--- Day changed Tue Nov 25 2008
00:04 < kanzure> VA?
00:04 < kanzure> xp_prg: I saw John's announcement, if that's what you mean
00:04 < xp_prg> ya
00:04 < xp_prg> VA = virignia
00:04 < xp_prg> kanzure I am going to put in an app, may I mention our work together?
00:04 < kanzure> I'd have trouble getting up there.
00:04 < kanzure> xp_prg: Only if you run it by me first.
00:04 < xp_prg> they are paying for transportation
00:04 < kanzure> I don't think you know the entire system yet.
00:05 < xp_prg> ok
00:31 < kanzure> ybit: You should tell him that.
00:31 < kanzure> ybit: Seriously. I don't know why he doesn't idle in here.
00:31 < kanzure> It's not like it would hurt him.
00:31 < kanzure> Arrgh! The pain! Moderately relevant topics!
00:34 -!- mmarmer_ is now known as mmarmer
00:34 < kanzure> Hi mmarmer.
00:34 < mmarmer> hey
00:54 < kanzure_> http://www.artas.nl/ [commercial link] User-friendly software for design, motion/force analysis & optimization of planar linkage, gear and belt mechanisms
00:56 < kanzure_> Designers will synthesize a linkage by starting with the required output motion, mechanical advantage, velocity and acceleration. A type of linkage is chosen and modified to deliver the required performance.
00:56 < kanzure_> Each link is treated as a vector and the vectors can be combined into a system of equations because they form a loop. The matrix is solved to create a closed form equation that relates input motion to output motion. The same is done for mechanical advantage, or any other important quantity. The equations of motion are differentiated with respect to time to find velocity and acceleration of the mechanism parts.
00:58 < kanzure_> http://commons.wikimedia.org/wiki/Linkage_(mechanical)
01:01 < kanzure_> http://www.youtube.com/watch?v=b0I65dJMQUc&NR=1 blender tutorial - theo jansen mechanism
01:02 < kanzure_> http://www.youtube.com/watch?v=JoQj-tobYOw&feature=related wheel-less bicycle. :)
01:05 < kanzure> hm, if I had a GA or other search algorithm for making bar/linked mechanisms with various degrees of freedom, how would you prevent it from just converging on the shape of a circle? I'm imagining a simualtor where you apply a force and try to see how far it goes or something
01:05 < kanzure> I guess it's a question of what to optimize for - diversity of generated designs to solving the same problem? That might be a sufficient metric.
01:06 < fenn> how does it steer?
01:07 < kanzure> "I've invented a brilliant bicycle that doesn't use wheels. The only downside is that it only goes in one direction. Good luck."
01:07 < kanzure> hehe
01:08 < fenn> i know the strandbeest could steer somehow to keep it on the beach
01:09 < kanzure> I don't remember what its overall shape was, was this just the linear one like a centipede?
01:09 < kanzure> and then it would use the pressure sensors made out of its feet to go in opposite directions with some internal logic
01:09 < kanzure> so it might have been a circular design or something
01:09 < kanzure> with eight legs or something.
01:09 < kanzure> eight "chunks" of legs.
01:10 < kanzure> so eight directions I guess.
01:13 < kanzure> hrm. 
01:13 < kanzure> That's odd.
01:15 < kanzure> Anyway, is this stuff anything more than just bars and joints?
01:18 < fenn> no
01:19 < kanzure> can you braindump a list of calculations that should be associated with these?
01:19 < kanzure> it shouldn't be hard for me to make up some objects somewhere and then randomly connect them and try them out. don't know why ADL has an entire person working on this.
01:22 < kanzure> I wonder if there's an interactive gnuplot for python dealy out there yet.
01:38 < ybit> some local is going to give me a few G3s, netvista machines, monitors, and keyboards. i <3 freecycle
01:38 < kanzure> Yep.
01:39 < ybit> http://www.vmunix.com/mark/blog/archives/2008/10/17/cloud-computing-is-a-sea-change-how-sysadmins-can-prepare/
01:39 < kanzure> Oh, I'm getting access to a moderate sized cluster soon.
01:40 < ybit> what is moderate?
01:40 < kanzure> Not super.
01:40 < ybit> heh :P
01:40  * ybit wants more details
01:41 < ybit> aside from "not quite super, not quite negligible"
01:41 < kanzure> I'll get you the details when I get them.
01:41 < ybit> okey
01:41 < ybit> this guy told me to bring a pickup and take as many machines as i wanted, he bought hundreds at a recent auction o.O
01:41 < kanzure> It looks like for mechanical kinematics all I'd have to do is encode degrees of freedom constraints
01:41 < kanzure> heh
01:42 < kanzure> Yeah, I met a guy like that, didn't go up to his house though
01:42 < kanzure> he was a freecycler somewhere up in North Texas and basically has old AT terminals in giant junk piles on a few acres
01:42 < kanzure> does a lot of 'research' on his own time, that sort of thing
01:42 < kanzure> Can't get him oline because of poor eyesight apparently.
01:43 < kanzure> Bill Snodgrass, I think.
01:47 < fenn> parallel kinematics gets ugly quick
01:47 < fenn> sometimes there isn't a direct analytical transformation from joint space to cartesian space
01:47 < kanzure> oh?
01:50 < fenn> hmm. somehow i find it hard to take statements like this seriously:
01:50 < fenn> "That's right, as an industry we're truly about to automate ourselves out of our jobs. Which, of course, is what we've been nobly aspiring to do for 30 years, but, folks, this time we're actually going to do it"
01:51 < kanzure> source?
01:52 < fenn> ybit's link
01:52 < fenn> cloud-computing-will-kill-us-all
01:52 < kanzure> so what's so hard about this kinematics stuff
01:52 < fenn> dunno
01:52 < kanzure> I mean, when I look at it, it's like animation skeletons for video games 
01:53 < fenn> that's not parallel links
01:53 < ybit> fenn, he says there are other tasks that sysadmins will be performing
01:53 < kanzure> what's a parallel link?
01:53 < ybit> not the ones typically performed today
01:53 < fenn> well, like theo janssen mechanism has two parallel links
01:53 < fenn> er, two links in parallel
01:54 < kanzure> so just a rectangle.
01:54 < fenn> the stewart platform (hexapod) is another parallel topology
01:54 < fenn> your neck uses parallel linkage
01:54 < kanzure> what's the linking mechanism?
01:54 < fenn> muscles
01:55 < kanzure> no, in let's say the theo janssen mechanism
01:55 < fenn> oh, pins and bearings i guess
01:55 < kanzure> so what's hard about doing that taking analogies from animation skeletons
01:55 < fenn> doesn't really matter for the purposes of calculating where it will be given a certain pose
01:56 < kanzure> again, degrees of freedom. those linkers being highly constrained of course.
01:56 < fenn> "why is anything hard"
01:57 < kanzure> well I guess there's some hardness in figuring out how to draw interesting things out of these beams, Hod had a lot of video clips of things that didn't even barely flo
01:58 < kanzure> flop
01:58 < kanzure> and they were using some internal linear actuators
02:01 < kanzure> I guess a 'hard' part about such generators for these mechanism designs would be to make sure that parallel structures crop up at some frequency, rather than allowing a full range of irrational numbers to specify placement coords
02:01 < kanzure> although presumably all patterns can be broken down into simpler more fundamental patterns, right? and that's just a single bar, no?
02:04 < fenn> how many foglets could a foglet chuck
03:08 < wrldpc> anyone have that open currency link fenn posted earlier?
03:08 < wrldpc> I think it was fenn ..
04:37 < fenn> wrldpc: http://openmoney.info/sophia/index.html http://evolvingtrends.wordpress.com/2008/10/21/p2p-social-currency-money-20/  probably one of those
04:37 < wrldpc> ty
04:37 < fenn> i'm rather against calling it "currency" or "money" since there are too many preconceived notions about what money is already
04:38 < fenn> and it's too easy to just say "money sucks" without providing any solutions
04:40 < kanzure> blah, now I'm hooked on figuring out how to pull off a koyaanisqatsi for manufacturing. It's so easy with classical music + trance paired to random hypnotizing shots.
04:41 < kanzure> "look, a spinny thinger"
04:50 < fenn> a lot of "industrial" music came from people using sampled sounds of machinery
04:51 < fenn> would be neat to do some sort of video sequencing where you see it and hear it at the same time
04:51 < kanzure> so when I heard about Black Sabbath and the finger accident, I thought it was kind of cool that they went on to do metal
04:51 < kanzure> but then I realized it was just a finger
04:51 < kanzure> people lose half their bodies, most of their brains, etc.
04:51 < kanzure> where's *their* heavy metal? :/
04:53 < fenn> heh maybe you should just start recruiting brain damaged people to play music
05:16 < kanzure_> "You could be the first on your block to own a residual gas analyzer which has an open filament in the ion source. Replacement filaments are available for $485 and I do not wish to spend this much to repair the instrument."
05:16 < kanzure_> freecycle contact I think
05:18 < fenn> hmm how much could some tungsten wire cost
05:20 < fenn> still that's not a bad price for a working mass spec.
05:21 < kanzure> what, free? :)
05:21 < kanzure> why is it that I get equipment when I least need it / least have room for it
05:21 < kanzure> this is silly. we have excellent mass spec machines on campus
05:21 < kanzure> and no room in a dorm
05:21 < kanzure> where's the bryanlab when I need it
05:21 < fenn> how big is it?
05:21 < kanzure> dunno, I've asked though
05:22 < fenn> some pics on the intarweb look lunchbox sized
05:24 < fenn> of course it would be sorta useless without a high vacuum system
05:24 < fenn> http://www.pascaltechnologies.com/Products/SpecialThumbs/Complete_XTx00big.jpg
05:27 < fenn> "glow discharge between the quadrupole rods is the main limit to high pressure operation, not the mean free path length of the residual gas molecules, so they have roughly the same high pressure operating limit. Operation at high pressures will result in a very short filament life also."
08:52 -!- Utopiah is now known as UtopiahGHML
11:21 < UtopiahGHML> 7 Beautiful Data Visualizations (With Videos)
11:21 < UtopiahGHML> November 24, 2008 http://mashable.com/2008/11/24/data-visualizations/
11:21 < UtopiahGHML> (not really new though)
12:39 < kanzure> hm, a friend that I've known forever at 2 am: "I am officially scared of /b/"
12:39 < kanzure> "they can find people so easily"
12:39 < kanzure> we have agents everywhere, rawr
13:09 < kanzure_> "Elysia chlorotica is a lurid green sea slug, with a gelatinous leaf-shaped
13:09 < kanzure_> body, that lives along the Atlantic seaboard of the US. What sets it apart
13:09 < kanzure_> from most other sea slugs is its ability to run on solar power."
13:09 < kanzure_> http://www.newscientist.com/article/dn16124-solarpowered-sea-slug-harnesses-stolen-plant-genes-.html
13:09 < kanzure_> Mary Rumpho of the University of Maine, is an expert on E. chlorotica and has
13:09 < kanzure_> now discovered how the sea slug gets this ability: it photosynthesises with
13:09 < kanzure_> genes "stolen" from the algae it eats.
13:11 < kanzure_> http://www.pnas.org/content/105/46/17867/suppl/DCSupplementalhttp://www.pnas.org/content/105/46/17867/suppl/DCSupplemental
13:11 < kanzure_> http://www.pnas.org/content/105/46/17867/suppl/DCSupplemental
13:14 < kanzure_> http://holy-web.blogspot.com/2008/11/amazing-video-excavating-giant-538.html Amazing video: excavating a giant (538 square feet) ant colony
13:17 < kanzure> at 4:16 it starts to get interesting
13:18 < kanzure> gah, they so need to digitally scan all of that :/
13:19 < UtopiahGHML> the equivalent of the Blue Brain Project for eusocial species
13:20 < UtopiahGHML> could provide some interesting "good pratices" recommendation for MAS/Mashup of WebAPIs 
13:20 < UtopiahGHML> automatized biomimetics
13:20 < kanzure> MAS?
13:20 < UtopiahGHML> multi-agents systems
13:22 < UtopiahGHML> could submit results to sth like Software Patterns and Quality 09
13:28 < UtopiahGHML> (automatized biomimetics that would be like... real machine learning)
13:30 < fenn> searching cyberspace for signs of intelligent life
13:49 < kanzure> No results found. :(
13:51 < wrldpc> fuck /b/
13:51 < wrldpc> no (no offense)
13:54 < wrldpc> that psychopath jihad who occasions this room is a barometric for the kind of misguided intellects that populate 4chan.
13:54 < kanzure> I used to be one of the occupants of 4chan. :/
13:55 < wrldpc> So did I .. for like a noncontiguous week.
13:55 < kanzure> I was so excited when I learned that there was another 4channer in my high school when I was a freshman. Ah, those were the days.
13:55 < kanzure> heh
13:56 < kanzure> All games are either too short or too hard (to the point where you're pretty sure that the level boss is impossible to beat).
13:56 < kanzure> Anything in between? besides, like, tetris.
13:57 < wrldpc> If they had any real skills and any real intelligence (yes, intelligence) they would hijack the Super Bowl broadcast this year and transmit a neo bla bla bla
13:57 < wrldpc> tetris is nice ..
13:58 < wrldpc> good book by Campre I think called Finite and infinite games
13:58 < wrldpc> infinite games>finite games
13:58 < wrldpc> sorta ..
13:58 < wrldpc> a neo-golden record http://en.wikipedia.org/wiki/Voyager_Golden_Record
13:59 < wrldpc> a mish mash of Aubrey de Grey, Kurzweil, Gershenfeld, ........
13:59 < wrldpc> dunno
13:59 < wrldpc> BRYAN
13:59 < wrldpc> hehe
14:00 < wrldpc> Just hijack the transmission and broadcast heybryan.org for 2 consecutive hours ... redirect everyone to heybryan.org
14:00 < wrldpc> bleh
14:01 < kanzure> I'm amazed that people who run television networks still know how to do "live" footage.
14:01 < wrldpc> why isn't there anyone out there with robust spam technologies filling everyone's email box with the H+ memes?
14:02 < kanzure> Why would that be productive?
14:02 < kanzure> besides "male enhancement" spamvertisements
14:02 < wrldpc> heh
14:02 < wrldpc> it's very theoretical
14:02 < wrldpc> the gist is ... the memes are like software the brains process ... 
14:02 < kanzure> Everybody always wants to do something to "help H+" but it always turns into some stupid publicity thingy.
14:03 < wrldpc> more processors ..
14:03 < kanzure> bullfucking shit
14:03 < wrldpc> you're right.
14:03 < wrldpc> LOL!
14:03 < kanzure> <--- is enough of a processor
14:03 < wrldpc> :)
14:03 < wrldpc> damn straight.
14:03 < kanzure> now give me metals
14:03 < wrldpc> k
14:03 < kanzure> and a lab
14:03 < wrldpc> kk
14:03 < kanzure> and a bajillion virgins
14:03 < wrldpc> kkk
14:23  * fenn puts in a request for a space elevator
14:47  * wrldpc seconds fenn's request!
14:57 < kanzure> Aren't you supposed to be the one filling the order though?
17:04 < fenn> i'm rather disappointed that a car exhaust costs more than a decent laptop
17:08 < kanzure> I'm disappointed that nobody has made a full pokemon clone engine yet. :/ it's like ff3 or something except less.
18:46 -!- xp_prg2 is now known as xp_prg
19:10 < bkero> What?
19:10 < bkero> fenn: You're looking at stupid designer exhaust.
19:10 < bkero> My exhaust cost me $160 for a nice downpipe, $35 for a cherry bomb, $50 for a muffler shop to provide extra pipe and weld it up.
19:55 < kanzure_> http://www.pge.utexas.edu/faculty/srinivasan.cfmv <- Meeting with this guy today at 4 pm. Any suggestions?
20:27 < xp_prg> nothing showed for me kanzure_
20:27 < kanzure> Where?
20:27 < xp_prg> that link
20:27 < kanzure> I probably added an extra v at the end.
20:28 < xp_prg> I suggest you tell him synthetic biology is the future and "step off bitch"
20:28 < kanzure> ??
20:28 < kanzure> What do you think he does?
20:29 < xp_prg> Petroleum engineering
20:29 < xp_prg> he works at stanford?
20:29 < kanzure> And what do you think I do?
20:29 < kanzure> biofuel bioreactors don't have the word fuel in it for nothing
20:29 < kanzure> while it's not petro, I just don't see what you're getting at
20:29 < kanzure> synthetic biology assists in variosu fuel scenarios, fyi
20:29 < xp_prg> yes I know
20:29 < xp_prg> tell him that
20:30 < xp_prg> and then to "step off"
20:31 < xp_prg> does that make sense kanzure?
20:32 < kanzure> Not really. I don't understand, do you have a bias against petro people?
20:32 < xp_prg> no, dude, making a joke
20:32 < xp_prg> just trying to say synth bio is where its at
20:32 < xp_prg> he probably doesn't understand its potential
21:00 < kanzure_> "Today, a Japanese startup called Lunascape has released an alpha version of its Lunascape browser... that allows you to switch between all three of these prominent rendering engines. The company says that the Japanese version of Lunascape has been downloaded 10 million times and touts it as the fastest browser available.""
21:00 < kanzure_> I wonder how they're managing to do that. I guess they either have backend deals with Microsoft, and some tie-in with Gecko, and webkit. 
21:00 < kanzure_> oh, Gecko, Trident and Webkit. Okay.
21:00 < kanzure_> Meh.
21:00 < kanzure_> Not interesting. Nevermind.
21:09 < kanzure_> re what I was talking about on blender + kinematics + animation - http://www.lysator.liu.se/~torkel/graphix/blender/ikman.html
21:10 < kanzure_> http://www.squidoo.com/BlenderCharacterRigging
21:13 < kanzure_> http://wiki.blender.org/index.php/BSoD/Introduction_to_Rigging/Character_Rigging_In_Review
21:29 -!- nshyah is now known as nsh
21:33 < gene> blender huh?
21:33 < kanzure_> just showing that mechanical kinematic components have been done in blender.
21:36 < gene> well the beauty of rigidchips isn't that you can make mechanical kinematic components, it's that you can quickly and easily make something with them that you can control
21:36 < kanzure> why do you think I was comparing blender to rigidchips?
21:36 < kanzure> I wasn't ..
21:36 < gene> ok
21:36 < kanzure> and also, it's not beauty because it's closed source and proprietary.
21:37 < gene> I wouldn't exactly call it proprietary
21:37 < gene> it's free
21:37 < gene> neither is it open source
21:38 < bkero> There's some pretty impressive stuff done in blender.
21:39 < gene> though there is a resources folder with all the textures, basic model meshes and stuff like that
21:39 < gene> there might be some other stuff in there too
21:39 < gene> but I haven't really checked it out
21:39 < gene> because it's all in japanese
21:39 < bkero> This was done in Blender: http://downloads.sketchworkproductions.com/peach/big_buck_bunny_1080p_h264.mov
21:39 < kanzure> ah yes the "open source movie" crazies :)
21:39 < gene> I can't see it
21:41 < gene> hey, there might be source code in one of the folders included with rigid chips
21:41 < gene> but it's all in japanese
21:42 < gene> can I post some of it?
21:42 < gene> ---------------------^???VB6?T?"?v?<(Form + Button(Command1) ,??\??)
21:43 < gene> 'RigidChips,??f?[?^?t?@?C?<,??J,-???b?Z?[?W,?'--,?(VB6?T?"?v?<)
21:43 < gene> 'WinAPI"o~^
21:43 < gene> Private Declare Function RegisterWindowMessageA Lib "user32" (ByVal msgid As String) As Long
21:43 < gene> Private Declare Function PostMessageA Lib "user32" (ByVal hwd As Long, ByVal msg As Long, ByVal wpara As Long, ByVal lpara As Long) As Long
21:43 < gene> Const HWND_BROADCAST = -1 '0xFFFF
21:43 < gene> '???b?Z?[?W??o?'??"
21:43 < gene> Const UMSG_RCLOAD_START = 0   'o???--?'--?M?J?n
21:43 < gene> Const UMSG_RCLOAD_CHAR = 1    '^?o???,,,?'--,?
21:43 < gene> Const UMSG_RCLOAD_END = 2     'o???--?'--?M?I--?
21:43 < gene> Dim WmRCLoad As Long
21:43 < gene> Private Sub Form_Load()
21:43 < gene> ' <?--p?E?B?"?h?E???b?Z?[?WID(WM_RIGHTCHIP_LOAD)"o~^
21:43 < gene>   WmRCLoad = RegisterWindowMessageA("WM_RIGIDCHIP_LOAD")
21:43 < gene> End Sub
21:43 < gene> Private Sub Command1_Click()
21:43 < gene>     
21:43 < gene>     ''--,?, 1/2 ,?o???--??i?p?Xot,<<?t?@?C?<- 1/4 ?j'?:VB,?String,?Unicode
21:43 < kanzure_> yawn, basic.
21:43 < gene>     filenameW = "C:\Rigid15B19\Data\BigFoot.txt"
21:43 < kanzure_> You can stop now.
21:43 < gene>     'Unicode,(C),??V?t?gJIS,?,?o???
21:43 < gene>     filenameA = StrConv(filenameW, vbFromUnicode)
21:43 < gene>   
21:43 < gene>     '?E?B?"?h?E???b?Z?[?W'--?M
21:43 < gene>     'o???--?'--?M?J?n
21:43 < gene>     PostMessageA HWND_BROADCAST, WmRCLoad, UMSG_RCLOAD_START, 0
21:43 < gene>     '^?o???,,,?'--,?
21:43 < gene>     For i = 1 To LenB(filenameA)
21:43 < gene>       PostMessageA HWND_BROADCAST, WmRCLoad, UMSG_RCLOAD_CHAR, AscB(MidB(filenameA, i, 1))
21:43 < bkero> Looks like code to me
21:43 < gene>     Next
21:43 < gene>     'o???--?'--?M?I--?
21:43 < gene>     PostMessageA HWND_BROADCAST, WmRCLoad, UMSG_RCLOAD_END, 0
21:43 < kanzure_> http://heybryan.org/bioreactor/membraneless_filtration/2008-11-25_something_wrong.png <- gnuplot stuff.
21:43 < gene> End Sub
21:43 < gene> would that be considered open source?
21:44 < gene> if it includes what the code is?
21:44 < bkero> No
21:44 < UtopiahGHML> gene: http://paste-it.net/
21:44 < bkero> gene: If 'seeing the code' means it's open source, then all of php is inherently open source.
21:44 < gene> this is in the documentation folder
21:45 < kanzure_> Anyway, what do you want to do with it gene?
21:45 < gene> make it better
21:45 < kanzure_> do you expect me to use this Visual Basic code to do something?
21:45 < kanzure_> *source code
21:45 < gene> NO
21:45 < kanzure_> CPE 3.118 apparently.
21:45  * kanzure_ leaves
21:45 < gene> bye
23:07 < kanzure> I'm surprised. Sanjay knew about VTK and OpenFOAM and was quite enthusiastic about these two packages being used.
23:10 < gene> kanzure?
23:10 < kanzure_> gene: apt-get install freefem3d ifrit libvtk5 libvtk5-dev mayavi mayavi2 octaviz python-vtk python-pyvtk vtk-doc vtk-examples vtk-tcl vtkdata 
23:10 < gene> hold on a second, I have to install linux
23:11 < gene> are you serious?
23:11 < kanzure_> Yes, the packages really are useful.
23:11 < gene> that's a lot
23:11 < kanzure_> it's like 566 MB of packages, so you should do it in the lab
23:11 < kanzure_> will take about 5 minutes.
23:12 < gene> what lab?
23:13 < kanzure_> the one that you have access to
23:13 < kanzure_> namely Sata's
23:13 < kanzure_> because he doesn't have download restrictions
23:13 < gene> I ain't gonna be here all week so it don't matter if I use up all the band width
23:13 < kanzure_> If you really want to use apt-get, which I recommend, then you'd need to get linux first, yes.
23:14 < gene> wait a minute kanzure, your using arctan2 to plot polar equations?
23:14 < gene> may I ask why?
23:14 < kanzure_> that's cartesian.
23:14 < gene> it makes no sense
23:14 < gene> oh
23:14 < gene> you want to convert to cartesian?
23:15 < gene> polar to cart is easy
23:15 < kanzure_> well, ultimately GPUs use cartesian coordinates, so whatever I end up using should be in cartesian, but I can convert to and from polar to cartesian with the cosines and sines.
23:15 < kanzure_> I don't want to convert that equation that I pasted,
23:15 < kanzure_> you mentioned knowing of something better
23:15 < kanzure_> which more adequately represents what we're doing
23:16 < kanzure_> uh, for spirals I mean.
23:16 < gene> yeah man
23:16 < gene> sines and cosine are what you use
23:17 < kanzure_> What do you have in mind?
23:38 < kanzure> what's with the VTK documentation being 60 MB?