--- Day changed Sun Apr 12 2009 00:04 < fenn> some rather funny movies from Liu's thesis work: http://cg.cis.upenn.edu/hms/research/archive/NASA/images.html 00:09 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has joined #hplusroadmap 00:10 < kanzure> "Center for Human Modeling and Simulation" 00:10 * kanzure wonders if they have a braincount > 0 00:12 < genehacker> hahaha 00:14 < genehacker> http://resistorbeterminated.com/skynet-model-101-seeks-blood/comment-page-1/#comment-948 00:14 < genehacker> check it out cheap dna sequencer 00:17 < genehacker> just have to faraday cage the thing 00:18 < fenn> who wants to bet this is a marketing campaign for an upcoming video game 00:27 < genehacker> bah it's for a game 00:27 < genehacker> that would be what's in the box 00:28 < fenn> reminded me too much of ilovebees 00:28 < genehacker> what's in the box? 00:28 < genehacker> hahaha bre's getting help making the makerbot from skynet 00:29 < genehacker> if only 00:29 < genehacker> damn thing might not cost as much 00:29 < genehacker> link to that mill about same size as maker bot but cheaper? 00:35 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: katsmeow-afk, faceface, myelinzar, UtopiahG1ML, chizu_, oublei, bkero, nsh, genehacker, Overand, (+7 more, use /NETSPLIT to show all of them) 00:38 -!- Netsplit over, joins: cis-action, drazak, genehacker, tarbo2, davidsjo, bkero, boogles, chizu_, myelinzar, faceface (+7 more) 00:48 -!- Netsplit calvino.freenode.net <-> irc.freenode.net quits: katsmeow-afk, faceface, myelinzar, UtopiahG1ML, chizu_, oublei, bkero, nsh, genehacker, Overand, (+7 more, use /NETSPLIT to show all of them) 00:50 -!- Netsplit over, joins: cis-action, drazak, genehacker, tarbo2, davidsjo, bkero, boogles, chizu_, myelinzar, faceface (+7 more) 01:16 -!- boogles [n=boogles@osuosl/staff/Boogles] has quit [Killed by brown.freenode.net (brown.freenode.net (boogles(?) <- ballard.freenode.net))] 01:16 -!- boogles [n=boogles@ash.osuosl.org] has joined #hplusroadmap 01:34 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has quit [Read error: 110 (Connection timed out)] 02:19 -!- elias` [n=me@cs78208074.pp.htv.fi] has joined #hplusroadmap 02:25 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has joined #hplusroadmap 03:42 -!- cis-action [n=cis-acti@146-115-127-170.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com] has quit [] 03:54 -!- UtopiahG1ML is now known as UtopiahGHML 04:15 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has quit [Read error: 110 (Connection timed out)] 04:32 -!- jm [n=jm@p57B9EA5E.dip.t-dialin.net] has joined #hplusroadmap 11:00 < kanzure> bre's a jerk for not being on om 11:03 < fenn> he's too busy making youtube videos to write text 11:09 -!- jm|space [n=jm@p57B9D8BB.dip.t-dialin.net] has joined #hplusroadmap 11:09 < kanzure> heh 11:09 < kanzure> I feel like bre's as much of an ass as tito is now 11:13 < fenn> katsmeow-afk: looks like you could make a stewart platform with those lexus steering units 11:25 -!- jm [n=jm@p57B9EA5E.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] 11:57 < kanzure> fenn, so I was thinking that with the "doll model" we could just let people assign tasks to The Human when they are making their model. i.e., "need to be able to translate 20 meters in x direction" and they would say, "yes, that's reasonable for a human to do" 11:59 < fenn> how is that better than just not saying anything at all? 12:04 < kanzure> what would not saying anything at all look like? 12:11 -!- cis-action [n=cis-acti@146-115-127-170.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com] has joined #hplusroadmap 12:13 < kanzure> "An object that does not alter the color of light reflected from its surface will appear white" 12:15 < kanzure> fenn: what I was thinking of was providing a dummy package, a doll package, with a specific API of different movements that are possible (like for common lab motions that are required) 12:15 < kanzure> the real sciency way to do this would be to use motion-capture of somebody performing 20 common lab protocols in a lab, and then do regression analysis to figure out all categories of common movements 12:16 < kanzure> but we know from experience that we can say three or four basic different movements like "move(obj,pos)" and "shake" and so on 12:16 < kanzure> so we would just have to ask whether or not some future 6-axis robo-arm package is compatible with "dummy-doll-1.1.3" 12:17 < kanzure> (in particular, the API with the move() calls and such) 12:17 < kanzure> right? 12:17 < fenn> i say if you're concerned about compatibility/interchangeability then specify the process, i.e. what happens to the thing you're doing stuff to 12:18 < kanzure> we were talking about a constraints-based representation of the process of PCR, such that we could say pre/post conditions after each step 12:18 < kanzure> and those conditions might involve something like moving something somewhere else 12:19 < kanzure> so you'd somehow specify that you need to move something 12:19 < kanzure> gah, this is too abstract 12:20 < kanzure> how are automatic pipetter robots programmed? in terms of CAM.. 12:20 < fenn> no idea 12:20 < fenn> i recall seeing some CSV files on openwetware 12:20 < kanzure> I think they assume a specific array of pipette tips 12:20 < kanzure> and then it's just a matter of passing a few variables about the xy location of where something interesting is 12:21 < kanzure> they don't even do positioning/staging- they just have an array of pipette-connections that move down on to the array :p 12:21 < kanzure> (well, some of them, I mean) 12:22 < fenn> i think it would be rather hard to move drops from one well to another without positioning 12:23 < kanzure> was thinking of something else then, maybe from The Robot Scientist video where they were doing things in bulk and just transferred things from one array/plate to another 12:23 < fenn> probably 12:24 < kanzure> they get to assume static frames of reference. 12:24 < fenn> it's quite common to have 8 pipettes in a row all doing the same thing 12:24 < kanzure> wait, why can't we assume static frames of reference? 12:24 < kanzure> so, when I was doing benchwork, most people just had little tubes laying around all over the place 12:25 < kanzure> and they never really had a good organization scheme 12:25 < kanzure> some people were smart and kept them in their plastic-container-array thingies. 12:25 < kanzure> of course, you could get confused when trying to place something at a particular xy location in the array 12:25 < kanzure> but not if you had instructions or something 12:25 < fenn> uh, who doesn't keep tubes in a tube rack? 12:25 < kanzure> what are those tiny tubes-that-aren't-tubes that you put in thermocyclers? 12:26 < fenn> mini tubes? 12:26 < kanzure> some people store their NTPs in them 12:26 < fenn> there's two sizes of eppendorf tubes 12:26 < kanzure> hrm. maybe? anyway, those 12:26 < fenn> microcentrifuge tubes i think they're called 12:26 < kanzure> yeah, that's the size 12:26 < kanzure> the desktop centrifuges and the small tubes for 'em 12:26 < kanzure> anyway, my labmates didn't have a systematic way of placing them 12:27 < kanzure> sometimes they would just be randomly thrown on ice while something else was being done 12:27 < kanzure> (maybe I'm just the most anal lab monkey ever?) 12:28 < kanzure> anyway, I mentioned this because if we can assume a 'static frame of reference'- like an 'array cartridge' for storing microcentrifuge tubes- then we don't have to deal with positional constraints really 12:28 < kanzure> we just say whether or not the PCR machine is compatible with a 10x10-skdb-blah-name-microcentrifuge-tube-array 12:28 < fenn> ok this is nothing new 12:28 < kanzure> (and of course, it wouldn't be, since we'd probably want to design one that didn't work on an array first? I dunno) 12:28 < kanzure> yeah it is 12:28 < kanzure> because now we don't have to do all that doll stuff :p 12:29 < fenn> you invented this doll stuff to make things harder on yourself 12:29 < kanzure> anything that implements the microcentrifuge-tube-array will just need to implement move() and transfer() and a few other thingies 12:29 < kanzure> thanks mr. psychologist 12:29 < kanzure> ~ :p 12:30 < kanzure> now how do you tell whether or not you have something that implements the microcentrifuge-tube-array holder and its API? 12:30 < fenn> yes it's nice to have a model of what a human is capable of, for checking whether something is compatible 12:30 < fenn> no you don't have to specify every little movement in every protocol 12:31 < fenn> if it 'implements' something it'll have the tag in it's functionality list 12:32 < kanzure> ok, so a pipette will be a compatible part- what my question is asking is how do you know that a human can hold the pipette 12:32 < kanzure> we know that humans can hold them (heck, they are designed for grasping) 12:32 < kanzure> oh, it seems I've just pushed the same problem to another place 12:32 < kanzure> drats 12:34 < kanzure> if you let me specify that a human can grasp it (this is what I called a "stopper" or "terminator" in terms of the recursion of whether or not the person has a tool laying around to do something, or if they have something equivalent already (i.e., their hands)) 12:34 < kanzure> then that would be so much easier .. 12:34 < kanzure> (than the doll stuff) 12:34 < fenn> if it has a handle, feel free to tag it as a handle 12:35 < kanzure> what good does that do? ok, you have it tagged 'handle' now what? 12:35 < fenn> now your doll human knows where to pick it up 12:36 < kanzure> oh, you mean tagging a surface/interface on the model as 'handle' and then checking whether or not the "hand" geometrically fits there? 12:36 < fenn> something like that 12:36 < kanzure> but some other grasping part in the db might match too 12:36 < fenn> sure 12:37 < kanzure> hm. does there need to be a bias for finding solutions that the "doll" is involved in? 12:37 < fenn> no reason to discriminate against artifical humans :P 12:37 < fenn> it's a matter of cost optimization 12:37 < kanzure> but then you'll be building things forever in an attempt to get this weird artificial hand working, giant pipes and manufacturing plants wired up all the way up to Escher's hand (the hand-that-makes-a-hand) 12:38 < fenn> if the human is cheaper/already available, then use it (if you're optimizing based on cost) 12:38 < kanzure> whereas you could have just used a human hand for that silly PCR protocol so many hundreds of years ago :p 12:38 < kanzure> I see 12:38 < kanzure> so the search algorithm just needs to have some weights that the user can adjust I guess 12:39 < fenn> btw have you seen youtube.com/watch?v=ZaHQ3ZYy4IE.flv and youtube.com/watch?v=u7KIp7A8iqs.flv 12:39 < fenn> oops 12:39 < fenn> s/.flv// 12:39 < kanzure> neat. no, I hadn't 12:40 < kanzure> there was one that I showed you a while back that was proprietary and was doing the typical egg-pick-up and such 12:40 < fenn> the shadow robot hand? 12:40 < fenn> same technology, festo just copied and made a better arm 12:41 < kanzure> heh, a dumbell 12:41 < kanzure> they so need to show that robot operating a lathe or mill 12:41 < fenn> i wonder why it moves so slow 12:41 < fenn> airmuscles can move really quickly 12:41 < kanzure> afraid of breaking it? 12:44 < kanzure> hm, crazy mad scientist time: they need a way to convert a given 'muscular definition file'- of some random model- and then a converter to to translate it into an air muscle design 12:47 < kanzure> even with the "tag an interface as a handle" method, you still need to do 'translational constraints' or some way to specify the required number of degrees of freedom 12:49 < kanzure> otherwise you get a search result that can fit the handle but just sits there 12:49 < kanzure> imagine a giant block of granite but with a sculpted handle-holding-part on one of the faces :p 12:50 * fenn imagines 12:50 < kanzure> has zero DOF 12:50 < kanzure> *DOFs 12:50 < kanzure> so there needs to be a way to distinguish it, I mean 12:50 < kanzure> but if there's a way to do it without 'dolls' .. since you figure it's just me making things more complicated .. 12:51 < fenn> the doll should be able to figure out how to grasp the handle 12:52 < fenn> you shouldn't provide grasp trajectories because they change all the time anyway 12:52 < fenn> similarly for how to lay out a lab bench, or how to move pipette tip around 12:53 < fenn> you just say what you want done "move this over there" 12:53 < fenn> or even better "this should be over there now" 12:53 < fenn> and at runtime the doll says 'hey i can do that' 12:53 < kanzure> wha? 12:53 < kanzure> and at runtime the granite block says "mu" ? 12:54 < fenn> or simulation time or whatever 12:54 < kanzure> so why did you spend all that time building the granite block? 12:54 < fenn> dunno, wtf is this granite block for? 12:54 < kanzure> it matched the handle 12:54 < fenn> wall-climbing exercise? 12:54 < kanzure> nope, it had a hand-sculpture on one of its surfaces 12:54 < kanzure> and it happens to match the pipette handle 12:54 < fenn> was it tagged as a handle? 12:55 < kanzure> the pipette had a part tagged handle yeah 12:55 < fenn> if you want to have a reasonable chance of ever getting around to evaluating the sculpture's suitability as a handle, you should tag it as a handle 12:55 < kanzure> ok, so fine, it's tagged handle too 12:55 < fenn> otherwise, dont blame me for not being able to read your mind 12:55 < kanzure> my point is that it just sits there 12:56 < fenn> why, because it's heavy? 12:56 < kanzure> because it's a block of granite 12:56 < kanzure> it doesn't do anything really 12:56 < fenn> sure it does 12:56 < kanzure> oh? 12:56 < fenn> ever heard of a surface plate? 12:56 < kanzure> yes? 12:56 < fenn> it just sits there 12:56 < fenn> but for some reason people pay thousands of dollars for them 12:57 < kanzure> so why would the doll be chosen over the granite block 12:58 < fenn> because the doll is a positioner 12:58 < kanzure> how does the computer know that? 12:58 < fenn> that's one of its functionalities 12:58 < kanzure> that didn't help :( 12:58 < fenn> it's tagged that way 12:59 < kanzure> so a linear actuator is also a positioner 12:59 < fenn> the positioner class allows you to move things within their limits 12:59 < fenn> yes a linear actuator is also a positioner 12:59 < kanzure> within what limit? 12:59 < kanzure> so you need a way to distinguish positioners 12:59 < kanzure> .. 12:59 < fenn> force limits, accuracy limits, travel limits 12:59 < kanzure> what is a travel limit? 12:59 < fenn> work envelope 13:00 < kanzure> but I thought we agreed it changes with respect to force exertion 13:00 < fenn> how far you can move your arm 13:00 < kanzure> depends on how much I'm lifting/pushing 13:00 < fenn> yeah it's a complex problem 13:00 < kanzure> so tagging it positioner doesn't help 13:00 < fenn> yes it does 13:00 < fenn> it cuts down the search space from literally everything in the database, to just positioners 13:01 < fenn> presumably the search heuristics would look at things you already have first 13:01 < kanzure> um, on another note, 13:01 < kanzure> I've never bought a linear actuator 13:02 < kanzure> but does it come with information on how much force it can exert or how it changes with respect to motion 13:02 < kanzure> blah, that's hard to describe 13:02 < kanzure> in the case of the Festo hand+arm, you can't lift anything by the fingers that can't be lifted by the first link/arm-part 13:02 < fenn> me either 13:02 < fenn> unless you count a paintball gun :P 13:03 < fenn> linear actuators are usually constant force with respect to travel distance 13:04 < fenn> airmuscles are unusual in this respect 13:04 < fenn> ignoring the whole complicated arm kinematics for a sec 13:04 < fenn> the airmuscle exerts the most force when it's fully extended 13:04 < fenn> and yes, festo provides graphs showing that 13:05 < fenn> but what you're talking about with the hand/arm is totally different 13:05 < fenn> that has to do with wrench space and changes in leverage and moment of inertia as the arm moves around 13:07 < kanzure> ok, how about this, what if we classified work envelopes, and we just said that such-and-such is compatible with a class-blah work envelope 13:07 < fenn> most robot arms and human arms have similar extension/leverage characteristics, but for example a hexapod robot is also nonlinear but has totally different kind of nonlinearity 13:07 < kanzure> and a work envelope classification would involve something like a range of masses and motions 13:07 < kanzure> "here's what this work envelope looks like. does this look like your work envelope?" and later we'll write some better software 13:07 < fenn> sure that would probably work fine in practice 13:08 < fenn> 'work envelope' means the range of motion in terms of distance 13:08 < fenn> payload and acceleration are described separately 13:08 < kanzure> ok, so we need a force-work envelope or something 13:08 < kanzure> erm 13:08 < fenn> no need to lump it all together 13:08 < kanzure> I think it should be described together as a package 13:08 < fenn> why? 13:08 < kanzure> because payload limits change with respect to acceleration and distance 13:09 < kanzure> it's coupled. 13:09 < fenn> but a 1kg payload robot is guaranteed to be able to move 1kg across its work envelope 13:09 < kanzure> o rly? 13:09 < kanzure> ok, great 13:10 < fenn> it could probably do more in certain poses/distances but that's too hard for stupid humans to figure out 13:10 < kanzure> kind of like "overclocking" or pushing the performance limits 13:10 < fenn> not really 13:10 < kanzure> which might not actually be really "pushing it"- but stupid humans and such don't know that ;-) 13:10 < kanzure> oh 13:10 < kanzure> hrm 13:10 < fenn> i mean it's like not trying to pick up a jug of milk with your arm fully extended 13:12 < kanzure> I think I can go write some code now 13:22 < kanzure> the latest xkcd is cute. 13:41 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has joined #hplusroadmap 13:50 < kanzure> aw, so I was making some simple models in heekscad and it keeps crashing :( 13:50 < kanzure> time to update. 14:22 < kanzure> fenn: um. importing OCC in a python script is a lot of overhead. I don't know how else to get surfaces/faces though :/ 14:22 < fenn> dont import all of OCC at once 14:22 < fenn> just get what you need 14:23 < fenn> from OCC import FaceStuffs 14:23 < kanzure> and FileReaderStuffs 14:24 < kanzure> what I'm doing is getting a list of surfaces and seeing if it makes sense to reference the internal OCC face IDs with the constraints that are supposedly 'attached' to certain faces/interfaces 14:24 < kanzure> which will really make the system throw up every time a model is changed significantly (oh well?) 14:29 < kanzure> hrm, I wonder why "from OCC import BRepPrimAPI" also imported OCC.IGESControl_Writer 14:29 < kanzure> erm, I mean, OCC.IGESControl_Controller() 14:30 < kanzure> oh, I see, it's assuming InteractiveViewer.py 14:50 < kanzure> fenn, do you know which version of pythonocc might be stable? 14:50 < kanzure> I tried out some of the topology stuff, but the faces stuff doesn't work, although on 2009-04-03, Jelle posted some example code, but svn up isn't giving me code that compiles. 14:52 < fenn> none are stable 14:52 < fenn> where's the code? 14:53 < kanzure> search the mailing list archives for BRepFeat 15:00 < fenn> BRepFeat.py never gets built 15:02 < kanzure> oh? 15:03 < kanzure> I see BRepFeat.i 15:03 < kanzure> I also see BRepFeat.py in pythonOCC/src/OCC/ 15:03 < kanzure> hm, there's nothing in there though really 15:04 < kanzure> why is SWIG separated into two folders (SWIG_src_modular_linux_darwin and SWIG_src_modular_win32) ? 15:06 < fenn> because he runs the generator script and puts the output in those folders (instead of just having people build the swig files on their own system) 15:06 < fenn> doesn't make sense to me, and just seems to screw things up more often 15:14 < kanzure> if I remove the swig files, will they be regenerated by the scons/make script? 15:14 < fenn> maybe 15:15 < fenn> that's my understanding but it's hard to search for manual changes vs just things he changed in the generator script 15:19 < fenn> i keep getting __PYTHONOCC_MAXINT__ undefined in AIS_wrap.cc 15:19 < kanzure> me too 15:21 < fenn> ./environment.py: ('__PYTHONOCC_MAXINT__',sys.maxint) 15:21 < fenn> maybe that needs to go in SConstruct 15:27 < fenn> add that to DEFINE_MACROS in SConstruct 15:30 < kanzure> I don't see how. DEFINE_MACROS=['WNT','WIN32','_WINDOWS','CSFDB','NDEBUG'] 15:30 < kanzure> oh 15:31 < kanzure> blah 15:31 < kanzure> okay, nevermind, I see it in the linux side of things :) 16:11 -!- kanzure- [n=bryan@66.112.232.249] has joined #hplusroadmap --- Log opened Sun Apr 12 16:38:21 2009 16:38 -!- kanzure [i=bryan@66.112.232.233] has joined #hplusroadmap 16:38 -!- Irssi: #hplusroadmap: Total of 21 nicks [0 ops, 0 halfops, 0 voices, 21 normal] 16:38 -!- Irssi: Join to #hplusroadmap was synced in 1 secs 16:49 < kanzure> fenn: updating pythonocc didn't seem to make it work, though I don't know why I thought it would 16:50 < fenn> well, that's not surprising 16:51 < kanzure> how do you know BRepFeat isn't being included? 16:54 < kanzure> oh. 16:54 < kanzure> BRepFeat.i in the windows version has the word 'Face' in it, but in the linux_darwin/ dir, that word doesn't appear in the file 16:54 < kanzure> hrm 16:59 < kanzure> that's probably not much of an indicator of anything being wrong though 17:05 < fenn> do you know anyone named emily daniels? 17:05 < fenn> oh hmm nevermind, just a hash collision 17:05 < fenn> emily_daniel != emily_daniels 17:10 < kanzure-> what you be doing? 17:25 < kanzure-> I wish all these programmers that really only program on windows would just admit that they are using ubuntu discs to log in every once and a while and check compatibility with linux or something, instead of advertizing it as completely linux compatible or something 17:32 < fenn> apparently my mom talks to Sung won Lim (4phlebas on diybio) on twitter, so she forwarded him to me.. one of his twitter friends is 'emily daniels', and one of my contacts in hotmail from god knows when is emily_daniel and i have no idea where it came from 17:32 < fenn> he's also following david nunez, so it wasnt totally unthinkable 17:35 < fenn> wow firefox is "so cool" ~~ 17:35 < fenn> AutoCopy. The basic use: It copies anyt text you select on the web as soon as you select it—no Ctrl+C necessary. For pasting into text forms, you simply hit the middle mouse button rather than Control+V. 17:35 < fenn> ironically it's always firefox that fucks up my x selection buffer 17:41 < fenn> if i say "star system" does that mean (to the layperson) a star and the crap floating around it, or multiple stars orbiting each other? 17:43 < kanzure-> single star and other crap floating around it 17:44 < kanzure-> star cluster is the cluster of stars. 17:44 < kanzure-> odd that your mom knew a diybio fellow 17:45 < kanzure-> maybe it's just an effect of being in college, but everything seems different. more crazy-tech getting in the news. more people knowing about transhumanism/singularity stuff. more people interested in diybio. what's going on? 17:48 < fenn> you're asking the wrong guy 17:48 * fenn mumbles something about 2012 18:00 < kanzure-> pfft 18:27 -!- fenn_ [n=fenn@cpe-72-177-52-180.austin.res.rr.com] has joined #hplusroadmap 18:43 < fenn_> sung won lim seems pretty cool http://bookhling.wordpress.com/tag/transhumanism/ 18:43 -!- fenn [n=fenn@cpe-72-177-52-180.austin.res.rr.com] has quit [Nick collision from services.] 18:43 -!- fenn_ is now known as fenn 18:44 < kanzure> nice. 18:50 < kanzure> hm, copying BRepFeat_* and BRepFeat.i from Windows, plus the dependencies that it demands (of the other packages, which you can discover by running scons and letting it fail), seems to let it work. I'm currently compiling via python setup.py install to check to make sure it actually installs or something this time, rather than just assuming my scripts are failing 19:29 < kanzure> neat, it worked 19:29 < fenn> would you dig through your history file and see what files those are? 19:30 < kanzure> how do I get multiple xterms and konsoles to save my histories right? 19:31 < fenn> close them out, the file is written when the shell exits 19:31 < fenn> it will be jumbled though, not chronological 19:31 < kanzure> hrm. okay. 19:32 < fenn> also you have to set something like HISTTIMEFORMAT="%k:%M; HISTSIZE=1000000 19:32 < fenn> "%k:%M " i mean 19:33 < fenn> to save times to the file 19:33 < genehacker> http://www.newscientist.com/article/dn12056 19:33 < fenn> not a very reliable method but i dont know how else to do it 19:33 < genehacker> liquid neural interface 19:35 < kanzure> http://www.pythonocc.org/wiki/index.php/Install_Linux 19:35 < kanzure> fenn: there you go. 19:35 < fenn> tak 19:35 < kanzure> ? 19:36 < fenn> it's malaysian or something 19:36 < fenn> "thanks" 19:37 < kanzure> genehacker: How does it keep the monolayers separated for separate electrodes? What's the resolution? 19:37 < fenn> and danish too, apparently 19:39 < kanzure> the paper: http://heybryan.org/books/papers/Electrochemical%20polymerization%20of%20conducting%20polymers%20in%20living%20neural%20tissue.pdf 19:40 < genehacker> no idea 19:41 < genehacker> it's so your neurons don't get scarred 19:41 < kanzure-> yeah 19:42 < fenn> one way to get high resolution is to make the electrodes as tubes and seep monomer liquid into the tissue around each tube 19:42 < kanzure-> I was thinking it would be neat if there'd be a way to have a high voltage mode where the electrodes would get super hot to polymerize a polymer 19:42 < kanzure-> but that's probably deadly. 19:47 < fenn> water catalyzed reactions could work 19:47 < fenn> but those are often gnarly chemicals 19:52 < genehacker> yeah no shit kanzure 19:56 < genehacker> I don't want to cook my cortices 19:57 < kanzure-> genehacker: when I get some more free time over the summer, maybe I'll show you how to design a brain implant 19:59 < genehacker> cool 20:04 < fenn> insert rfid chip + coiled antenna-wire into a syringe; drill hole in skull; insert under meninges 20:04 < kanzure-> um, rfid chips have electrode arrays? 20:04 < fenn> got a better idea? 20:05 < kanzure-> http://heybryan.org/books/Biology/neuro/ 20:05 < kanzure-> many? 20:05 < fenn> well, pick one 20:05 < kanzure-> microelectrode array 20:05 < fenn> and how do you make that, and how do you implant it 20:05 < kanzure-> you implant it very carefully :( 20:05 < kanzure-> you get it via the link genehacker gave you 20:06 < kanzure-> $80 for a somewhat large array 20:06 < fenn> 'large' meaning what precisely? 20:06 < genehacker> you sterilize so you don't get a nasty brain infection 20:06 < kanzure-> > 10x10 20:07 < fenn> also i'm not terribly concerned about price for a brain implant.. 20:07 < fenn> as long as it's <$100k 20:07 < kanzure-> since when do you have teh monies? 20:07 < fenn> i mean it's not a frickin USB gadget 20:09 < fenn> kanzure: how much do you think the surgery would cost alone? 20:09 < fenn> i'm talking real surgery here, not power drill + buddy with some ice 20:09 < kanzure-> how about some medical student without a medical license, or doctor in foreign country? 20:11 < fenn> well? 20:12 < kanzure-> I haven't asked them how much it would cost 20:13 < fenn> i'm guessing around $5-10k 20:14 < fenn> maybe lots more 20:15 < kanzure-> hm, I don't know how much brain surgery costs 20:16 < kanzure-> also, what happened to all that talk about ultrasound, infrared, or rTMS? 20:16 < kanzure-> much less costly (for the moment) 20:20 < fenn> well, they aren't implants, for one.. 20:28 < kanzure-> yes, but if you wanted to really play with implants, you'd do some culturing or testing on rats or rabbits or something first 20:28 < kanzure-> what's wrong with the "a buddy and some ice" approach? 20:33 < fenn> you're likely to die or get an infection 20:33 < fenn> also hard to find said buddy 20:33 < kanzure-> say, .. buddy. 20:34 < kanzure-> nudge nudge wink wink 20:34 < fenn> why dont i have OCC.Utils? 20:34 < kanzure-> hm 20:34 < kanzure-> did I do something special? 20:34 < fenn> sorry i'm not drilling into your skull 20:36 < ybit> i wonder what the protocols are in China for ordering certain chemicals 20:36 < kanzure-> probably a few drowned female babies 20:37 < ybit> lol 20:37 < ybit> i would like to live there to know and study Mandarin on the side, but I wouldn't go over there first before knowing what their laws are concerning these things 20:37 < ybit> -"to know and" 20:37 < ybit> um, +to 20:37 < ybit> :P 20:39 < ybit> i want to learn another language, it's something i've thought about for awhile now. i think 2 is enough, more than that and your just wasting your time imo 20:40 < ybit> but english is the de facto for communication, even the hong kong public library has english on the front 20:40 < ybit> it's statue in the front says "the world is my book", my or our i forget 20:40 < ybit> its* 20:41 < ybit> any insights, surely some of you have thought about this 20:41 < fenn> i have thought about moving to china 20:42 < kanzure-> firewall 20:42 < fenn> had intended to learn mandarin but then i forgot the name of the asian culture house or whatever, and couldnt find the schedule for free classes 20:43 < fenn> kanzure: firewall is really no big deal in practice 20:43 < kanzure-> what is the technical spec of their firewall anyway? 20:43 < ybit> mandarin is a hideous language though, if it and their culture were more like japanese, that would be great 20:43 < fenn> just doesn't forward packets from certain IP's 20:44 < fenn> ybit surely you know that half of japanese is derived from mandarin 20:44 < ybit> the kanji, sure, but the rest makes it easy to pick up the language quickly 20:45 < fenn> http://en.wikipedia.org/wiki/Sino-Japanese_compound 20:45 < fenn> its sort of like saying 'half of english is derived from latin' though 20:45 < ybit> the rest: katakana and hiragana 20:48 < kanzure> "raising a string exception is deprecated" 20:48 < kanzure> blah 20:49 < kanzure> huh, face.Complement() 20:49 < fenn> swap the normals 20:49 < ybit> fenn: i know what you're talking about when mentioned the chinese learning centers, and i can't find them, looking for it still 20:49 < fenn> ybit: it was something associated with indiana university 20:50 < ybit> found it: Confucius Institutes 20:50 < ybit> http://en.wikipedia.org/wiki/Confucius_Institute 20:50 < fenn> if that's not a missionary i dont know what is 20:51 < kanzure> neat, face.IsSame(other_face) 20:51 < ybit> it's sponsored by the chinese government, it's not a missionary :P 20:52 < kanzure> http://www.opencascade.org/support/products/coldet/ 20:52 < kanzure> blah, screw them 20:52 < ybit> http://en.wikipedia.org/wiki/Confucius_Institute 20:52 < ybit> hmm, did i just paste that again... 20:52 < ybit> don't know because i cleared the screen by accident 20:53 < ybit> anyway, seems the indiana has a one on campus, and i bet that's where the classes were 20:53 < fenn> no, that's IUPUI which is 50 miles north 20:53 < kanzure> so now I'm confused about collision detection because that page doesn't make it sound easy 20:53 < kanzure> but I'm used to making simple collision detection algorithms 20:53 < fenn> this wasnt a chinese specific thing, they taught all sorts of languages 20:56 < kanzure> maybe I could do a volume/displacement method 20:57 < kanzure> or, actually, not implementing collision detection for now is ok 20:58 < ybit> kanzure: i'm guessing you are simulating parts of an exoskeleton, right? 20:58 < kanzure> nope 20:58 < kanzure> part mating for skdb. 20:58 < kanzure> I even have a (small) sample catalog of parts made up of pegs and holes (I know, I know, the world sucks) 20:59 < fenn> so tell me, does a round peg fit into a square hole? 21:00 < kanzure> what type of material? 21:00 < fenn> wood 21:00 < kanzure> can it be forced into shapes with only 1 to 100 Newtons of force? 21:01 < fenn> depends on the size 21:01 < kanzure> wait, am I the one who is supposed to be right here? or are you trying to bring me down? 21:01 < kanzure> I forget 21:01 < fenn> dunno 21:02 < kanzure> sometimes I've put round pegs in square holes, yes 21:07 < kanzure> neat, moving a face doesn't separate it from its object. 21:08 < ybit> thinking i may give up on mandarin. i couldn't trust others to interpret the chinese laws correctly, nor could i trust myself for 7-10 years, or until i could read and actually understand the meaning of the literature. 21:08 < ybit> may just wait until i can download the language to my brain :) 21:08 < ybit> or use open rTMS to enhance my memory of their 1k+ alphabet 21:09 < kanzure> there is no evidence of any enhancements with any particular rTMS stimulation pattern 21:09 < kanzure> just saying. 21:09 < kanzure> don't be like the last guy who ran in here wearing a helmet of infrared LEDs claiming to have "positive effects" 21:10 < kanzure> because of his n=1 trial testing "with my cat, Whiskers" 21:10 < ybit> heh 21:10 < ybit> hmm, no evidence... 21:10 * ybit searches 21:10 < kanzure> well, there is some evidence of really interesting things happening 21:10 < kanzure> like arms flailing around, regions of the brain shutting off etc 21:10 < kanzure> but that's not necessarily enhancement 21:10 < ybit> yeah yeah, i've seen that 21:11 < ybit> not what i'm going by though 21:11 < ybit> not even sure what i'm going by, i could have swore i saw a paper on this... 21:12 < kanzure> see: http://heybryan.org/~bbishop/docs/open-rtms/ 21:13 < ybit> http://scholar.google.com:80/scholar?q=memory+rtms&hl=en&lr=&ie=ISO-8859-1&btnG=Search 21:13 < ybit> Statistical comparison of memory performance during rTMS and baseline testings 21:13 < ybit> without stimulation revealed no significant chan 21:13 < ybit> :\ 21:13 < ybit> hanges.* 21:13 < ybit> http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6T0D-3VM6JW8-4&_user=10&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=9324bfb03c3576a5dca3584c54ed6ef9 21:13 < ybit> bah, i can get that paper 21:14 < kanzure> ok, maybe this: http://heybryan.org/~bbishop/docs/open-rtms/Repetitive%20Transcranial%20Magnetic%20Stimulation%20Dissociates%20Working%20Memory%20Manipulation%20from%20Retention.pdf 21:16 < kanzure> " 21:16 < kanzure> Personality traits are correlated with motor cortex excitability a transcranial magnetic stimulation (TMS) study" 21:16 < kanzure> neat. 21:16 < kanzure> "Modulation of the neuronal circuitry subserving working memory in healthy human subjects by repetitive transcranial magnetic stimulation 21:18 < kanzure> "Localization of the motor hand area using transcranial magnetic stimulation and functional magnetic resonance imaging" <- bwahaha 21:21 < ybit> "Consensus: Can transcranial direct current stimulation and transcranial magnetic stimulation enhance motor learning and memory formation?" http://filebin.ca/thjmft 21:22 < ybit> sciencedirect's search is terrible 21:24 < kanzure> http://heybryan.org/~bbishop/docs/open-rtms/?C=M;O=D 21:24 < kanzure> ok, added it 21:24 < kanzure> yeah, it is. 21:24 < kanzure> google scholar for the win 21:25 < kanzure> btw, I have a subscription to all sciencedirect journals that arrives in my inbox- the equivalent of their "notify me of new interesting shit" and "notify me of anything new", so I have a rather large (10k+) email archive collection of various goings on 21:26 < ybit> http://filebin.ca/pwbryh 21:26 < ybit> Transcranial magnetic stimulation can measure and modulate 21:26 < ybit> learning and memory 21:26 < ybit> """ 21:26 < kanzure> "Visual working memory revealed by repetitive transcranial magnetic stimulation" 21:27 < ybit> anyway, i suppose this is why i was thinking rTMS could help me with Mandarin 21:28 -!- cis-action [n=cis-acti@146-115-127-170.c3-0.smr-ubr1.sbo-smr.ma.cable.rcn.com] has quit [] 21:29 < kanzure> haha, I love this: http://heybryan.org/~bbishop/docs/open-rtms/Transcranial%20magnetic%20stimulation-induced%20switch%20into%20mania%20-%20a%20report%20of%20two%20cases.pdf 21:29 < kanzure> Case 1, she's awesome. 21:30 < kanzure> where can I sign up to get her? 21:30 < kanzure> case 2 is also interesting 21:31 < kanzure> too bad I'm not depressed. hrm. 21:31 < kanzure> how does one become depressed? do I have to "go goth"? 21:31 < kanzure> (I'm kidding, of course) 21:33 < ybit> heh 21:45 < fenn> 'increased goal-directed activity' is a symptom? 21:48 < kanzure> heh 21:48 < kanzure> bad boy! bad! silly human, having goals! 21:51 -!- ybit [n=heath@unaffiliated/ybit] has quit [Read error: 104 (Connection reset by peer)] 21:52 -!- ybit [n=heath@unaffiliated/ybit] has joined #hplusroadmap 21:55 -!- any52619123 [n=someone@75-120-35-130.dyn.centurytel.net] has joined #hplusroadmap 21:59 < kanzure> "Prolonged peripheral nerve stimulation induces persistent changes in excitability of human motor cortex" 22:00 < fenn> is that a good thing? 22:01 < kanzure> dunno, google scholar doesn't come in "good/bad"-color-vision 22:10 -!- any66208037 [n=someone@75-120-38-93.dyn.centurytel.net] has joined #hplusroadmap 22:10 -!- katsmeow-afk [n=someone@75-120-21-158.dyn.centurytel.net] has quit [Nick collision from services.] 22:10 -!- any66208037 is now known as katsmeow-afk 22:13 < ybit> ;P 22:13 < ybit> meh, wrong emoticon, whatever. good joke though :) 22:15 < genehacker> I think it is bad 22:16 < kanzure> don't go wild just yet, but: 22:16 < kanzure> http://heybryan.org/books/papers/bibliographies/rTMS.bib 22:16 -!- any25030207 [n=someone@75-120-22-51.dyn.centurytel.net] has joined #hplusroadmap 22:17 < ybit> i'm pretty sure i have everything public on heybryan 22:18 < kanzure> nah, /books/papers/bibliographies/ is new. 22:18 < ybit> but thanks for the heads up, it takes awhile to cipher through all the docs to find what i need 22:18 < ybit> how recent? 22:18 < kanzure> um, four hours ago? 22:18 < ybit> okay then, i don't have it 22:18 < kanzure> yeah .. 22:21 -!- katsmeow-afk [n=someone@75-120-38-93.dyn.centurytel.net] has quit [Read error: 60 (Operation timed out)] 22:25 -!- any78219784 [n=someone@75-120-35-28.dyn.centurytel.net] has joined #hplusroadmap 22:28 -!- any52619123 [n=someone@75-120-35-130.dyn.centurytel.net] has quit [Read error: 110 (Connection timed out)] 22:31 -!- any25030207 [n=someone@75-120-22-51.dyn.centurytel.net] has quit [Read error: 60 (Operation timed out)] 22:33 -!- any78219784 is now known as katsmeow 22:37 -!- katsmeow is now known as katsmeow-afk 22:40 -!- jm|space [n=jm@p57B9D8BB.dip.t-dialin.net] has quit [No route to host] 22:45 -!- any92105174 [n=someone@75-120-32-71.dyn.centurytel.net] has joined #hplusroadmap 22:51 -!- katsmeow-afk [n=someone@75-120-35-28.dyn.centurytel.net] has quit [Read error: 60 (Operation timed out)] 23:05 < kanzure> does anyone know a good source for acquiring "velvet bean"? 23:05 < kanzure> ybit: that rTMS.bib file has been updated. 23:08 -!- genehacker [n=chatzill@wireless-128-62-99-202.public.utexas.edu] has quit [Read error: 110 (Connection timed out)] 23:09 < kanzure> where are weight-dosage charts published for pharmaceuticals? 23:22 -!- any26168626 [n=someone@75-120-43-250.dyn.centurytel.net] has joined #hplusroadmap 23:27 -!- any26964776 [n=someone@75-120-31-92.dyn.centurytel.net] has joined #hplusroadmap 23:28 -!- any92105174 [n=someone@75-120-32-71.dyn.centurytel.net] has quit [Read error: 60 (Operation timed out)] 23:45 -!- any26168626 [n=someone@75-120-43-250.dyn.centurytel.net] has quit [Read error: 110 (Connection timed out)] 23:50 -!- any51719035 [n=someone@75-120-31-27.dyn.centurytel.net] has joined #hplusroadmap