--- Day changed Mon Nov 17 2008
03:59 < kanzure_> http://www.miamiherald.com/459/story/773145.html 'The 16-year-old girl was mauled by a cougar and rushed to the hospital after she and 21-year-old Anthony Zitnick illegally entered a wildlife collector's home about 1:30 p.m. Neighbors believe Zitnick was trying to impress the younger girl."
03:59 < kanzure_> That just made my day a little brighter.
04:00 < kanzure_> "Dried blood covered Rigerman's concrete floor at the spot where the 4-year-old cougar pounced on the girl and clenched his jaws around her head."
04:00 < kanzure_> "'I heard her scream, `I'm gonna die, I'm gonna die,' '' he said, hours after paramedics had left."
04:01 < kanzure_> "At first, Miralles said, he tried to push Chaos off, but he wouldn't budge. He then resorted to punching and kicking the cougar, who quickly balled up in a corner, he said."
04:08 < fenn> so, which character in this drama would you most like to be?
04:08 < fenn> the cougar, the owner, the guy, or the girl
04:08 < kanzure> poor sap who had to straighten out the story
04:09 < kanzure> "Wait, the cougar's name is chaos? Well that's just flat out improbable."
04:09 < fenn> ''I'm worried about my absolutely spotless reputation with what I do with these animals,'' 
04:10 < kanzure> the only way that the sentence parses is if he means to say that he does terrible things with the animals
04:18 < ybit> http://pewinternet.org/datasets.asp
04:19 < kanzure> Doesn't pew do internet surveys?
04:19 < fenn> can anyone spot what i'm doing wrong?  foo = re.match('nan', '\tDefinition: nan\n')
04:19 < fenn> it should match but it doesnt
04:20 < kanzure> python?
04:20 < fenn> ya
04:20 < kanzure> is this find-and-replace?
04:20 < fenn> no, it should return a match object
04:20 < kanzure> what's the second parameter?
04:20 < fenn> the stuff i'm matching the pattern against
04:21 < fenn> bar = re.match('a', 'abc') works
04:21 < bkero> match only does the beginning of a string
04:21 < bkero> Try search
04:21 < bkero> match implies ^
04:21 < bkero> So you're actually doing foo = re.match('^nan', '\tDefinition: nan\n')
04:23 < fenn> i wonder how i managed to avoid that with every other regex i've written
04:25 < procto> oh shit... `i'm pickling an object and python-' memory usage is at 350mb
04:25 < procto> i'd have though cPickle would not be that heavy...
04:25 < bkero> lol pickles
04:26 < procto> shit, it takes as long to unpickle as to generate the object in the first place...
04:48 < drazak> so
04:49 < drazak> when doing a genetic sequencing, we cleave chromosomes in half, and then (Generally) preform double barrel shotgun sequencing
04:49 < drazak> what if we left them whole, and did a 4 barrel squencing, using more computer power and allowing for checks bassed on percent difference from results in the hg project
04:49 < drazak> you'd be atleast twice as fast, probably even twice as fast -1/3rd
05:08 < gene> So Kanzure do you know about Tipler?
06:59 < fenn> trying to preserve the original units (inch, celsius etc) is quite a pain in the ass
07:11 < kanzure> For some reason I decided to go to sleep at three yesterday.
08:22 < kanzure> gene: Yes, I am aaware of Tipler.
08:34 < kanzure> fenn: Under what circumstances are units being changed?
09:04 < kanzure_> http://heybryan.org/mediawiki/index.php/Towards_automatic_generation_of_linear_genetic_regulatory_networks_via_biobricks
09:06 < kanzure> So if the codebase is finished before that gets published, I just remove the "Towards" from the title.
09:14 < kanzure_> http://www.biomedcentral.com/1752-0509/2/91/abstract  <-- Hop on the gravy train.
09:15 < kanzure_> oh wait, 50 pg pdf. Nevermind.
09:57 < kanzure> fenn: if it makes you happy, ADL references lisp on the site.
09:57 < kanzure> I was trying to go through the papers there to steal images but now I'm realizing that I can't even get the majority of the papers cited ..
09:57 < kanzure> and I supposedly "have access"
10:53 < kanzure_> http://www.youtube.com/watch?v=NKCZ9jRCYT0 3-O-Matic - success
14:07 < UtopiahGHML> anybody know studies on the most human active gene pools? what genes are currently changing the fastest and what are the hypothesised environmental pressure that pushes this phenomenum?
14:07 < UtopiahGHML> maybe sth from Bruce Lahn (Chicago Uni)
16:05 < UtopiahGHML> Id appreciate if someone could share The Mechanical Mind in History, Cambridge, MA: MIT Press. 978-0-262-08377-5 
17:00 < kanzure_-> Hello.
17:25 < gene> you there kanzure
17:35 < kanzure_-> yes
17:36 < kanzure_-> fenn: http://heybryan.org/books/Manufacturing/campbell/ just fyi.
17:57 < kanzure_-> oh, I guess I should have linked to the .zip in the parent dir.
17:57 < kanzure_-> since the filenames are hardly informative.
18:43 < fenn> and they're all empty
18:47 < fenn> i love how todo.txt is from 1999
18:47 < fenn> hmm probably a date error
19:18 < fenn> campbell said no.
19:20 < fenn> i think
19:43 < gene> kanzure
20:01 < kanzure_-> yes?
20:01 < kanzure_-> fenn: yeah, looks like he really wants a grad student
20:02 < kanzure_-> but he replied. which is impressive?
20:02 < fenn> heh
20:02 < fenn> there is that, at least
20:03  * fenn rings up cat-lady to see if she still needs a housekeeper
20:03 < fenn> so i think i should steam ahead with SMIRF
20:03 < fenn> since it's the only way to get anything done these days
20:03 < kanzure_-> what, "the only way to get things done is to throw it into an MMORPG" ?
20:03 < fenn> flashy graphics and social software, that is to say
20:05 < kanzure_-> sounds boring.
20:05 < fenn> billions of people can't be wrong!
20:07 < kanzure_-> huh, so there is an alternative
20:09 < kanzure_-> well, no there isn't. nevermind.
20:09 < kanzure_-> http://www.utexas.edu/ce/college/ doesn't really help with the income factor.
20:12 < kanzure_-> (nor the undergraduate standing)
20:12 < fenn> ideally i'd be able to migrate south for the winter
20:12 < fenn> them geese are smart
20:15 < gene> ok Kanzure did you finish the thing?
20:15 < kanzure_-> gene: The thingy?
20:15 < kanzure_-> What?
20:16 < gene> yeah
20:16 < kanzure_-> What thing?
20:16 < gene> materials and manufacturing report?
20:16 < kanzure_-> that's due next week
20:16 < gene> oh
20:16 < kanzure_-> it was assigned last week.
20:16 < gene> ok
20:17 < gene> hmm.... why not apply an mmorpg to SKDB?
20:17 < gene> put stuff into database
20:17 < gene> get points
20:17 < gene> and win things
20:17 < gene> like hats
20:17 < kanzure_-> because people suck at data entry.
20:17 < kanzure_-> that's why the current repository sucks.
20:17 < gene> indeed
20:18 < kanzure_-> (also because the data entry app sucks too, but that's just because people sucked at understanding what they were doing)
20:18 < gene> ok
20:19 < fenn> gene: that's the idea
20:19 < fenn> behind smirf
20:19 < kanzure_-> yesterday you said to get rid of human data entry
20:19 < fenn> i did?
20:19 < gene> indeed
20:19 < kanzure_-> after commenting on the Mike mockup
20:19 < kanzure_-> arid_shadow comcast link thing that Paul links to
20:19  * fenn has amnesia
20:20 < fenn> what Mike mockup?
20:21 < kanzure_-> the one that showed an entry app to their "OSCOMAK" but it was really just this thing with a giant human language entry form
20:21 < kanzure_-> and a small link on the side saying "Upload CAD file button"
20:21 < fenn> not that kind of data entry
20:21 < fenn> more like Alice/croquet
20:21 < kanzure_-> and then on the left there was stuff like "category blah blahb lha"
20:21 < fenn> http://home.comcast.net/~arid_shadow/interface.png
20:21 < kanzure_-> yes
20:22 < fenn> perhaps my production budget is too low for my ambition, but certainly a more graphical "tactile" interface is desirable
20:23 < fenn> and semantic wikipedia already exists so why reinvent it
20:23 < gene> tactile would be awesom
20:23 < gene> shape parts with our hands
20:23 < fenn> well, nobody has haptic input devices so will have to get by with mouse and keyboard :(
20:23 < kanzure_-> shape isn't a really important factor in all of this
20:23 < kanzure_-> it's one of them, but not really.
20:24 < gene> yeah
20:24 < fenn> there is a lot of "this"
20:24 < gene> unfortunately
20:24 < kanzure_-> what's unfortunately?
20:24 < fenn> gene wants play-doh land
20:24 < gene> tactile interfaces aren't sufficiently cheap yet
20:24 < gene> not play-doh
20:24 < gene> land
20:24 < fenn> i cant say i blame him really, it works for spore
20:25 < fenn> and would work for the "magic" stuff in smirf
20:25 < gene> more like dimension parts with your hands
20:25 < kanzure_-> but we have 3D shape generators already.
20:25 < gene> put lines and stuff down with your fingers
20:25 < gene> just an idea
20:25 < kanzure_-> just a month of running would generate the possibility space of constructable components that users would ever be able to make within their lifetimes
20:25 < fenn> bah
20:26 < fenn> show me
20:26 < kanzure_-> well, give the user a finite set of operations and be amazed
20:26 < kanzure_-> for continuus variables and parameters, like "offset by 2132.5414.141" sure, that's not going to happen via the generators unless that number has some damned good definable reason
20:27 < fenn> ah arid_shadow = mike harris
20:27 < kanzure_-> yes
20:28 < fenn> is "energy minima" a damned good reason?
20:28 < fenn> state space
20:28 < kanzure_-> "energy minima" in terms of what, something like tension?
20:28 < fenn> optimized for some function
20:29 < kanzure_-> no. because you're being too vague here.
20:29 < fenn> like mass or thermal dissipation
20:29 < fenn> or rigidity
20:29 < kanzure_-> k, that looks like end-result parametric tweaking more than anything
20:29 < kanzure_-> since this doesn't do FEA-down-the-tree (whatever that means / looks like)
20:29 < fenn> well, it should
20:29 < kanzure_-> what I mean by that is that it does generate->load into FEA->run FEA on that item
20:29 < kanzure_-> what would "FEA-down-the-tree" look like?
20:30 < kanzure_-> FEA-down-the-tree means that the operations that are allowable are in terms of the FEA-results as well as manipulation of the topology/operations-that-you-can-make-on-the-design
20:30 < fenn> you mean "back propagation"? http://fennetic.net/cadwiki/index.php?back-propagation
20:30 < kanzure_-> if that makes any sense. probably doesn't because I don't know of anybody who has this method.
20:30 < kanzure_-> not quite.
20:31 < fenn> oh so you're treating FEA as a design constraint (but soft)
20:31 < fenn> o yonder whatsit
20:31 < fenn> What light through yonder window breaks? 
20:31 < kanzure_-> hm, the way that you word backpropagation I'd be tempted to say yes
20:32 < fenn> btw i think iteration is way more practical than PDE's
20:32 < fenn> in terms of actually getting something finished
20:32 < kanzure_-> fenn: iteration instead of PDEs?
20:32 < fenn> yeah
20:32 < kanzure_-> does not compute for the way I was going to use PDEs (FEA methods etc.)
20:33 < kanzure_-> or not "going to" but "would use if I were to"
20:33 < kanzure_-> iterate the graph and then given a final 3D cad model, run FEA
20:33 < kanzure_-> erm, not iterate the graph
20:33 < kanzure_-> traverse the tree of all possible graphs up to some point defined by various metrics yay handwaving
20:33 < fenn> why would you do analysis as the last step?
20:33 < kanzure_-> and then take that representative design and run FEA
20:33 < kanzure_-> not quite the last, but at some well-defined points in the tree
20:34 < kanzure_-> like at points that you might be interested in showing the user
20:34 < fenn> decision points
20:34 < kanzure_-> but this isn't because of any sort of "informed search" where the information making up the "informed" aspect is due to FEA information .. I'd like it to be due to FEA precognition information, yes of course
20:34 < kanzure_-> hm
20:34 < kanzure_-> FEA foresight?
20:34 < kanzure_-> FEA precognition information?
20:34 < kanzure_-> projected results?
20:34 < kanzure_-> whatever.
20:35 < kanzure_-> some confluence of the operations in the manipulation of the traversal tree, making it so that you get reasonably useful predictions of what the FEA results are going to be from one manipulation of the design-graph to another.
20:35 < fenn> fea is very low cost with coarse meshes
20:35 < kanzure_-> but pretty inaccurate, no?
20:35 < fenn> but good enough for a first order survey
20:35 < kanzure_-> are you sure?
20:35 < fenn> no
20:35 < kanzure_-> heh
20:36 < kanzure_-> I can approximate a human with 20 surfaces maybe (think pre-N64), but in reality some of those surfaces are complex, high-impact craters that impact FEA results dramatically (I'm making this up)
20:36 < fenn> that was the strategy i was planning on using for magical-cad-app
20:37 < fenn> generally if the topology doesn't work at low res it wont work at high res either
20:37 < fenn> (sometimes it will work at low res but not at high res)
20:37 < kanzure_-> ok, so for a quick check, sure.
20:37 < kanzure_-> I do all kinds of quick checks as it is, so a few more isn't terrible.
20:38 < kanzure_-> Do you know of a python library for exporting 3D data into some format that blender or 3D modeling packages will like?
20:39 < fenn> export from what?
20:39 < kanzure_-> I've been looking around for this but haven't found much for python nor perl. which is awkward.
20:39 < fenn> there was some opengl screen capture to STL hack somewhere
20:39 < kanzure_-> export from a collection of points, vertices, lines, or some other similar primitive objects.
20:39 < kanzure_-> screen capture?
20:39 < kanzure_-> or buffer capture?
20:39 < fenn> buffer capture
20:39 < kanzure_-> neat
20:40 < fenn> stl cant do primitives though, only mesh
20:40 < kanzure_-> that's fine, I can convert to quads
20:40 < kanzure_-> (I think)
20:41 < fenn> i dont think there's much 3d geometry stuff for python
20:41 < fenn> (would love to be proven wrong of course)
20:41 < kanzure_-> same with perl. :/
20:41 < kanzure_-> I wonder why I'm talking about perl/python
20:41 < kanzure_-> the glxgears.c stuff I'm doing is still in c
20:42 < kanzure_-> so uh, I'm sure gamedev.net and nehe will have something.
20:42  * fenn grunts
20:42 < kanzure_-> ?
20:43 < fenn> "good luck"
20:43 < fenn> everyone loves to write importers, nobody writes exporters
20:45 < fenn> praps you could simply use the blender python api
20:45 < kanzure_-> I could generate a python script perhaps
20:45 < fenn> oh right it's still C
20:45 < kanzure_-> nah, generating a script isn't bad
20:45 < fenn> why didnt you use the glxgears.py i linked to :)
20:46 < kanzure_-> does glxgears.py run in blender?
20:47 < fenn> no
20:47 < fenn> there is a blender gears script tho (makes anatomically accurate gears too)
20:47 < kanzure_-> it's not terrible for me to write "print <command>" instead of just calling <command> .. just means that I'll have to generate a .py script, then run blender in batch mode or whatever
20:47 < kanzure_-> hm, I'll have to look int othat
20:47 < kanzure_-> *into that
20:47 < kanzure_-> plus the buffer->STL stuff.
20:48 < fenn> kewl: http://www.youtube.com/watch?v=kcZsMhU7H3o
20:51 < fenn> i dont think you're going to find anything better than this: http://www.swineworld.org/blender/gears/
20:52 < fenn> of course no fscking license info
20:53 < kanzure_-> Hello ferrouswheel.
20:53 < fenn> oh, i guess "(cc) by 3.0 nl" means something
20:54 < ferrouswheel> hi kanzure_ 
20:54  * fenn runs it through his license multiplier matrix </dream>
20:54 < ferrouswheel> hey UtopiahGHML - fancy seeing you here :)
20:55 < kanzure_-> I'm confused. Do we know ferrouswheel?
20:55 < ferrouswheel> No, I've never been in this room, but I work on OpenCog w Ben Goertzel.
20:55 < fenn> sounds familiar
20:55 < fenn> ah that's probably it
20:56 < ferrouswheel> Joel Pitt IRL
20:57 < ferrouswheel> Randomly saw this room in an email by Bryan Bishop, while randomly searching for other things in gmail.
20:57 < ferrouswheel> So thought I'd drop by :)
20:58 < UtopiahGHML> hi ferrouswheel 
20:59 < UtopiahGHML> since you are here and alive ;) have you read The Mechanical Mind in History, Cambridge, MA: MIT Press ? seems pretty interesting but didn't check reviews yet
21:00 < ferrouswheel> UtopiahGHML, no I haven't... recent publication?
21:00 < UtopiahGHML> April 2008 
21:01 < ferrouswheel> Ahh... well. If you do read it, let me know if it's worth buying/reading
21:01  * ferrouswheel already has a backlog of reading to get through
21:01 < ferrouswheel> ...as I suspect many of us do!
21:01 < UtopiahGHML> TOC http://mitpress.mit.edu/catalog/item/default.asp?ttype=2&tid=11479&mode=toc
21:01 < fenn> i'm procrastinating my backlog with my other backlog
21:02  * UtopiahGHML started a meta-todo list recently
21:02 < fenn> my todo list is more meta than yours :P
21:02 < UtopiahGHML> I meta-doubt it!
21:03 < ferrouswheel> I have "finish my todo list" on my meta todo-list... etc
21:04 < fenn> i meta-dog dare you to finish it
21:05 < UtopiahGHML> ferrouswheel: how is the transhumanism/posthumanism/... movement perceived in NewZealand?
21:05 -!- ferrouswheel is now known as metaferrouswheel
21:05 < UtopiahGHML> any cursus? groups? labs? sth?
21:05 < metaferrouswheel> ferrouswheel doesn't know of any such movement :(
21:06 < metaferrouswheel> he plans to try and start a gathering of folks at his house though...
21:06 < UtopiahGHML> maybe some MeetUp group
21:06 < fenn> NZ is disproportionately represented for its population size
21:06 -!- metaferrouswheel is now known as ferrouswheel
21:06 < fenn> as is Boston, San Diego, and Austin
21:07 < ferrouswheel> I miss being in SF - I may have to go on another extended trip.
21:07 < UtopiahGHML> :)
21:07 < UtopiahGHML> the BayArea is pretty cool
21:07 < fenn> bunch of hipsters with too much money, wah
21:07 < UtopiahGHML> and very open on those questions
21:08 < fenn> i guess that's the going definition of "cool" though
21:08 < ferrouswheel> fenn, yes, but having people with too much money leads to them sponsoring crazy projects ;)
21:09 < ferrouswheel> UtopiahGHML, I had a great time hanging out with folks who'd I'd previously only known online.
21:09 < ferrouswheel> Damned expensive to live there though... especially compared to NZ, haha
21:10 < UtopiahGHML> really?
21:11 < UtopiahGHML> well you also have to consider the average local salary in MountainView/PaloAlto
21:13 < ferrouswheel> That's true... however I'm working on not all that amazing rates at the moment (pretty good for NZ, just not so much for SF), so it felt expensive
21:14 < ferrouswheel> plus the NZD is low against the USD at the moment... which I don't quite understand since I though it was the US economy that was in trouble ;P
21:14 < UtopiahGHML> yes I didn't get it either
21:14 < UtopiahGHML> I changed USD this summer
21:14 < UtopiahGHML> and expected to have an even higher gap with euro
21:14 < UtopiahGHML> but it went up
21:15 < UtopiahGHML> (which is good for me since I still have those USD)
21:16 < UtopiahGHML> guess Im not smart enough to play on the currency market yet
21:17 < UtopiahGHML> and also talking about NZ have you tried ThinkBase? from Auckland uni, CS lab http://thinkbase.cs.auckland.ac.nz/
21:19 < ferrouswheel> i haven't... what is it? (complains java isn't installed, but it is... :/ )
21:19 < UtopiahGHML> visual wiki navigation using Freebase (using itself Wikipedia) as a network to explore
21:19 < ferrouswheel> (it should really at least give people the option of finding out about it, even without Java)
21:21 < UtopiahGHML> their blog (not requiring java) is at http://thinkbase-blog.blogspot.com/
21:21 < UtopiahGHML> another NZ project I like is http://www.eurekster.com/
23:01 < kanzure_-> fenn: Around?
23:02 < kanzure_-> fenn: it turns out that he was concerned because your undergrad degree was microbiology.
23:02 < kanzure_-> he should be writing another email to you in a day or two, and if not, I'd say you have grounds to phone him
23:04 < kanzure_-> so I told him you were computer-sciencey, perhaps more than I am (if that's possible) slash mechanical, so you just need to strut or something
23:04  * kanzure_- goes to eat
23:21 < kanzure> yawn, freebase.
23:21 < kanzure> ferrouswheel: So, hi. I'm Bryan.
23:21 < bkero> Google dinnertime in 9 minutes
23:21 < bkero> ferrouswheel: Good evening.
23:21 < kanzure> Why?
23:21 < bkero> kanzure: That wasn't a command.  I was saying dinner at the Google complex occurs at 6:30PM EST.
23:23 < kanzure> Some good that does me.
23:23 < kanzure> google: me a dinner.
23:23 < kanzure> erm, I mean, why can't I google up dinner.
23:23 < kanzure> Scuse me. I need to go wander around.
23:24 < bkero> Go troll your dorm
23:29 < fenn> try sudo google me a dinner
23:30 < bkero> fenn: come to new york and I'll bring you to google dinner
23:30 < fenn> ok be right there..
23:30  * bkero goes to google dinner
23:30 < fenn> hmm beans or lettuce tonight..
23:31 < fenn> oo maybe i'll have _both_
23:44 < kanzure> So what's up with bkero and all the rad jobs?
23:44 < kanzure> Intel, Google, is he a player?
23:56 < bkero> Sort of
23:56 < bkero> I'm just super awesome