=3): #just print out each value and be done with it 13:31 < kanzure> my_method({'first': 39401}) 13:32 < kanzure> (such that my_method does "first = ", first, "\nsecond = ", second, etc.) 13:32 < kanzure> and that dict is given to the first named parameter ("first"), in other words, first becomes a dict 13:32 < kanzure> so my_method prints out "first = {'first': 39401}" 13:32 < kanzure> so when from_yaml calls "cls(data)", where "data" is a dict and "cls" is a bad way of saying "self", 13:33 < kanzure> I don't understand how the yaml peeps expect that to work 13:33 < kanzure> they are just sending a dict 13:37 < kanzure> guess I shouldn't use to_yaml and from_yaml 13:40 < fenn> you probably shouldnt use to_yaml and from_yaml 13:42 < fenn> cls is not "self" 13:42 < kanzure> oh? 13:43 < fenn> since to/from_yaml is @classmethod, it's the name of the class that's calling it 13:43 < fenn> er, the name of the class of the instance 13:43 < fenn> whereas self is the instance itself 13:43 < kanzure> so, in my example yaml file I have a list of interfaces that a part has 13:44 < kanzure> each interface should have a generic name 13:44 < kanzure> so I thought that calling one of them 'hole' would be ok. (so the line is "hole: !interface") 13:44 < kanzure> the yamlauthor still has to add a 'name' attribute though right? 13:44 < kanzure> or is there a way around this 13:45 < fenn> because you can't access the key "hole" right? 13:45 < kanzure> right 13:45 < fenn> i would really like to figure out how to do that 13:45 < kanzure> or because we don't know that the key is necessary "hole" 13:45 < kanzure> *necessarily 13:45 < kanzure> I guess you can look for keys that are of the correct type 13:45 < fenn> the key could be anything 13:45 < kanzure> no, I mean, er.. 13:45 < fenn> the key isnt the type, it's the type of the value it points to 13:46 < kanzure> when you have a dict that yaml spits at you from yaml.load(foo), 13:46 < kanzure> mydict = yaml.load(foo) 13:46 < kanzure> for each in mydict.keys(): 13:46 < kanzure> if type(each) == skdb.pymates.Interface: ... 13:46 < kanzure> er. wait. 13:47 < kanzure> yeah ok I think I see what you mean 13:48 < fenn> yaml doesnt give you a dict if you use a custom tag ( unless your object inherits from dict, and even then i think it just sets everything as attributes) 13:48 < kanzure> bbl 13:49 < fenn> i was reading about the 'property' keyword; apparently you can have property access via dict syntax? 13:50 < fenn> http://adam.gomaa.us/blog/2008/aug/11/the-python-property-builtin/ 13:50 < fenn> Or we could make a .url property with dictionary access: book.urls['absolute'] 13:50 < fenn> Or heck, even with attribute access book.urls.absolute 13:50 < fenn> or maybe i'm totally missing the point 13:51 < fenn> anyway i want to use the property function for stuff like Screw.breaking_strength 13:52 < fenn> sometimes you have an actual measured value, and other times you have to estimate/calculate based on possibly crappy data (so its best to have both options instead of always forcing you to calculate) 13:53 < fenn> what's the python syntax for the parent of an attribute? like for foo.bar, bar.__something__ == 'foo' 13:53 < fenn> or just foo, not 'foo' 14:01 -!- samrose [n=samrose@c-24-11-214-181.hsd1.mi.comcast.net] has quit ["Ex-Chat"] 14:43 < fenn> damn. BRepPrimAPI_MakeSweep isnt implemented yet 14:45 < kanzure> fenn: dash in #python claims that your foo.bar.__something__==foo is impossible in python 14:45 < fenn> that's lame 14:45 < kanzure> why not just get a list of attributes and look for those that are of type 'foo' 14:46 < kanzure> oh, you're asking for a '__parent__' or something 14:46 < fenn> because i dont have foo to begin with, only bar 14:46 < fenn> yeah 14:46 < kanzure> every bar necessarily creates a new foo? 14:47 < fenn> no. what? 14:47 < kanzure> just got off the phone with someone who wanted to touch bases with me before he presents at DEFCON in a few weeks. huh. 14:47 < fenn> now i forget what i was doing. thanks~~ 14:47 < kanzure> hah 14:49 < kanzure> for each foo, for each bar in foo, set bar.__parent__ to its foo. 14:49 < fenn> i told you i dont have foo to begin with 14:51 < drazak> fenn: if it's not too much trouble, could you email that to me? 14:51 < kanzure> the zip file? 14:51 < kanzure> wtf 14:51 < fenn> can i get a monitor with lots more than 60fps? 14:51 < fenn> and if so, will any graphics card support it? 14:52 < fenn> graphics cards claim to render at 300fps but if your monitor is only running at 60 then what's the point 14:53 < fenn> drazak: i object on general principle 14:56 < kanzure> I still don't see why you can't loop through the key,value pairs in the dict 14:56 < fenn> there is no dict 14:56 < kanzure> that's how yaml reads it 14:56 < fenn> there's a class with attributes 14:56 < fenn> not if you have a custom tag 14:56 < kanzure> fooey 14:58 < kanzure> wait, why is that? 14:58 < fenn> because tags describe the data type 14:59 < kanzure> could you join #yaml? 14:59 < fenn> and the way i'm doing it, the tags come from yaml.YAMLObject 15:06 < kanzure> the hacky way to do it would be to have something like "hole: !interface\n name: &*" where &* is a special identifier referring to the parent name. but this would be auto-inserted and not written by the user. kind of stupid though. 15:19 < kanzure> "too many values to unpack" 15:19 < kanzure> at the for loop 15:19 < fenn> you might as well just s/((.*): !interface)/\1\n name: \2/ 15:20 < fenn> please dont actually use sed, that's a terrible hack 15:21 < kanzure> aren't we all. 15:22 < fenn> speak for yourself 15:23 < fenn> i wonder if i can convert a solid to a shell and _then_ sweep it 15:23 < kanzure> are shells implemented? 15:23 < fenn> yes, basically shell == surface 15:23 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has joined #hplusroadmap 15:24 < fenn> if you sweep a wire you get a shell 15:24 < fenn> somehow this is different from sweeping an edge to get a face.. dont ask me 15:25 < fenn> http://adl.serveftp.org/lab/fenn/pngs/prism_not_sweep.png shows why i need MakeSweep 15:26 < fenn> sort of confusing since there's two overlapping objects 15:34 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has quit [Read error: 60 (Operation timed out)] 15:38 < CIA-73> skdb: kanzure master * r4c502fd / (pymates/models/blockhole.yaml pymates/pymates.py): pleasant way to add interfaces - http://bit.ly/NaUQJ 15:39 < kanzure> fenn: why the fuck is hall-heroult.process in / ? 15:39 < fenn> because it didnt belong in inventory and i was making silly little changes to test the cia script 15:39 < kanzure> what about change.txt ? 15:39 < kanzure> may I delete change.txt ? 15:39 < fenn> same 15:39 < fenn> yes 15:40 < CIA-73> skdb: kanzure * r4c502fd12b03 /pymates/ (models/blockhole.yaml pymates.py): pleasant way to add interfaces 15:40 < CIA-73> skdb: kanzure * r4fa0eacad95e / (4 files in 2 dirs): Merge branch 'master' of ssh://bryan@adl.serveftp.org/var/www/skdb 15:40 < fenn> now i cant figure out how i made a shell :\ 15:40 < CIA-73> skdb: kanzure * rd7b1e5873ab8 /change.txt: clean up yer messes, fenn. 15:40 < kanzure> yay CIA is smart enough to not send duplicates 15:41 < fenn> eh? 15:41 < kanzure> I committed to adl and github 15:41 < kanzure> both have the CIA hooks 15:41 < kanzure> CIA only mentioned the commit once 15:41 < fenn> adl has post-receive only 15:41 < kanzure> so? 15:41 < kanzure> oh 15:41 < kanzure> yeah, so? 15:41 < fenn> and it did say the same commit twice 15:41 < kanzure> no it didn't 15:42 < kanzure> it just said: r4c502fd12b03, then r4fa0eacad95e, then rd7b1e5873ab8 15:42 < fenn> r4c502fd showed up twice, maybe that's not what you meant? 15:42 < kanzure> oh crud it did 15:42 < kanzure> fuck.. 15:43 < kanzure> what's the point of that? 15:43 < fenn> i'm not sure what the point of having two "main" repositories is 15:44 < kanzure> github shows pretty sources 15:44 < kanzure> but now since we have the hook, I'll go disable the hook 15:44 < kanzure> on github i mena 15:44 < kanzure> *mean 15:45 < fenn> who the fuck brings crying babies to the engineering building 15:45 < kanzure> maybe the little kid wanted to see the time machines 15:45 < fenn> move along, nothing to see here 15:46 < kanzure> http://www.singularitysummit.com/ 15:47 < kanzure> nice to see Ed Boyden there. wonder what he's been up to lately. 15:51 < fenn> huh it worked this time 15:53 < fenn> http://adl.serveftp.org/lab/fenn/pngs/prism_not_sweep2.png 15:54 < kanzure> I like it when things work. 15:55 < fenn> i like it when they work the first time 15:55 < fenn> and also the second time 15:55 < fenn> one or the other is not so good 15:55 < kanzure> but the third. who cares about the third. she's a bitch. 15:55 < fenn> poor third 15:55 * fenn is reminded of Armitage 15:56 < fenn> or battlestar galactica.. same thing i guess 15:58 < fenn> any word from jata? 15:58 < kanzure> rescheduled to next week. but she hasn't bought the tickets yet either. 15:58 < kanzure> maybe I'll call Alex soon 15:59 < kanzure> I haven't actually explained to them the lease expiration date etc 15:59 < kanzure> why wouldn't she show up in here before flying us out there. she's kind of lazy. 16:00 -!- mage2 [n=mage@66.179.208.36] has quit [Remote closed the connection] 16:00 < fenn> weirdness is a given 16:02 < kanzure> hm. so. should I do the matrix math stuff now, or the pythonOCC visualization stuff? 16:04 < kanzure> can't determine the variational priorities here. 16:05 < fenn> i'm afraid i dont see the difference 16:05 < fenn> matrix math stuff = translations and rotations in OCC, no? 16:05 < kanzure> with visualization I could just say "look! it loads a CAD model!" 16:05 < kanzure> oh. hrm. 16:05 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has joined #hplusroadmap 16:05 < kanzure> yeah I guess so 16:05 < kanzure> I guess it's also easier to say "hey look! an image thingy!" 16:06 -!- splicer [n=patrik@h189n3c1o261.bredband.skanova.com] has joined #hplusroadmap 16:06 < kanzure> hello splicer 16:08 < splicer> hi kanzure 16:16 < CIA-73> skdb: kanzure * r964eec078000 /pymates/ (models/blockhole.yaml pymates.py): added coordinates support 16:29 < genehacker> are you messing around with cad stuff? 16:29 < kanzure> yes 16:29 < kanzure> http://adl.serveftp.org/lab/fenn/pngs/prism_not_sweep2.png 16:29 < genehacker> are you messing around with cad stuff? 16:29 < kanzure> http://adl.serveftp.org/lab/fenn/pngs/prism_not_sweep2.png 16:29 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has quit [] 16:29 < kanzure> pft. 16:30 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has joined #hplusroadmap 16:30 < genehacker> are you messing around with cad stuff? 16:31 < genehacker> if you can extrude stuff from a 2d sketch I'll write you a script for making really really weird gears 16:32 -!- Smari [n=spm@89-160-140-46.du.xdsl.is] has joined #hplusroadmap 16:34 -!- Smari [n=spm@89-160-140-46.du.xdsl.is] has quit [Client Quit] 16:34 -!- Smari [n=spm@89-160-140-46.du.xdsl.is] has joined #hplusroadmap 16:34 < kanzure> genehacker: yes 16:34 < kanzure> http://adl.serveftp.org/lab/fenn/pngs/prism_not_sweep2.png 16:34 < kanzure> you can extrude stuff from a sketch, yes. 16:34 < kanzure> genehacker: you can install pythonOCC on Windows if you want 16:35 < kanzure> genehacker: here's how: http://www.pythonocc.org/wiki/index.php/InstallWindows 16:35 < Smari> material properties? Does anybody have a definitive non-redundant list? 16:35 < kanzure> ha, ha. no. :( 16:35 < kanzure> Smari: I have a rip of matweb if that's what you're looking for 16:35 < Smari> kanzure, matweb? 16:36 < kanzure> Smari: http://matweb.com/ 16:36 < kanzure> has about 80,000 materials listed 16:37 < kanzure> it's 479 MB in size: http://heybryan.org/books/Manufacturing/matweb.zip 16:37 < genehacker> so I have a matlab program that makes gears 16:37 < genehacker> gears like you've never seen before 16:38 < genehacker> btw does matweb have mooney constants for RTV silicone? 16:38 < kanzure> it only had the most generic of information 16:39 < kanzure> so I'd doubt it 16:39 < genehacker> dang 16:39 < kanzure> you should check 16:39 < genehacker> just did 16:39 < Smari> kanzure, nice. 16:39 < Smari> kanzure, is that just the database? 16:39 < kanzure> no. that's all of their HTML bullshit. sorry. 16:39 < genehacker> trying to get this hyperelastic modeling thing to work 16:39 < Smari> kanzure, parse.. 16:39 < Smari> parse... 16:39 < kanzure> bah 16:40 < kanzure> I haven't got around to it yet 16:40 < kanzure> I'm not convinced of the general utility 16:40 < genehacker> it's been punching away at my mesh for 8 minutes 16:40 < genehacker> I wonder if it crashed 16:40 < kanzure> Smari: but if you bug me enough I'll get around to it 16:40 < genehacker> or I did something wrong 16:40 < kanzure> do you really think you'd have a use for it? 16:41 < genehacker> oops 16:42 < Smari> kanzure, no. But it would be good to have that kind of dataset parsed nicely anyway. 16:42 < kanzure> right 16:43 < Smari> It could be culled quite a bit. 16:44 < kanzure> did you just wake up? 16:44 < Smari> no 16:45 < Smari> just got off a boat. 16:52 < Smari> Mmm 16:52 < Smari> Properties properties properties 16:53 < Smari> So what I'm doing is generating the model/DTD stuff for a "material".. 16:54 < Smari> what is CTE? 16:56 < genehacker> FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF 16:56 < genehacker> this thing takes a long time to run 16:56 < Smari> Ah 16:56 < Smari> That. 16:56 < genehacker> just trying to simulate a part 16:57 < genehacker> it's gonna take awhile 16:58 < genehacker> it might even take days 17:00 -!- tarbo2_ [n=me@unaffiliated/tarbo] has joined #hplusroadmap 17:05 -!- tarbo2 [n=me@unaffiliated/tarbo] has quit [Read error: 110 (Connection timed out)] 17:14 -!- genehacker [i=genehack@wireless-128-62-165-165.public.utexas.edu] has quit [Read error: 60 (Operation timed out)] 17:19 * kanzure compiles pythonOCC after making a correction in environment.py 17:21 * Smari is trying to write code while experiencing a live performance of "Dánarfregnir og Jarðarfarir" on vibraphone. 17:21 < kanzure> vibraphone? 17:22 < Smari> it's an instrument. 17:22 < Smari> It's similar to a xylophone I guess. 17:22 < kanzure> if I was going to go outside to eat tonight, where would I eat 17:22 < Smari> The park! 17:23 < Smari> The moon! 17:23 < Smari> The freeway! 17:23 < Smari> The airport! 17:23 < Smari> Subway? 17:27 -!- genehacker [n=noko@cpe-67-9-157-136.austin.res.rr.com] has joined #hplusroadmap 17:39 < kanzure> TopoOpeBRepDS_tools.hxx: No such file or directory 17:39 < kanzure> blargh 17:40 < Smari> You said you had a GNU Units wrapper? 17:40 < kanzure> right 17:40 < kanzure> it's in skdb.py in the skdb git repo 17:40 < kanzure> try class Units and class Measurement etc. 17:42 < kanzure> aha the source of fleas was the garbage disposal 17:43 < Smari> I've got a fairly substantial list of properties.. 17:43 < kanzure> hm? 17:43 < kanzure> start from the beginning please 17:43 < Smari> Wondering whether I should add them all individually to the DTD or create a "subclass" for it.. 17:43 < kanzure> a list of potential properties for a material in a material database? 17:43 < Smari> mechanical properties. 17:43 < Smari> yes 17:43 < kanzure> well this is why yaml is so fun 17:44 < Smari> mechanical, optical,.. 17:44 < kanzure> you don't have to make a DTD with yaml. 17:44 < kanzure> it just loads in the attributes that you specify 17:44 < Smari> No I WANT to make a DTD. 17:44 < kanzure> O.o 17:44 < Smari> Because having a DTD means I have an explicitly defined protocol that can be sanity checked.. 17:44 < Smari> and I can autogenerate databases and stuff from... 17:45 < kanzure> yeah I hear that last one I guess 17:45 < kanzure> so, hey, 17:45 < kanzure> I was looking for a tool to automatically generate DTDs given an XML document 17:45 < Smari> YAML means that I end up with a gazillion edgecases unless it just happens to have been the same autistic typist who wrote all the files. 17:45 < kanzure> where you get to assume that the XML document uses everything 17:45 < kanzure> but I've never seen anything that does this 17:46 < kanzure> it seems like an obvious thingy to me 17:46 < kanzure> but I can't find it. you know of anything like this? 17:46 < kanzure> autogendtd example.xml > example.dtd 17:46 < Smari> yeah.. should be fairly easy: Traverse the tree and write down every tag and property you encounter. 17:46 < Smari> Never seen anything like that though 17:47 < kanzure> do you have a list of properties already in a human readable format? 17:47 < kanzure> and what are you going to use this dataset in? 17:49 < Smari> I'm converging the map frontend with your SKDB purposes. 17:49 < Smari> so I'm making a few things: 17:50 < Smari> 1) XML DTD's for the core stuff 2) Django Model definitions (which are essentially SQL tables) 3) XMLRPC and JSON interfaces to the Django stuff. 17:50 < kanzure> so your map would be a django model? 17:50 < kanzure> is there a generic Google Maps AJAX voodoo magic plugin for django yet? this seems like something taht should exist anywway 17:51 < kanzure> *anyway 17:51 < kanzure> sorry, *django view (not a model) 17:51 < kanzure> although I'm sure it will have a few models 17:52 < Smari> kanzure, I've not seen many pick-up-and-play Django apps.. 17:52 < Smari> kanzure, but I'm using OpenStreetMap in this. 17:52 < kanzure> but wasn't that the idea 17:52 < Smari> My map is a Django model yes. 17:53 < Smari> well, several models and a few views. 17:56 < Smari> Pull the repo. 17:56 < Smari> i.e. tangiblebit 17:58 < kanzure> blah, I need to start keeping track of addresses 17:59 < kanzure> aha git pull worked 17:59 < kanzure> you deleted the store? 18:04 < Smari> yes 18:04 < Smari> Moved it elsewhere. 18:05 < Smari> Want to keep this clean for this project. 18:05 < kanzure> Anything I should look at in particular? 18:06 < Smari> doc/* sources/fabmap/xmlrpc.py sources/fabmap/models.py 18:23 < Smari> Like where this is going? 18:24 < kanzure> eternal bash history: http://www.debian-administration.org/articles/543 18:25 < drazak> oh that'd drive me nuts 18:26 < Smari> ctrl+r would become dreadfully slow I suspect. 18:26 < kanzure> oh just running `history 2` is nice 18:28 < kanzure> history -w ~/remember/my_history 18:28 < kanzure> yay 18:28 < kanzure> it's like a christmas wish 18:39 -!- QuantumG [n=qg@nuclear.biodome.org] has quit [Remote closed the connection] 18:39 -!- Jata [n=chatzill@cpe-76-173-244-243.socal.res.rr.com] has joined #hplusroadmap 18:40 < kanzure> hello Jata 18:40 < Jata> hello bryan 18:40 < Jata> i gave in to peer pressure, it seems 18:41 < kanzure> don't do irc, it will ruin your life 18:41 < kanzure> so maybe you could reask the question 18:41 < Jata> which? 18:41 < Smari> IRC skemir heylan. 18:42 < Smari> oh wait, that joke's not funny in this crowd. 18:42 < Smari> sorry. 18:42 < kanzure> Jata: in particular whether or not fenn and i are worth it 18:42 < Jata> well then 18:42 < Jata> are you and fenn worth it? 18:42 < Smari> worth what exactly? 18:42 < fenn> that's what i was going to say 18:43 < kanzure> what do you want to know though? 18:43 < Jata> if you guys fly out here to chat with alex and myself, what are we going to discuss? what magical breakthroughs are we going to pretend to make, and what is alex going to get out of it? 18:44 < Jata> is the essence of it 18:44 < fenn> i want to know what you're pretending to do in the first place, if we aren't there 18:44 < Jata> a great question 18:44 < kanzure> well actually the real reason would be to see if we can actually work together in person 18:44 < Jata> fair enough 18:44 < kanzure> there's really no other legitimate excuse for it 18:44 < kanzure> well, besides my skepticism of alex being real 18:45 < Jata> yeah, this is my thought as well. but it's not me buying the tickets... hah. i have my suspicious sometimes too... 18:45 < fenn> all you zombies 18:45 < Jata> so, i assume you're familiar with h+, formerly wta 18:45 < Jata> yes? 18:45 < fenn> yes 18:45 < Jata> mkay. so, alex is the new exec dir of h+ 18:46 < fenn> does alex want to do anything beyond fundraising and media bullshit? 18:46 < Jata> he wants h+ to stop doing nothing, and he hired me to fix that 18:46 < Jata> well, that's the thing. i don't know what he wants to do, i'm figuring out what it should do 18:46 < fenn> i see he's interested in wearables (to put it mildly) 18:46 < Jata> hah, yeah 18:46 < Jata> have you spoken to him? 18:46 < fenn> no 18:47 < Jata> well, you will 18:47 < kanzure> so basically the task is to figure out what they should have been doing 20 years ago 18:47 < kanzure> or how to catch up 18:47 < kanzure> and then to do it 18:47 < fenn> but that doesnt help us 18:47 < kanzure> or do it then figure it out (the other way around) 18:47 < Jata> yes, and yes it could help us 18:47 < fenn> we should be figuring out what to do _now_ 18:47 < Jata> that is the idea 18:47 < Smari> I'm tempted to say something along the lines of, "the people you really want to be calling to a meeting, then, are..." and then give you a list of five people who have more or less plotted out what is going to have to happen in order for humanity to survive the next fifty years. But I'm not going to. Unless specifically asked. 18:48 < kanzure> does jata know smari? 18:48 < Jata> well, these people have been shouting at the top of their lungs about it 18:48 < Jata> and no, i don't think so 18:48 < Smari> I don't think so either. 18:48 < fenn> Smari: five people would never agree on anything 18:48 < kanzure> there's one 18:48 < Smari> fenn, depends on which five people. 18:48 < kanzure> Jata: so, 18:49 < Smari> fenn, they needn't agree on everything. Just more than most. 18:49 * fenn mumbles something about hive mind 18:49 < Jata> hah 18:49 < Jata> well, here's the summary of what i would like to see happen with h+: 18:49 < kanzure> Jata: so, what is it that you're already doing? 18:49 < kanzure> I've told you about skdb and all of these exciting things happening from my end 18:49 < kanzure> or not happening but that need to happen more 18:49 < Jata> hah 18:49 < Jata> fair enough 18:49 < Smari> they're certainly happening at my end ;) 18:49 < fenn> skdb is coming along 18:49 < Smari> sorry. 18:50 < kanzure> smari's ass is being burned and charred as we speak 18:50 < Smari> I'll go stand in the corner. 18:50 < Jata> glad to hear it 18:50 * fenn waits patiently for the summary 18:51 < Jata> so: at the moment, we're building a diybio lab, writing lab manuals, talking to people who want to help us build the open source diagnostic chips, etc. pretty straightforward, small scale 18:52 < Jata> i would like h+ to actually be useful to crazy people like you, and other small communities of... creative science 18:52 < kanzure> it sounds like alex wanted to be the center of his "communities of creative science" 18:52 < kanzure> and building a one-off diybio lab isn't going to cut it (sorry) 18:52 < Jata> yeah, it does, but that's sort of beside the point 18:52 < kanzure> (of course, like any good salesman, I have the solution no less) 18:52 < Jata> obvi, bryan 18:53 < kanzure> :p 18:53 < Jata> haha, okay. let's hear it 18:53 < kanzure> er, I thought I already told you about skdb 18:53 < Jata> you did 18:53 < kanzure> apparently not well enough 18:53 < kanzure> there's no reason that diybio labs can't be built DIY-style 18:54 < kanzure> the idea of skdb is to package that information and the instructions 18:54 < Jata> how long will it take you to actually make skdb happen? 18:54 < kanzure> such as instructions for both procedures and building tools 18:54 < kanzure> well it's coming along 18:54 < Jata> how long will it be before i can use it? 18:54 < kanzure> what does that mean? 18:54 < Jata> my point is this: 18:54 < kanzure> there are some development milestones we are approaching, if that's what you're asking 18:54 < kanzure> for instance, by monday I plan to have the part mating algorithms done 18:54 < fenn> it turns out that everything takes longer than you think 18:55 < Jata> exactly 18:55 < Smari> Jata, documentation and starter packages. 18:55 < fenn> so i expect part mating algorihms done in (3 * 2 to the next unit) 6 weeks 18:55 < kanzure> Jata: skdb is kind of the central project to all the clutter.. but I still think pursuing other projects is appropriate. 18:55 < fenn> it's not actually central 18:56 < fenn> it's the glue that holds everything together 18:56 < Jata> for me, it's not about whatever silly diybio projects i do right now. there are a lot of really cool projects being worked on. they always take longer than you think. they are expensive. they are time consuming. what would help you? 18:56 < Smari> It permeates through the æther of Things to Be Done. 18:56 < kanzure> so, instructables and all of these other websites just take pictures of hardware 18:56 < fenn> honestly i think diybio is a waste of time 18:56 < kanzure> and this isn't sufficient for advancing diy and transhumanism 18:56 < kanzure> a photograph isn't sufficient. we need to be more serious about this :p 18:56 < fenn> until people have a basic manufacturing and chemical infrastructure, how do you do anything? 18:57 < Jata> agreed. the point of diybio, imho, is to make people pay attention 18:57 < Jata> but that's beside the point 18:57 < Jata> what do you need to make skdb happen? 18:57 < kanzure> it's already happening, but slowly because we waste our time at the lab 18:57 < fenn> time and smart people 18:57 < fenn> and minions, lots of them 18:57 < kanzure> oh yes definitely minions 18:58 -!- QuantumG [n=qg@nuclear.biodome.org] has joined #hplusroadmap 18:58 < Jata> if you could suddenly afford these minions and smart people, would you be able to get results? 18:58 < fenn> but i expect that to happen on its own eventually 18:58 < Smari> I'm fed up of standing in the corner. I'm piping in. Look: Transhumanism is facing a situation where there will be no transhumanism if the current running model of humans doesn't solve a bunch of very very big problems right now and no later than now. These questions all more or less boil down to infrastructure. 18:58 < fenn> Jata: that's sort of hard to answer.. how would i know? 18:59 < kanzure> I can't point to a website and say, "look! it's spitting out instructions right now!" 18:59 < fenn> Jata: i'm supposed to answer yes, and you're supposed to believe me, because you're a primate and if i believe in myself then it must be right, right? 18:59 < Jata> thank you, smari. you're allowed to pipe in, and yes, i agree. it's very frustrating. what would you see done? 18:59 < QuantumG> Smari: oh do tell, what problems would they be? 18:59 < Jata> fenn, you aren't supposed to answer 18:59 < Smari> Fab Labs and DIYBio Labs and all of that stuff are manufacturing infrastructure, and unless we have those and all of that jazz in every single populated area on earth, we're going to need a distribution system for manufactured goods. 18:59 < fenn> huh. 18:59 < Jata> yet.. 18:59 < kanzure> yes, I've been kind of yelling at jata about infrastructure issues 19:00 < kanzure> deployable downloadable infrastructure is kind of important glue. 19:00 < fenn> we already have a distribution system for manufactured goods 19:00 < kanzure> it kind of sucks 19:00 < fenn> (well in most populated areas) 19:00 < Jata> fenn, you and bryan are going to come to LA (next weekend?) and hit alex over the head with the importance of infrastructure. then we are going to figure out what we can do to help you. this is one of the major points of the meeting. 19:01 < Jata> the other point is: as smari pointed out, transhumanism is in a pretty bad state 19:01 < kanzure> heh 19:01 < kanzure> yes, it is 19:01 < Jata> this is annoying to me. we need to fix it. 19:01 < Jata> because the old school transhumanists aren't really helping much 19:01 < kanzure> so while I would like to hit alex over the head about infrastructure, 19:01 < kanzure> I think that alex has the right idea 19:01 < QuantumG> we all'd like to hear the plan 19:01 < kanzure> the other day he said something about "ok jata, what do you need to make this happen" 19:01 < fenn> part of the problem is it's a pseudo religious -ism instead of a political -ism 19:01 < fenn> religions don't have agendas 19:01 < Jata> yeah, this is a huge part of the problem 19:02 < Jata> this needs to all go away 19:02 < kanzure> yay. that makes me happy. I've been bashing my head against these guys for years now. something has changed. :) 19:02 < Jata> i'm currently rewriting all the h+ docs, redoing the website, revamping the thing 19:02 < Jata> hahah 19:02 < kanzure> yeah have you poked the new FAQ thing? 19:02 < Jata> any input any of you have is welcome 19:02 < Jata> which one? 19:02 < kanzure> the huge one 19:02 < Jata> there are like 4 19:02 < Jata> oh, god, it's a disaster. i've been tweaking it 19:02 < kanzure> the monster that shouldn't be touched with a four foot pole 19:03 < Smari> QuantumG, In short 2 billion people living extremely unsustainably, to the degree of possibly irrevertably fucking up the planet and, if not, causing a permanent materials disequilibrium. Another 4 billion, soon to be 5 billion, living more or less sustainably but are largely miserable and want to become part of a future 6 billion person suburbia, where every human being has an SUV and goes to the Canary Islands for Christmas. 19:03 < fenn> "the super FAQ" typical kanzure textsplosion 19:03 < kanzure> hey now.. 19:03 < Jata> smari: energy crisis won't be an issue 19:03 < Jata> but that's irrelevant 19:03 < Smari> Jata, I keep hearing people say that but I haven't seen anybody solve that on the large scale. 19:04 < QuantumG> Smari: and how does this lead to extinction? 19:04 < fenn> Jata: why did you agree (or think you agreed) with smari earlier? 19:04 < Smari> Jata, although I do agree. 19:04 < kanzure> Jata: so, how would alex justify h+ funding, say, an h+ diybio fablab fabratory skdb thingy 19:04 < Jata> about h+? 19:04 < kanzure> as in, "ok wta, we're going to do this now because I think it's cool" 19:04 < kanzure> yeah, I'm just wondering 19:04 < fenn> about planet earth being doomed 19:04 < kanzure> you can't just snap your fingers and get everyone convinced 19:04 < Jata> yeah, that's pretty much it 19:04 < kanzure> er, me? 19:04 < Jata> i don't think planet earth is going to be doomed. i think our transhumanist hopes coul be doomed 19:04 < Jata> yes you, bryan 19:05 < Smari> QuantumG, not necessarily extinction, but do you not agree that for highly philosophical concepts such as transhumanism to exist, people need to have a moderate degree of comfort - you know, not be want of food, shelter, etc, etc.? 19:05 < fenn> why do we need to convince everyone? 19:05 < kanzure> Jata: why should h+ fund *our* particular transhumanist hopes? 19:05 < kanzure> fenn: because I don't know why it should be us rather than someone else 19:05 < kanzure> I guess maybe there is nobody else 19:05 < kanzure> maybe that's why.. 19:05 < fenn> oh. minions, right. 19:06 < Jata> yes 19:06 < kanzure> it's not like h+ members get "free body upgrades" as part of their membership 19:06 < QuantumG> if you want to predict a war or a famine or a dark ages, sure, I could buy that.. with a nice grain of salt.. but when you start talking about there being "no more humans" I gotta chuckle 19:06 < Jata> ...yet 19:06 < Jata> this is exactly the problem with h+ 19:06 < QuantumG> now, if you were talking about killer asteroids... 19:06 < Smari> Here's my point: I want to live for another 1000 years, minimum, but as things are going right now I'm not sure I'm going to survive the next 20 years. 19:06 < Jata> arguing over whether we're all going to die... not useful 19:06 < kanzure> Jata: so getting the tools to the transhumanists is important, like educating them and so on- and servicing the tools etc.- 19:07 < kanzure> maybe that's sufficiently legit 19:07 < Jata> yes. most important is making the technology actually happen 19:07 < Jata> skdb, if successful, would be damned awesome 19:07 < Smari> QuantumG, there will be humans. I'm not worried about that. What I'm worried about is that the humans won't be in any position to worry about genetic modifications or the theoretical underpinnings of the universe because we'll be too busy trying to survive. 19:07 < kanzure> I talked with someone on the phone about DEFCON today and mentioned to him that I think of transhumanism in terms of .. being a systems administrator of the brain, sort of. 19:07 < Jata> i would like to live in a world where skdb exists; therefore, i will do what i can to ensure that you are capable of making it happen, however you choose to make it. it's that simple 19:08 < Jata> oh, is anyone going to defcon? we could meet 19:08 < Smari> I wish 19:08 < kanzure> Jata: that sounds great, but I'm afraid I may have spoken myself into a hole- for there is a great deal more that can be done simultaneously while the glue is drying 19:08 < Smari> I wish I could be at HAR too. 19:08 < Smari> I'll be at FAB5 though 19:08 < QuantumG> Smari: so in under 5 minutes you've changed your argument.. no wonder you're not gunna live forever 19:08 < Jata> bryan, what? 19:09 < kanzure> Jata: fenn made an analogy of skdb as infrastructure glue 19:09 < Jata> i got that part 19:09 < Smari> QuantumG, no, my argument hasn't changed at all. You've just not been following my point. 19:09 < QuantumG> no.. you've just failed to make one 19:09 < kanzure> Jata: so while the glue is drying there are other interesting things that would be kickass to work on 19:09 < Smari> *sigh* 19:09 < Jata> yes. 19:09 < kanzure> Jata: for instance, alex indicated interest in the rapamycin experiments 19:09 < QuantumG> that happens if you don't think before you talk 19:09 < kanzure> although I think there might be some other interesting targets 19:10 < Jata> yes! adorable mice. 19:10 < fenn> how about wearable computers that dont suck 19:10 < Jata> there are zillions of topics 19:10 < Jata> fenn, i like you 19:10 < kanzure> like how todd might be funding the myostatin inhibitor project too 19:10 < Jata> all of these things should be worked on 19:10 < kanzure> etc. "millions" 19:10 < kanzure> right 19:10 < kanzure> so we just need to make it happen 19:10 < Jata> exactly 19:10 < QuantumG> how about direct neural interface research that don't suck 19:10 < kanzure> QuantumG: have you seen innerspacefoundation.org ? 19:10 < Jata> how about a lot of things that dont suck? 19:10 < kanzure> heh 19:11 < fenn> hum. the difference is i think i actually know how to make a decent wearable computer in a reasonable timeframe 19:11 < Jata> how about -- and i dont know who is in here, so i don't know who i'm inviting -- but why don't interested people with interesting projects get themselves on some hypothetical list 19:11 < fenn> but 'neural interface' could mean 50 things 19:11 < kanzure> yet another list? 19:12 < Jata> not a mailing list 19:12 < Jata> necesarily 19:12 < fenn> a hit list 19:12 < Jata> precisely. 19:12 < kanzure> let's just get it done.. submit a patch. 19:12 < Jata> okay, here's the deal: h+ might be able to fund your independent research. if you want in, you need to make that known 19:13 < kanzure> make that known how? 19:13 < Jata> however possible. my email is parijata@gmail.com 19:13 < QuantumG> well, making just a battery powered computer with vga out, usb in, and not much else is a challenge.. as I haven't found anywhere you can buy one. Last wearable I built I used an eeepc 19:13 < fenn> are you asking for a grant proposal? 19:13 < Jata> unofficially, yes 19:13 < kanzure> I don't know what's happening 19:13 < Jata> okay guys 19:14 < Jata> this is why i was reluctant to jump into this channel 19:14 < Jata> i'm telling you what's in progress 19:14 < Jata> i need your input, because it's not happening yet 19:14 < Jata> i have a lot of resources; i need to direct them 19:14 < fenn> QuantumG: i'm using beagleboard 19:14 < QuantumG> there are some low power computers you can get with no embedded keyboard/screen/etc.. they're for home media center use.. and go on about how low power they are.. and yet there's no battery systems for them. 19:14 < Jata> the best place to direct them, it seems, is to these sort of projects. especially the ones that could be serious business. 19:15 < fenn> i have pages and pages of projects 19:15 < kanzure> Jata: so the reason why I was asking a lot last week about the itinerary was because you've caught me at a good time 19:15 < Jata> great. bring them. 19:15 < Jata> oh? 19:15 < kanzure> Jata: because if fenn and I were to come up north to meet with you and alex, 19:15 < kanzure> and if something was to click, like if we were to do what we're doing in austin (the fabratory) but up there, 19:15 < kanzure> this would be a good time 19:15 < kanzure> because of leasing contracts 19:16 < kanzure> involving the apartment etc. 19:16 < fenn> (everything's north from texas) 19:16 < Jata> hah, okay 19:16 < kanzure> so it's just good timing 19:16 < Jata> gotcha 19:16 < Jata> is this upcoming weekend still good timing? 19:16 < kanzure> I think that if we had a place to work and implement our infrastructure ideas things could go better 19:16 < kanzure> yes 19:16 < Jata> would you like a place to work? 19:16 < fenn> four letter word 19:16 < Jata> so is play... 19:16 < fenn> i guess play is also a four letter word 19:17 < kanzure> fenn has this strange dream of living and playing and building in a giant warehouse of sort 19:17 < fenn> gah 19:17 < Jata> well, aren't you in luck 19:17 < kanzure> stop reading each others' minds, damn it 19:17 < kanzure> that's off-topic and not allowed 19:17 < Jata> i have a giant warehouse that i was hoping you guys would be interested in using 19:17 < kanzure> we need a place to stash machines, tools, materials 19:17 < kanzure> so that we could say one morning "I need to make xyz" 19:17 < kanzure> and then go make it 19:17 < kanzure> without fussing about where the hell to get shit 19:17 < Jata> i didnt mention it because i was trying to figure out some clever way to talk you away from texas 19:17 < fenn> that sort of thing takes a while to build up enough junk 19:17 < Jata> but yes. if you want a warehouse in LA, it's yours 19:18 < kanzure> an inventory system would also be nice 19:18 < kanzure> Jata: you don't need to be clever to bring me to the coasts 19:18 < Jata> anyway 19:18 < Jata> hahaha 19:18 < kanzure> austin is a bad idea for many reasons 19:18 < kanzure> why am I here 19:18 < kanzure> who are you people 19:18 < QuantumG> fenn: thing the beagleboard could produce output for this: http://www.vuzix.com/iwear/products_wrap920av.html it's not vga 19:18 < kanzure> fenn just came down here because I told him about the fabratory project 19:19 < QuantumG> I have the VR920 which is vga. 19:19 < fenn> QuantumG: MyVu crystal is VGA and costs about $200; you get two displays that can be used separately 19:19 < kanzure> Jata: did you ever see the inventory lists in the skdb repo? they were sort of like wish lists 19:19 < Jata> okay, well, fenn, bryan: the purpose of your trip is to tell alex about the importance of infrastructure, about skdb, and to decide what you actually want/need for your work. we will also discuss h+ and where we'd like to see that go. 19:19 < kanzure> of stuff that would be useful to have or to eventually build 19:19 < Jata> yeah, i figured 19:19 < kanzure> Jata: okay, great. if you happen to have h+ers wanting to do some serious projects or something, 19:19 < fenn> fabratory fumbled because people were trying to make money off it instead of getting it done 19:19 < kanzure> fenn and I have piles of notes and documents 19:20 < QuantumG> fenn: the wrap 920 is/will be semitransparent 19:20 < Jata> this is great news 19:20 < kanzure> Jata: so it sounds like we have next weekend planned out then? 19:20 < Jata> h+ is a non-profit, we couldn't make money off of you if we wanted to 19:20 < Jata> yessir 19:20 < Jata> let me call alex and make sure it works 19:20 < kanzure> there's actually a way to make money off of skdb, and alex mentioned that he might want to hear it 19:20 < kanzure> but it's more of a consulting gig version of cygnus 19:21 < kanzure> but that doesn't matter at the moment. 19:21 < fenn> what about offering 'buy it now' button service 19:21 < kanzure> yeah that too 19:21 < kanzure> (joseph jackson wants to do "diybio premium" for a cost or something, but he never got around to convincing me) 19:21 < fenn> that doesnt make sense anyway 19:22 < QuantumG> commercial makes it possible to jump in at different parts in the production tree.. not to mention the motivation factor. 19:22 < kanzure> right 19:22 < QuantumG> unfortunately it has negative side-effects too 19:22 < fenn> lots of negative side effects 19:22 < Jata> absolutely. but i dont thikn skdb is there yet... correct me if i'm wrong 19:22 < fenn> no, skdb is a pile of poo 19:23 < kanzure> well the thing is that the parts about ordering stuff 19:23 < QuantumG> fenn: so how are you powering your beagleboard? 19:23 < kanzure> are actually much easier parts 19:23 < fenn> QuantumG: 3.7V lithium batteries 19:23 < kanzure> I mean, anybody can slap up a web interface to a store or something 19:23 < kanzure> srsly 19:23 < Jata> yeah... well, i think we've got some ideas of what we need to discuss. fenn, bryan, unless you need something, i'm ditching this channel 19:23 < kanzure> Jata: glad you showed up 19:23 < QuantumG> fenn: any ideas why that web site doesn't sell packs? 19:23 < fenn> nice chatting 19:23 < Jata> indeed 19:23 < Jata> take care :-) 19:24 -!- Jata [n=chatzill@cpe-76-173-244-243.socal.res.rr.com] has quit ["ChatZilla 0.9.85 [Firefox 3.5/20090624025744]"] 19:24 < kanzure> I bugged her so damn much about showing up in here 19:24 < Smari> haha 19:24 < Smari> well, she came 19:25 < kanzure> Smari: sorry I was ignoring you 19:25 < Smari> no worries. 19:27 < fenn> i often feel like i'm not as smart as i used to be.. i wonder what it means 19:27 < kanzure> I feel that way about me and programming 19:27 < kanzure> I was a much better programmer when I was 12. 19:27 < Smari> So was I. 19:27 < Smari> I cry damnation every day on this account. 19:28 < kanzure> is that before or after the beer 19:28 < Smari> Before. 19:28 < QuantumG> I wasn't a better programmer when I was younger.. but I certainly wrote more code.. I guess its easier to churn out crap. 19:28 < Smari> Sweet alcohol eases the pain ;) 19:29 < fenn> i've noticed that code almost writes itself if i'm sleep deprived 19:29 < fenn> sort of the opposite of what you'd expect 19:30 < fenn> but the price is that you dont get any new ideas 19:30 < QuantumG> hehe.. I watch my cow-orkers check in code when its 2am their time and it is consistently their worst work. 19:33 < fenn> QuantumG: what's the resolution on the wrap 920? 19:35 < QuantumG> presumably it'll be similar to the VR920.. which is twin 640x480.. and some down-scaling so you can support 1024x768 video modes 19:35 < fenn> why did you say 'not vga' then? 19:36 < QuantumG> not a vga connector 19:36 < fenn> oh. what's the connector then? 19:36 < fenn> sorry. myvu is RCA composite 19:36 < QuantumG> the VR920 has a vga connector.. 19:36 < QuantumG> the AV920 is composite 19:37 < QuantumG> and the Wrap 920AV will be composite 19:37 < QuantumG> I do wonder if they'll make a vga version 19:39 < fenn> hmm "please call for pricing" http://www.vuzix.com/iwear/products_taceye_lt.html 19:39 < QuantumG> the Wrap 920AV with the two optional accessories (6DoF head tracking, and USB stereo camera pair) and the PhasAR input device or some other input device is starting to get to a decent augmented reality system 19:40 < QuantumG> yeah.. they have a military line.. I ignore that stuff. It's less painful to just pretend it doesn't exist. 19:41 < fenn> i dont get it.. it's the same shit, why does it cost 10X just to get one display instead of two 19:41 < fenn> reminds me of imitation meat 19:41 < QuantumG> probably because the military is stupid enough to pay? 19:41 < fenn> 100% imitation costs 10x the price of 95% imitation meat 19:43 < QuantumG> Vuzix are, clearly, a struggling hardware company who figures going consumer is the only way to stay in business.. only problem is: they suck at it. 19:43 < QuantumG> I mean, shit 19:43 < QuantumG> make an augmented reality system already 19:43 < QuantumG> a complete solution 19:44 < QuantumG> stick a price tag on it.. it'll be excessive, sure, but if you put a price tag on every part then people can cobble the system together as they can afford it. 19:44 < QuantumG> ultimately, what's missing here is the vision. 19:45 < fenn> right 19:45 < QuantumG> there's no "this is what augmented reality is for" coming from Vuzix.. they just wanna sell widgits 19:45 < fenn> nobody wants to watch DVD's on some weird thing in public 19:45 < QuantumG> someone does 19:45 < fenn> i sort of figured it was for porn or something 19:46 < QuantumG> where public often means "planes" 19:46 < QuantumG> because you aint doing it on the train 19:46 < QuantumG> maybe with the Wrap 920AV you could though 19:46 < fenn> eh? why? 19:46 < QuantumG> because it is transparent 19:46 < kanzure> Smari: you should set up tangiblebit to cia.vc 19:46 < Smari> Why? 19:46 < kanzure> Smari: we have an announcer bot in here (CIA-73) 19:46 < kanzure> well it's so that we know when you commit 19:46 < Smari> Ah. 19:47 < kanzure> so that we can pull and play around with it 19:47 < fenn> QuantumG: you must live in one of those places where people commute by train 19:47 < Smari> yes 19:47 < Smari> Indeed. 19:47 < Smari> I'll do that. 19:47 < QuantumG> fenn: indeed :) 19:47 < QuantumG> I went on a train yesterday 19:47 < kanzure> Smari: we just set up skdb yesterday, so presumably it will help bring some awareness that we're actually doing some work :p 19:47 < QuantumG> there was commuters holding iPods watching tv 19:47 < QuantumG> one of them was watching Enterprise 19:47 < QuantumG> I silently mocked him 19:48 < kanzure> guy from bus stop that fenn and I randomly met: "told my sister about you.....give her a holler and see if she can help you !! world domination dude !!!!!!" - on my facebook wall 19:48 < fenn> just imagine, with pervasive AR you could summon a giant bird to take a dump on him 19:49 < QuantumG> it's kinda funny that their camera optional accessory totally destroys the "almost normal" look of the device 19:49 < QuantumG> without the cameras you might get comments like "hey, nice sun glasses dude" .. with the cameras you'll get people avoiding making eye contact 19:50 < fenn> yeah cameras definitely belong at the sides 19:50 < fenn> otherwise you look like a gecko at best 19:51 < fenn> steve mann had it half right with the 45 degree mirror 19:51 < fenn> (it should be a partially silvered mirror) 19:52 < QuantumG> unfortunately he never did anything more to commercialize it than "call me if you wanna commercialize this and I'll ignore you" 19:52 < fenn> commercial design firms just pump out mroe of the same crap over and over, with different plastic styling 19:52 < kanzure> fenn: and then campbell wants us to "automatically design" with that plastic crap 19:52 < kanzure> wtf. 19:53 < QuantumG> fenn: well this Vuzix stuff is pretty unique.. as such they must be struggling 19:53 < QuantumG> people buy what they know or what they're told by a personality to buy 19:54 < fenn> i had a nice depressing surf this morning of DK ahn's "web portfolio" or whatever it was supposed to be (the guy behind http://youtube.com/watch?v=rWgPH1vMFTk ) 19:54 < fenn> it's like people forgot how to play with legos 19:55 < QuantumG> its not just legos 19:55 < fenn> (previous comment was relevant because it shows the process by which "designers" are "educated") 19:55 < QuantumG> cad cam has destroyed tinkering 19:55 < kanzure> not true 19:55 < kanzure> you just don't have a CADCAM tool 19:55 < Smari> Where do I put ciabot.pl? 19:56 < QuantumG> ya.. 19:56 < fenn> Smari: if it's a bare repo it goes in the directory with config hooks/ etc 19:56 < Smari> ah 19:56 < Smari> I found it 19:57 < fenn> QuantumG: i tend to agree somewhat; that's more of a failing with the way cad software has been designed than an inherent property 19:57 < fenn> Smari: also take a look at http://kitenet.net/~joey/blog/entry/lazyweb:_git_cia_hooks/ 19:57 < fenn> i'm not sure why it doesnt do that automatically 19:59 < QuantumG> youtube is hating 20:01 < QuantumG> givin' me the wheel of death every 10 seconds 20:01 < fenn> youtube-dl.py ftw 20:03 < fenn> why is it EDA cad allows use of standardized components but mechanical cad makes you do everything from scratch every time? 20:03 < fenn> shouldnt there be an 'auto router' for mechanical stuff too? 20:04 < fenn> "hold these parts in place" 20:04 < kanzure> haven't we gone over this before? 20:04 < fenn> yeah i'm just bitching 20:05 < fenn> more of the same @ http://fennetic.net/cadwiki/ 20:07 < fenn> "[calculate cost] by machining time, painting time, electricity consumed, number of assembly steps (total movement necessary to assemble the product)" 20:07 < fenn> i guess i got that sort of backward 20:07 < Smari> Testing the bot now.. 20:08 < fenn> really you should be designing by choosing the cheapest processes in the first place 20:08 < fenn> and that would affect the shape 20:08 < Smari> I'm guessing that didn't work. 20:08 < Smari> I'll fix it later. 20:09 < Smari> gone! 20:09 < QuantumG> fenn: do you recommend any AR software? 20:09 < fenn> QuantumG: nope 20:09 < kanzure> what does that mean? 20:09 < fenn> QuantumG: it all looks like crap, unfortunately 20:10 < fenn> i havent looked very hard though 20:10 < QuantumG> k, so tell me, is there any open source software that can say "that's a face" on a live video stream? 20:10 < fenn> opencv 20:10 < fenn> or pretty much any vision library 20:11 < QuantumG> that's one of the things I'd like an AR system to do.. record every face it sees, and (off-line if necessary) classify them.. eventually to the extent that it can overlay a box around the face and heads up info next to it. 20:12 < kanzure> opencv yep 20:12 < kanzure> I keep on recommending it but I haven't actually used it yet 20:12 < QuantumG> you can assign names to the faces, notes, etc 20:12 < kanzure> so somebody give me a bloody status note on it or something 20:12 < QuantumG> I think I've downloaded it.. built it.. but I don't remember what, if anything, I did with it 20:13 < fenn> i guess i keep assuming everyone will have some wearable thingy with geo coordinates 20:13 < kanzure> ah finally got pythonOCC to continue compiling 20:13 < kanzure> fenn: there was a LocOpe* file that included the same file that didn't exist that your last sed line didn't grab 20:13 < kanzure> just because the list of files didn't include LocOpe* (it was TopOpe* only) 20:14 < fenn> patch plz 20:14 < kanzure> what do I diff against? 20:14 < kanzure> or isn't there an svn patch command? 20:14 < fenn> just be more verbose please 20:15 < fenn> "since it may not be possible to go from geometry to process specification without some sort of feature-recognition algorithm (unlikely in a user-generated process specification) we'll have to have process specification tools that "render" to 3d geometry, for instance a thread-turning process will render to a helix with all the thread geometry specifics" 20:16 < kanzure> that'd be nice to see in action 20:16 < kanzure> is that kokompe? 20:16 < fenn> that's what i'm doing this week 20:16 < fenn> with pythonOCC and skdb 20:16 < kanzure> did campbell ask you for that? 20:16 < fenn> er, yeah i guess 20:16 < kanzure> you're not one to spontaneously write large blocks of text 20:16 < fenn> that was my "half assed random process string" idea 20:17 < fenn> that text is from years ago 20:17 < fenn> it's barely two sentences anyway 20:19 < kanzure> for you that's a marathon 20:23 < kanzure> does anyone know where wx.aui comes from for python? 20:26 -!- Smari [n=spm@89-160-140-46.du.xdsl.is] has quit [Read error: 110 (Connection timed out)] 20:30 < kanzure> I can't seem to figure this out 20:30 < kanzure> fenn: how did you get pythonOCC/Tools/InteractiveViewer.py to run? 20:30 < fenn> worked first try 20:30 < kanzure> python version? 20:30 < fenn> 2.6 20:31 < kanzure> for some reason apt is telling me that there is only 2.5 available 20:31 < kanzure> and I'm back on testing 20:31 < fenn> do you have stuff in /usr/share/pyshared/OCC/Display/ ? 20:31 < kanzure> yes 20:31 < kanzure> it says that it cannot find wx.aui 20:31 < fenn> can you run the samples/ demos? 20:31 < kanzure> and frankly neither can I. I have no idea which package it is in 20:32 < kanzure> yes, but I need to set CSF_GraphicShr 20:32 < fenn> /usr/lib/python2.6/dist-packages/wx-2.8-gtk2-unicode/wx/aui.py 20:32 < kanzure> yeah I have that too 20:33 < kanzure> er, except for some reason I'm on python2.5/site-packages/..same.. 20:33 < fenn> can you do "import wx.aui"? 20:34 < fenn> (from pkg python-wxgtk2.8, now why didn apt-file find that?) 20:35 < kanzure> no, that's what's complaining 20:35 < kanzure> "import wx.aui" is what is failing 20:36 < fenn> something's wrong with your sys.path i guess 20:36 < fenn> if you really do have that file 20:39 < kanzure> wx-2.6-gtk2-unicode is in my sys.path 20:39 < kanzure> but not wx-2.8-gtk2-unicode 20:40 < kanzure> er how does removing python-wxgtk2.6 require me to get 23.6 MB of new archives? 20:41 < kanzure> math has defeated me 20:41 < fenn> there's gotta be some open source version of this boxes and arrows stuff http://developer.apple.com/macosx/images/CoreDataSimpleDataModel.jpg 20:41 < kanzure> UML? 20:42 < kanzure> or, UML renderers, rather? 20:42 < fenn> no, i just mean the boxes and arrows stuff 20:42 < kanzure> graphviz does it if you feed it CSS that it likes 20:42 < fenn> "graph widget" instead of tree widget 20:43 < fenn> blender does it 20:43 < kanzure> ? 20:44 < fenn> in the materials editor, and some of the animation stuff 20:44 < fenn> there's also some drawing program which i cant remember the name of 20:46 < fenn> wtf googling "open source procedural" leads to this? http://openfarmtech.org/images/7/7d/Wheel_drive_circuit.jpg 20:47 < fenn> the hive mind surely has me 20:47 < kanzure> damn it fenn 20:47 < kanzure> I was serious when I said no mind reading in here 20:47 < kanzure> I'll kick you if I have to 20:48 < kanzure> aha! 20:48 < kanzure> #if you get "wx.aui" not found: 20:48 < kanzure> sudo rm -f /usr/lib/python2.5/site-packages/wx.pth 20:48 < kanzure> sudo echo "wx-2.8-gtk2-unicode" >> /usr/lib/python2.5/site-packages/wx.pth 20:49 < kanzure> InteractiveViewer.py now runs 20:49 -!- xp_prg [n=xp_prg3@99.2.31.217] has quit ["This computer has gone to sleep"] 20:49 < kanzure> wow it runs slow here 20:51 < kanzure> awesome. all that work to get a slow-as-all-hell app. 21:10 -!- xp_prg [n=xp_prg3@99.23.56.166] has joined #hplusroadmap 21:23 < fenn> there's some open source compositing drawing program that has some kind of bird and trees as its logo, ring any bells? 21:24 < kanzure> fenn: is there some better way to do this? from OCC.Display.wxSamplesGui import start_display, add_function_to_menu, add_menu 21:25 < kanzure> oh wait nevermind 21:25 < fenn> that looks fine 21:25 < kanzure> well I could just do OCC.Display.wxSamplesGui.add_menu() 21:29 < CIA-73> skdb: kanzure * rac66a42460eb /pymates/pymates.py: pythonOCC+pymates integrated somewhat. need to write wrapper. 21:31 < kanzure> please note this is only there for the time being 21:31 < kanzure> there's no reason pymates should be a mainly graphical thing :p 21:31 < kanzure> but campbell insists, so. 21:36 < fenn> The concept of the fablab CAD/CAM toolset is to be able to describe almost anything, in a way that can be interpreted by almost any fabrication machine, to make parts for almost anything 21:36 < fenn> Just as desktop publication is enabled by algorithmic representation of page layout (page description languages), the fablab CAD/CAM software enables desktop fabrication through algorithmic representation of three dimensional shapes and systems with functional components. 21:37 < kanzure> yummy 21:37 < fenn> where 'fablab cad/cam' = kokompe 21:37 -!- xp_prg [n=xp_prg3@99.23.56.166] has quit ["This computer has gone to sleep"] 21:42 < kanzure> http://adl.serveftp.org/lab/2009-07-17_pymates.png 21:42 < kanzure> not that it's interesting 21:43 < fenn> it makes a blank window? 21:44 < kanzure> yep. fear me! 21:44 < kanzure> you're not fearing >:( 21:49 < kanzure> how am I going to run an efficient world domination co-op if you're not going to fear me? 21:54 < kanzure> fenn: how did you get Utils working? 22:06 < CIA-73> skdb: kanzure * r18ba9bcc0740 /pymates/pymates.py: pymates loads CAD files. todo: wx menu file selector. 22:55 < CIA-73> skdb: kanzure * rd7d56e1ac6ec /pymates/pymates.py: added wx.FileSelector() 23:24 < genehacker> don't they already have that? 23:28 < genehacker> sounds like you're reinventing the wheel 23:53 < ybit> david dalrymple, 13 year old who graduated from MIT in 2005, that david dalrymple? never heard of him ...'til just now 23:53 < ybit> kanzure: suppose you met him last summer during your internship? 23:55 < ybit> met a hot waitress today studying to be a neurosurgeon, needless to say, we hit off well 23:55 * ybit got the digits 23:55 < ybit> "back of the net!" 23:57 < bkero> digitz? 23:58 * ybit is persuading her to join the darkside of neuroeng 23:59 < ybit> back to reading teh logs