--- Day changed Sat Sep 13 2008 | ||
fenn | i'm so uninformed when it comes to things like matlab that i dont even want to know | 00:00 |
---|---|---|
fenn | like, wtf is matlab, why would anyone want to use it? | 00:00 |
kanzure | you want me to answer that? | 00:00 |
kanzure | matlab is all over the place in the labs | 00:00 |
fenn | "matlab allows easy matrix multiplication" but sfw, so does every other progamming language | 00:01 |
kanzure | it's as deeply rooted as perhaps fortran was | 00:01 |
kanzure | "nothing's better than fortran!" | 00:01 |
kanzure | then came matlab or something | 00:01 |
fenn | i thought everyone always hated fortran | 00:01 |
kanzure | mauk wants to rewrite his cerebellum code from visual basic to fortran | 00:02 |
kanzure | "when I was your age, and I entered academia, I too thought fortran was old .. and this was the 70s, mind you" | 00:02 |
fenn | tell mauk to spend 20 minutes surveying the past 50 years in computer language development | 00:02 |
kanzure | no kidding | 00:02 |
fenn | i wish python did lists the way matlab does (without having to do some import numpy or whatever) so that 2 * [1, 2] = [2, 4] not [1, 2, 1, 2] | 00:04 |
fenn | so i ordered one of those folding keyboards | 00:07 |
fenn | perhaps i can arrange things so that i will work on it in order to procrastinate working on something else | 00:07 |
kanzure | keep on procrastinating, soldier | 00:08 |
kanzure | it's good work. | 00:08 |
kanzure | meanwhile /me is still trying to figure out what it was that he was going to be doing | 00:08 |
fenn | good luck with that | 00:09 |
kanzure | I'm not sure I should write the skdb code stuff now, it looks like the lab's software sorta kinda does it | 00:09 |
* fenn off to interact with real world for a bit, sorta | 00:09 | |
kanzure | stalking? | 00:09 |
fenn | going to grocery w/roomie | 00:09 |
kanzure | does anyone remember their statistical thermodynamcis? | 00:29 |
kanzure | *thermodynamics | 00:29 |
kanzure | In particular, Boltzmann and his ensemble of particles; the importance of being able to set fire to a tree and get ashes and not be able to reconstruct the tree from the information within the ashes, and so on? | 00:29 |
kanzure | since we're just generally interested in matter/energy manipulation processes, I'm wondering if it would be helpful to just work at a slightly deeper level at those system effects rather than at the level of silly part names etc. - an ontology to approaching thermodynamics and | 00:31 |
kanzure | general systems can be selected, and from there different parts are simply the embodiment of conversions from matter/energy into various forms in order to complete given design requirements with a given context specification. Like the conversion of hydrogen clouds into sandwhiches. | 00:31 |
kanzure | Oh, guess that doesn't help the grounding issues. Nevermind, you still can't get a free lunch with those sandwhiches. | 00:31 |
kanzure | (by this I mean you still have to provide enough to tell it what a 'sandwhich' must mean) | 00:31 |
kanzure | though I don't suppose keeping strictly to optimization rules thereof is a terrible idea | 00:32 |
* fenn mumbles something about "new kind of science" | 01:27 | |
fenn | you can't reconstruct the tree because of the heisenberg uncertainty principle | 01:28 |
fenn | if you knew the positions of all particles/wavicles in the tree's light cone, you could reconstruct the tree | 01:28 |
fenn | or at least backtrack through time to see what it was like | 01:28 |
kanzure | I keep having this love/hate relationship with metadata entry :) | 01:29 |
kanzure | anywho, http://oem.cadregister.com/asp/PPOW_Entry.asp?elementID=73333734/HEATING/FANHEAT/CY0007/3&orderedResponses=0|&language=GB&oem=true&referrer=CNDynLink&css=&isWT=n&WTID=&cId=3374&appId=124&lsId=15617&ctId=0&vId=10&uId=&rn=02700.0-00&rm=PTC%20Fan%20Heater&rb=Woot | 01:29 |
kanzure | there's a java applet for viewing the models | 01:29 |
kanzure | meaning the model data is being sent to the client | 01:29 |
kanzure | although java applets apparently don't work on my system | 01:29 |
fenn | "Please install the Java Runtime Environment before refreshing this page. "? | 01:30 |
kanzure | although look under the 'email' select box thingy | 01:30 |
kanzure | no, I just get the java box fadded out thingy | 01:30 |
kanzure | even though I have, in fact, installed the jdk, vm, sdk, all sorts of java stuff | 01:30 |
fenn | what is this thing supposed to do exactly? | 01:30 |
kanzure | http://psv5.cadregister.com/enablement/ENparts/TIN41173.xml | 01:31 |
kanzure | is this just geometry? | 01:31 |
fenn | 404 | 01:31 |
kanzure | '<!-- This XML-document is generated by PowerParts. | 01:31 |
kanzure | It's well-formed and valid against genius_part.xsd XML Schema Definition. -->' | 01:31 |
kanzure | hrm | 01:31 |
kanzure | go back to the link, see the 'download' select box thingy | 01:32 |
kanzure | select 'XML' | 01:32 |
fenn | i guess 'information only' is product lifecycle stuff | 01:32 |
kanzure | doesn't seem to be much functionality | 01:32 |
kanzure | this is a heater unit thingy? | 01:32 |
fenn | well, i cant get it to work either way | 01:33 |
fenn | did i mention i hate web 2.0? | 01:33 |
kanzure | doesn't look web 2.0 to me | 01:33 |
kanzure | what's happening on your end? | 01:33 |
fenn | it's a web app | 01:33 |
fenn | == web 2.0 in my book | 01:33 |
kanzure | oh :) | 01:33 |
fenn | it says 'please wait...' and then goes back to the 'please install jvm' page | 01:34 |
kanzure | web 2.0 started in the 90s for you then :) | 01:34 |
kanzure | what a hell | 01:34 |
kanzure | odd | 01:34 |
kanzure | browser? | 01:34 |
fenn | konq | 01:34 |
fenn | waiting on firepig | 01:34 |
fenn | my laptop seems to be getting exponentially slower | 01:34 |
kanzure | you can make an entire campaign on "fire the fox" | 01:34 |
kanzure | why is it that computers tend to get slow in their lifetime? | 01:34 |
kanzure | even without using other computers, i.e. without a bias effect | 01:35 |
fenn | ok but it's actually iceweasel so, "ice the weasel" | 01:35 |
fenn | it's becausse each webpage has 500 fricking images on it | 01:35 |
fenn | so you get horrrid memory frragmentation, especially in the 256MB of ram that i have | 01:35 |
kanzure | works in ff for me | 01:36 |
kanzure | iceweasel, I mean | 01:36 |
fenn | me too | 01:36 |
kanzure | nope, nevermind | 01:36 |
kanzure | it doesn't give me my popup with the image I requested | 01:36 |
kanzure | opera works though. whatever. | 01:36 |
kanzure | hrm, elementID is a weird variable | 01:37 |
kanzure | guess I'll have to steal the index from thomasnet.com's html pages | 01:37 |
kanzure | but I'm not convinced this data set is useful | 01:37 |
kanzure | what's the point of just giving out the models? | 01:38 |
kanzure | there's not even a "send us money to give you the equations" etc. | 01:38 |
kanzure | maybe it's in the cad files, I haven't looked | 01:38 |
fenn | you mean non-parametric geometry? | 01:38 |
fenn | unlike industrial machinery, cad file format interoperability is a real Hard Problem | 01:39 |
kanzure | what? | 01:39 |
kanzure | I don't know if I mean any sort of geometry | 01:39 |
kanzure | I mean stuff like "here's what this *does*" | 01:39 |
fenn | hence monstrosities like STEP | 01:39 |
kanzure | STEP? | 01:39 |
fenn | STEP is sort of like the union set of all cad file formats | 01:40 |
kanzure | 214 or 213 or what do I want? | 01:40 |
fenn | but written in this awful EXPRESS language | 01:40 |
kanzure | 'STEP unigraphics' ? | 01:40 |
fenn | unigraphics is a cad company | 01:40 |
fenn | STEP is an ISO standard | 01:40 |
fenn | 214 is the second graphics/geometry subset of STEP | 01:40 |
fenn | for automotive/aerospace | 01:40 |
kanzure | is there a non-gfx subset? | 01:40 |
fenn | yes, at least 214 of them | 01:40 |
kanzure | hrm | 01:41 |
kanzure | /* ISO 10303-21 file written by STEP Caselib, ProSTEP GmbH, Germany */ | 01:41 |
fenn | 10303 is STEP | 01:41 |
fenn | 10303-21 is the file format, sorta like XML but not | 01:42 |
kanzure | STEP 203 looks exactly the same .. | 01:42 |
fenn | it's called EXPRESS and was made up solely for describing the STEP standard | 01:42 |
fenn | 21 is just the low level format | 01:42 |
kanzure | I don't see anything but geometry here | 01:42 |
fenn | the information within can be 203, 214, whatever | 01:42 |
kanzure | this tells me very little .. | 01:42 |
fenn | 203 is just geometry | 01:42 |
kanzure | I want non-geo info | 01:43 |
kanzure | argh | 01:43 |
fenn | what sort of non-geo info? | 01:43 |
kanzure | well, "what does this product do" | 01:43 |
kanzure | I don't care what sort of weird ontology it's in | 01:43 |
kanzure | I guess it's on thomasnet.com's stupid 67,000 categories | 01:43 |
fenn | please be patient while my computer cranks away.. | 01:43 |
kanzure | hm? | 01:43 |
kanzure | wtf is this shit? #703=EDGE_CURVE('EDGE85',#683,#667,#702,.T.); #704=ORIENTED_EDGE('COEDGE98',*,*,#703,.F.); #705=EDGE_LOOP('NONE',(#691,#697,#698,#704)); #706=FACE_BOUND('LOOP1',#705,.T.); | 01:44 |
kanzure | terrible :) | 01:44 |
fenn | i havent seen disk thrashing like this since 1999 | 01:44 |
kanzure | same old lappy? | 01:44 |
fenn | no, only had this lappy 2 years or so | 01:44 |
fenn | this s what happens when i run java | 01:44 |
fenn | #705=EDGE_LOOP('NONE',(#691,#697,#698,#704)) | 01:45 |
fenn | see #704, that references #704=ORIENTED_EDGE('COEDGE98',*,*,#703,.F.); | 01:46 |
fenn | those sort of reference links can be parametric | 01:46 |
fenn | oh, that's express format btw | 01:46 |
fenn | er, no actually it's 10303-21, sorry | 01:46 |
fenn | express is like XSD | 01:47 |
kanzure | I know what it's saying, just doubt its optimality for it | 01:48 |
fenn | so there is some non-geometry info if you download non-geometry XML (but not too much) | 01:48 |
fenn | most cad files are just "drafting" not "design" so dont get your hopes up | 01:49 |
kanzure | what use is it then | 01:49 |
kanzure | I mean, just algorithmically generate boxes and stuff | 01:50 |
fenn | it's a way to describe what you want, or what something is | 01:50 |
kanzure | problem solved, why download cad | 01:50 |
kanzure | drafting cad | 01:50 |
fenn | huh? | 01:50 |
fenn | ever try to draw precise boxes in blender? | 01:50 |
kanzure | if you just need polygons, run GenerateSphere(radius); | 01:50 |
* kanzure is confused. are they making drafting harder than it actually is? | 01:50 | |
kanzure | because I know there's more to it than just geometry | 01:51 |
fenn | you understand what a shop drawing is right? | 01:51 |
kanzure | well, no, because surely they come with different material requirements | 01:51 |
fenn | hell, most engineers don't know what a shop drawing is for :( | 01:51 |
fenn | this planet is so fucked | 01:51 |
kanzure | where are they getting the material information about the cad file they might be implementing? | 01:51 |
kanzure | eh? | 01:51 |
fenn | material information is usually marked somewhere on the drawing, it's a pretty standard cad thing | 01:52 |
fenn | sorry if i dont know whatever goofball XML format thomasnet uses | 01:52 |
kanzure | there's multiple file formats that they offer, did you see the list? | 01:52 |
fenn | thomasnet is like ebay for industry | 01:52 |
kanzure | except sucks a little bit more as far as I can tell | 01:52 |
fenn | naturally, everything in industry sucks | 01:53 |
fenn | for some reason the most pig-headed obstinate greedy fucks make it to the top | 01:53 |
fenn | and unlike the military or science, they dont have to actually show any progress | 01:53 |
fenn | just make it look like progress | 01:54 |
fenn | actually thomasnet is more like amazon, alibaba is like ebay | 01:55 |
kanzure | DXF 3D, STEP 214, STEP 203, STEP Unigraphics, ACIS SAT 4.0, ACIS SAT 3.0, ACIS SAT 2.0, IGES, VDA-FS, JPG, BMP, PNG, TIF, PCX, TGA, XML, XML Geometry only, XML Information only | 01:55 |
kanzure | bmp! pick bmp! | 01:56 |
fenn | i like IGES | 01:56 |
kanzure | does it have non-geom info | 01:56 |
fenn | no, it's strictly geometry | 01:56 |
kanzure | gahreloq | 01:57 |
fenn | those are all geometry or picture formats i think (not sure what step unigraphics is exactly) | 01:57 |
kanzure | ok, explain shop drawing and why thomasnet isn't giving me something more useful than just a picture and saying "make up materials and specs while you go kthnxbai" | 01:57 |
fenn | and xml can contain whatever obviously | 01:57 |
kanzure | I checked the xml, it was geom | 01:58 |
fenn | ok, shop drawing is a piece of paper you hand to the dirty factory guy at the end of the line, where the bits hit the road | 01:58 |
kanzure | at the end of the line? | 01:58 |
fenn | somewhere in china | 01:58 |
fenn | in this factory there's dudes that run the machines and dudes that wrangle cad files | 01:58 |
kanzure | ok, I was interpreting lines to be machines | 01:59 |
fenn | the cad file is wrangled into a g-code format by cad programs that only recognize certain features of cad file formats, so it usually takes lots of massaging with said cad programs to make it work | 01:59 |
kanzure | so why would he be at the end of the machines if he's interpreting the drawings? etc. | 01:59 |
kanzure | hrm | 01:59 |
kanzure | how awkward .. no material specifity? | 02:00 |
fenn | "line" goes something like research -> engineering -> development -> procurement -> manufacturing | 02:00 |
kanzure | so people with calculators and fancy equations are doing everything, separately, somewhere, off in invisible margins? | 02:00 |
fenn | ok wait i'm not done yet | 02:00 |
fenn | while the software jockey is wrangling cad formats, the shop floor guy is looking around for a block of steel, because it says "steel 1040 AISI blah blah" | 02:01 |
fenn | on the paper drawing | 02:01 |
fenn | he theoretically doesnt care what the part is or does, as long as he makes the part to specification, meaning all the dimensions are to within spec of the drawing, and it looks "right" | 02:01 |
kanzure | is it usually just single word strings "materialname" ? and if so, why isn't that on these drawings | 02:02 |
kanzure | I only briefly glanced at the xml and other files here, but I'm pretty sure it's not here | 02:02 |
fenn | but the problem is many engineers are clueless, so the shop guy has to reverse engineer the part from the drawing and figure out what's wrong with the drawing (usually because the dickhead didnt know that .001mm tolerance is infeasible, even though it's the default tolerance of his cad program) | 02:02 |
fenn | it's not there.. that's not really a part diagram, i think its just a sort of general sketch of the whole product | 02:03 |
kanzure | this is Wrong, right? | 02:04 |
fenn | of course it's wrong! | 02:04 |
fenn | these days the engineers dont even know how to make a "normal" shop drawing, the kind that everyone can read | 02:05 |
kanzure | what's the difference? the correct answer would include two links :p | 02:05 |
fenn | they only teach GD&T which is this cryptic symbology based on extracting the most detailed description of what you want in any possible situation | 02:06 |
kanzure | I haven't heard of this yet | 02:06 |
fenn | dont worry about it | 02:06 |
fenn | unless you're going to go work in a factory | 02:06 |
kanzure | 'geometric drawing and tolerance' | 02:06 |
fenn | geometric dimensioning and tolerance | 02:06 |
kanzure | but I thought you said it's engineers that use gd&t? | 02:06 |
kanzure | and supposedly I'm taking introductory engineering courses (whether or not this is true, I'm not sure) | 02:06 |
fenn | they do, but nobody else knows how to read it | 02:06 |
kanzure | so why a factory? | 02:07 |
kanzure | anyway | 02:07 |
fenn | because some poor sap has to translate between gd&t and regular shop drawings | 02:07 |
fenn | unless for some reason your factory workers went to engineering school | 02:07 |
fenn | its sort of like, one guy speaks calculus and the other speaks matlab | 02:08 |
* kanzure wonders who uses oem machines :p http://oem.cadregister.com/ I kid | 02:08 | |
kanzure | hm, if it's wrong, then this data set is pretty much useless | 02:08 |
kanzure | is there a way to make it useful without those annotations? | 02:08 |
kanzure | oh, right, forking it | 02:09 |
fenn | what annotations? | 02:09 |
kanzure | what the material is | 02:09 |
kanzure | and other stuff | 02:09 |
kanzure | like equations governing its operation, or how to use it, or manuals, or technical documentation, all sorts of relevant stuff | 02:09 |
fenn | that's not a part drawing, there probably isnt enough information there to manufacture the parts in it | 02:09 |
fenn | (or use it) | 02:09 |
kanzure | parts drawing != the geometry of the part? | 02:10 |
fenn | STEP is supposed to handle all that stuff but it rarely does, because you need like 9 billion dollars of software to be able to read/write all that info | 02:10 |
kanzure | http://heybryan.org/~bbishop/docs/me302/ hurray for crappy file formats | 02:10 |
fenn | parts drawing is a drawing, geometry of the part is just geometry | 02:11 |
fenn | most cad formats include tolerances with the geometry, but not all | 02:11 |
kanzure | let me use more convenient terminology | 02:11 |
fenn | drawing must have tolerances for someone to make it | 02:11 |
kanzure | the 'geometry' seems to be 3D point/edge/line plotting | 02:11 |
kanzure | part drawing then is what? | 02:11 |
kanzure | what other information does it encode | 02:12 |
fenn | geometry is a solid, in this case using boundary representation, but can be implicit geometry like spheres or paraboloids or whatever | 02:12 |
* kanzure notices that the oem.cadregister.com site must be made by a German. <title> is "Testeite" | 02:12 | |
kanzure | oh | 02:12 |
kanzure | and then drawing includes tolerances in terms of measurement variation? "within 10% of desired measurements" | 02:13 |
kanzure | 10% :-p | 02:13 |
kanzure | http://www.thomasnet.com/catalognavigator.html?cov=NA&what=Heaters&heading=37420106&cid=10061056&pdtl=A&CNID=&cnurl=http%3A%2F%2Fstegousa.thomasnet.com%2FCategory%2Ffan-heaters | 02:13 |
kanzure | ^ there's some metadata there | 02:13 |
fenn | drawing also must have material, surface finish. might have "clues" like wtf the part is for, or what each feature does | 02:13 |
kanzure | heating element, overheat protection, housing, axial fan, air flow free blowing, function control light, connection, mounting, position, dimensions, weight, operating temperature, storage temperature, protection type, protection class | 02:14 |
fenn | if you're lucky it will inlude fixturing/machining setups | 02:14 |
kanzure | oh | 02:14 |
kanzure | :) | 02:14 |
kanzure | so it'd be somewhat slightly usable | 02:14 |
kanzure | in comparison to these silly cad files that thomasnet is trying to sell me | 02:14 |
kanzure | 'sell' | 02:14 |
fenn | is that metadata for the part TIN31447 or just the ontology you found it within? | 02:14 |
fenn | (btw slight philosophical musing: "is" a fan a part or an assembly?) | 02:16 |
fenn | well the chinese finally made a cnc hobby mill: http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=66052 | 02:22 |
kanzure | http://cooperationcommons.com | 02:28 |
kanzure | argh, damn musings | 02:28 |
* fenn yawns | 02:28 | |
fenn | why cant all these cooperation websites cooperate | 02:28 |
kanzure | is gnu really all just stallman and torvalds? | 02:29 |
kanzure | this can't be true .. | 02:30 |
kanzure | I mean, it's the only way that explains anything getting done | 02:30 |
fenn | no, stallman is mostly a political figure these days | 02:31 |
fenn | and torvalds doesn't really finish what he starts | 02:31 |
fenn | but in general, yes, it's mostly just one or two guys per niche making 90% of the cool stuff happen | 02:31 |
kanzure | I'm pretty sure stallman is responsible for the original toolchain of the 80s | 02:32 |
kanzure | and then cygnus which had business support | 02:32 |
fenn | yes, stallman practically wrote all the posix stuff himself | 02:32 |
kanzure | but why did it catch on? | 02:32 |
fenn | cat ls du etc | 02:32 |
kanzure | I remember it being because people hated installing software | 02:32 |
kanzure | when they got new machines. | 02:32 |
fenn | because people were familiar with UNIX but hated the unix vendors | 02:32 |
kanzure | haven't we had this conversation before | 02:33 |
fenn | stupid license agreements, bugs that should have been fixed ten times already | 02:33 |
fenn | cygnus didnt add anything to GNU, just their own windows port | 02:34 |
fenn | also there was a fear that unix would just vanish into the encroaching sea of windows computers | 02:35 |
fenn | you were asking before about how stallman went about announcing his crusade | 02:35 |
fenn | and dredged up some old usenet postings | 02:36 |
fenn | (rather uninspiring i might add) | 02:36 |
fenn | it's kind of amazing anyone cooperated with him - oh wait, they didn't. | 02:36 |
fenn | that whole cathedral/bazaar thing | 02:36 |
jk4930 | kanzure: why did it catch on? -> "worse is better". it pushed unix, it pushed FOSS later. ;) | 02:37 |
fenn | yep | 02:37 |
fenn | plan 9 is superior to unix | 02:38 |
fenn | linux is agonizingly slowly moving towards a plan-9-like system | 02:38 |
kanzure | I'm not sure when it comes to programming | 02:39 |
kanzure | I've seen some awesome programmers before that seem to be able to do the impossible, | 02:39 |
kanzure | for instance, it would seem like some of them could just go implement their own plan 9 file system for linux if they wanted to | 02:39 |
kanzure | but they don't. | 02:39 |
kanzure | not sure how easily it would happen with minor patches | 02:39 |
jk4930 | maybe they did. but no one used it. | 02:39 |
kanzure | are big revisions doable with minor patches from multiple programmers? | 02:39 |
fenn | yeah and dont forget all the software that has to be fixed to use the new system | 02:40 |
jk4930 | network effect + switching costs | 02:41 |
kanzure | fenn: metadata | 02:45 |
kanzure | not ontology | 02:45 |
kanzure | as for part v. assembly musings .. what would assembly imply? | 02:45 |
fenn | more than one part | 02:45 |
fenn | assembled | 02:45 |
kanzure | hrm | 02:46 |
fenn | generally the assembly has some separate metadata | 02:46 |
fenn | you can make a bridge out of i-beams and rivets | 02:46 |
kanzure | all things are made of other things, infinite divisables, 'atoms' not being 'atoms' .. | 02:46 |
kanzure | oh, | 02:46 |
kanzure | assemblies do have extra data available? | 02:46 |
fenn | "atom" is not a part unless you're a nutase | 02:46 |
fenn | (nutase = enzyme that degrades nuts) | 02:46 |
fenn | i dont know as much about cad metadata as i'd like (they call it PDM) | 02:47 |
fenn | you might want to interrogate ben in #cad | 02:47 |
fenn | or scott, or whatever his name is | 02:47 |
fenn | scanf | 02:48 |
kanzure | Okay. | 02:48 |
jk4930 | ok, my compile has finished, here it's 5 am, good night. cu | 02:50 |
kanzure | well this is odd | 02:50 |
kanzure | http://www.thomasnet.com/catalognavigator.html?cov=NA&what=Boilers%3A+High+Pressure&heading=6113609&cid=10043403&pdtl=A&CNID=&cnurl=http%3A%2F%2Fmiura.thomasnet.com%2FCategory%2Fsteam-boilers | 02:50 |
kanzure | this has notably more data | 02:50 |
kanzure | some PDFs, 'system diagram', wiring diagram, schematic, dimensions, .. | 02:50 |
kanzure | me dislikes these interfaces that have a variable "surprise!" amount of data shown on each page | 02:51 |
fenn | that's the nature of the XML beast | 02:51 |
kanzure | hm? xml would be useful here though | 02:51 |
kanzure | if you come across something that you know you can deal with it | 02:51 |
fenn | er.. you are wishing for a schema right? | 02:52 |
kanzure | I'm wishing for a plain xml file for each of these products | 02:52 |
fenn | but any XML document can also have stuff that's not in th schema | 02:52 |
kanzure | schema would be extra / bonus | 02:52 |
kanzure | this is true | 02:52 |
kanzure | but at least it could theoretically be somewhat repetitive | 02:52 |
kanzure | i.e., if I come across something new, | 02:52 |
kanzure | I just need to look at it once and hit 1k other instances | 02:52 |
fenn | not following | 02:53 |
kanzure | why does this one use an 'autocad active x plugin' while the other heater thingy was using something else | 02:53 |
kanzure | uhm, well, if you have your xml script to process the xml data set | 02:53 |
fenn | oh, bad web design i'd guess | 02:53 |
kanzure | and you come across some element that makes you crash | 02:53 |
kanzure | then you look at it visually and make some updates and you're good for another few thousand xml files maybe | 02:53 |
kanzure | if they have random variables for each item, then it's totally useless | 02:53 |
fenn | one firm hired later on decided to just make their own system instead of extending what was there (?) | 02:53 |
kanzure | *for each xml file | 02:53 |
kanzure | right | 02:54 |
kanzure | that's what I'm guessing is going on | 02:54 |
kanzure | or they don't know what they are doing | 02:54 |
fenn | its not like anyone actually uses that info | 02:54 |
kanzure | heh | 02:55 |
fenn | you dont spend $500k and not call up the manufacturer for the datasheet/manual | 02:55 |
kanzure | let's say that you want to try to actually figure out which product to use, one out of a thousand alternatives, this needs to be somewhat ontologically meaningful .. not a phone number to go call and yell at somebody for the right information | 02:56 |
kanzure | for a thousand times | 02:56 |
kanzure | ugh | 02:56 |
* fenn wonders how much a 4500hp boiler costs | 02:56 | |
fenn | kanzure: this is why there are procurement departments | 02:57 |
* kanzure doesn't know what this is: http://cadupdate.thomasnet.com/mythomas/signup.html?goto=http://updates.cadregister.com/updates.asp | 02:57 | |
fenn | its someone's job to call up all these suppliers and yell at them | 02:57 |
fenn | i guess partspec/cadblock is some kind of integrated cad software database of stuff that's in thomasnet | 02:58 |
fenn | hm. a 'block' is like a class (like programming) in autocad | 02:59 |
fenn | except you cant do inheritance or anything cool of course | 03:00 |
fenn | Design professionals use the CADBlocks CD-ROM to insert pre-drawn architectural and building products into their CAD plans | 03:01 |
kanzure | hm | 03:01 |
fenn | Engineering professionals use the PartSpec CD-ROM to insert 2D and 3D drawings of parametrically designed mechanical parts into their CAD plans. | 03:01 |
fenn | etc | 03:01 |
kanzure | http://updates.cadregister.com/downloads/partspec.exe | 03:01 |
fenn | so not downloading that | 03:02 |
kanzure | 69 MB exe file | 03:03 |
kanzure | yeah .. | 03:03 |
kanzure | http://oem.cadregister.com/asp/PPOW_Entry.asp?language=GB&product=PS&referrer=V3Redirect&ori=/CADREgister.ppow& | 03:07 |
kanzure | I am confused | 03:07 |
kanzure | where did this come from? | 03:07 |
fenn | it's.. a list of manufacturers? | 03:08 |
kanzure | well, they said "please wait x seconds to be redirected to our new system" | 03:08 |
kanzure | but it looks like one half of their own website doesn't even know about this 'new system' | 03:09 |
fenn | so? | 03:09 |
kanzure | dunno, guess I'm still in hunt-for-technical-incompetence mode | 03:09 |
fenn | its pretty cool to see instant cad drawings of products i've looked at on the manufacturer's website | 03:10 |
fenn | sort of amazing that people put up with no antialiasing | 03:11 |
fenn | its 2008 ffs | 03:11 |
kanzure | terrible images. | 03:11 |
kanzure | 'Part drawings can be inserted with one click into: AutoCAD, Mechanical Desktop, Inventor, Solid Works, SolidEdge, ProEngineer, CATIA, Caddy++, HiCAD, TurboCAD, MegaCAD, TopSolid, SmartSketch, Technobox and IntelliCAD, for example. To setup your CAD system inside PartSpec go to: Options --> Configuration --> Defaults and select your CAD system. | 03:30 |
kanzure | PartSpec allows the user to access literally millions of components, to create these parts parametrically and to transfer them to the appropriate CAD systems. ' | 03:30 |
fenn | take "parametrically" with a grain of salt | 03:31 |
kanzure | lies? | 03:32 |
kanzure | but it says 'parts drawings' which is what you suggested might be slightly better than just plain 3d coordinate geom files | 03:33 |
kanzure | ' III. Known restrictions ----------------------- a) A correct presentation of shaded 3D-models is possible only with a | 03:37 |
kanzure | color palette of at least 256 colors.' | 03:37 |
fenn | part drawing != shop drawing | 03:50 |
fenn | though one might hope so | 03:50 |
fenn | anyway 'part' doesnt necessarily mean one solid component, it could be an assembly (like a fan for example) | 03:51 |
fenn | sorry if i'm confusing things, but it's not my fault | 03:52 |
fenn | one thing you can count on is that a "shop drawing" will be one solid manufacturable component (unless it's an assembly diagram) | 03:53 |
fenn | some examples | 03:58 |
fenn | here's what i'm calling a "shop drawing": http://www.laboratoryformicroenterprise.org/lme/Drawings/2.1.2-1.gif | 03:58 |
fenn | here's an assembly diagram: http://www.laboratoryformicroenterprise.org/lme/Drawings/2.1-1.gif | 03:59 |
fenn | and here's an assembly diagram where each "part" is another assembly: http://www.laboratoryformicroenterprise.org/lme/Drawings/2-1.gif | 03:59 |
procto | I have just become a 23andme customer | 05:00 |
procto | they dropped their prices from $1000 to $400 | 05:00 |
kanzure | procto: decodeme gives you a CVS file w/ SNPs listed | 05:07 |
kanzure | don't know if you get your data from 23andme | 05:07 |
kanzure | Hey spookact. | 05:07 |
spookact | ello kanzure | 05:08 |
procto | too late now. my personal genomics price ceilirg for thi- year was $500 | 05:08 |
kanzure | spookact: What brings you? :) | 05:08 |
procto | I saw tim o'reilly la-t night and he actually broke the price drop for me | 05:08 |
spookact | kanzure: noticed this channel mentioned in #bioinformatics, and transhumanism just happens to be my specific area of interest :) | 05:10 |
kanzure | Aha. | 05:10 |
kanzure | :-) | 05:10 |
kanzure | Yes, you're in the right place then. | 05:10 |
kanzure | Today we've been complaining about thomasnet.com | 05:10 |
kanzure | or, I have at least | 05:10 |
kanzure | spookact: So, what brings you to the concepts of transhumanism ? | 05:12 |
spookact | kanzure: I'm originally from a programming background, long story short gradually shifted toward neuroscience and now I'm back in school | 05:15 |
spookact | read a few too many books like Kurzweil's and now I'm all ready to evolve :) | 05:16 |
kanzure | programming background, excellent | 05:17 |
kanzure | I find that I don't care too much for Kurzweil | 05:18 |
kanzure | http://heybryan.org/fernhout/ for the f/oss programmer's perspective on kurzweil ;-) | 05:18 |
kanzure | sort of. | 05:18 |
spookact | haha I'll give this a read... I wasn't impressed by his "science" at all... it was more of a cheerleading effect | 05:19 |
kanzure | he's an "inventor" | 05:20 |
kanzure | Hey gene. | 06:53 |
gene | hello kanzure | 06:54 |
gene | apparently plastic seperation is a bit harder than it looks | 06:55 |
gene | which might make it harder to make a sea bound replicator | 06:57 |
gene | there's a bunch of plastic out there just floating around in a current | 06:59 |
gene | it's the size of Texas | 07:00 |
kanzure | coordinates? | 07:01 |
gene | it moves around I think | 07:01 |
gene | http://www.bestlifeonline.com/cms/publish/travel-leisure/Our_oceans_are_turning_into_plastic_are_we.shtml | 07:02 |
gene | the north pacific gyre | 07:02 |
gene | it's just a bunch of plastic garbage floating around in the water | 07:04 |
gene | plastic garbage just keeps on accumulating there | 07:04 |
gene | unfortunately, not all of it is thermoplastics | 07:05 |
gene | some of it is fishing line | 07:07 |
gene | from my experience, it doesn't melt | 07:09 |
kanzure | fenn: words suck, | 07:43 |
kanzure | isn't the solution typically to therefore acronymitize everything? | 07:43 |
kanzure | an ACP, acronym creation process | 07:44 |
* kanzure has been reading too many papers | 07:44 | |
kanzure | "KARMA represents ourfirst steps in designing a testbed for the knowledge-based generation of maintenance and repair instructions using a head-mounted, see-through display." | 07:56 |
kanzure | http://heybryan.org/~bbishop/docs/Knowledge-based_augmented_reality.pdf | 07:57 |
kanzure | http://heybyan.org/~bbishop/docs/Understanding_and_automating_algorithm_design_-_Elaine_Kanf.pdf | 08:00 |
kanzure | Ooh. | 08:30 |
kanzure | http://heybryan.org/~bbishop/docs/manufacturing/An%20intelligent%20agent%20for%20conceptual%20design%20-%20informed%20search%20using%20a%20mapping%20of%20abstract%20qualities%20to%20physical%20form%20-%20Koile%20-%20good.pdf | 08:30 |
kanzure | A way to deal with crappy user input. Consider it as having to be "repaired" to the baseline level, letting the user help to conduct such 'repairs'. Progression from ambiguity/abstraction -> parts in the db. | 08:30 |
kanzure | 'Abstract. In early stages of design, the language used is often very abstract. In architectural design, for example, architects and their clients use experiential terms such as "private" or "open" to describe spaces. The Architect's Collaborator (TAC) is a prototype design assistant that supports iterative design refinement using abstract, experiential terms. TAC explores the space of possible | 08:31 |
kanzure | designs in search of solutions satisfying specified abstract goals by employing a strategy we call dependency-directed redesign: It evaluates a design with respect to a set of goals, uses an explanation of the evaluation to guide proposal and refinement of design repair suggestions, then carries out the repair suggestions to create new designs.' | 08:31 |
kanzure | 'dependency directed redesign' pg 4 in particular | 08:32 |
kanzure | ' DESIGN GRAMMARS FOR CONCEPTUAL DESIGNS OF SPACE STATIONS' http://www.isd.uni-stuttgart.de/forschung/publikationen/publikationen.html | 08:47 |
kanzure | bah, stupid paywalls | 08:48 |
nsh | kanzure++ | 14:01 |
nsh | lol | 14:06 |
nsh | i searched "paywalls" and found a physicsforums thread you'd posted in | 14:06 |
-!- fenn_ is now known as fenn | 15:01 | |
kanzure | mom says http://rockler.com/ for swivels | 18:17 |
kanzure | nsh: I was being lazy. I don't know why, but I figured I'd rather crosspost to 20 different forums asking the same damn question, rather than phone 50 different university libraries asking them for a list of the journals and databases they subscribe to. | 18:30 |
nsh | quite | 18:32 |
nsh | mmm | 18:32 |
* nsh is trying to think of a way we could share journal access credentials | 18:33 | |
kanzure | I still need to torrent some of the paper archives I've been downloading. | 18:33 |
nsh | without giving each other our passwords | 18:33 |
kanzure | You can set up a proxy on your own machine. | 18:33 |
nsh | i suppose i could set up a proxy | 18:33 |
nsh | yeah | 18:33 |
kanzure | nsh: Are you aware of the ezproxy password trading circles? | 18:33 |
nsh | vaguely | 18:33 |
nsh | or i remember something similar | 18:33 |
nsh | please go on | 18:33 |
nsh | o/ biopunk | 18:34 |
kanzure | Botnets sniff out passwords of unsuspecting users, then the Chinese/Russians farm those out for a small fee to Chinese students that really, really need it, then they post those on forums to share with their brothren. | 18:34 |
kanzure | It's how I was reading papers before I realized I could cough up $40 for the local community college to give me access (by taking a class). | 18:34 |
nsh | interesting | 18:35 |
nsh | did you purchase or intercept? | 18:35 |
kanzure | Pardon? | 18:35 |
nsh | oh, nevermind, you probably found on one of the forums | 18:35 |
kanzure | yes | 18:35 |
kanzure | nsh: I think it would be wiser to write more spiders and archive papers for future reference. | 18:36 |
kanzure | I know that I might not be in a university forever | 18:36 |
nsh | aye | 18:36 |
* nsh is working at the lab until the end of the year | 18:37 | |
nsh | then i'll lose access | 18:37 |
kanzure | Hrm, it seems I have the full indices to sciencedirect. I just need to write some HTML::TreeBuilder code to save each paper to the write file name and with the right extra metadata (I will not repeat the same mistakes I made with nature ...). | 18:38 |
kanzure | Since it's 9 million papers I figure that I don't actually have enough hdd space for it. I was going to look for some structured data describing "impact factor" of journals by name, but .. | 18:39 |
nsh | mmm | 18:39 |
kanzure | well, there is a way to get that information - each journal on sciencedirect has a link to elsevier's other server, through some cryptic link of course, where they have this brief text snippit saying "impact factor: 3.2" which is, of course, yet another silly regex | 18:39 |
* nsh nods | 18:40 | |
nsh | btw | 18:43 |
nsh | how are classes going? | 18:43 |
fenn | is the amount of data available in journal articles keeping pace with the cost of storage technology or will we eventually be able to swap "all" of the articles out there? | 18:44 |
nsh | storage will win, i'd wager | 18:44 |
nsh | though there is exponential increase in the number of papers in some fields (biosciences), it's far more strongly bounded growth than storage capacity | 18:45 |
nsh | well, there is exponential increase in scientific data, at least | 18:45 |
nsh | submitted papers is probably not growing at quite that | 18:46 |
fenn | with bio stuff yes, but either it's a trade secret or patented or publically available | 18:46 |
fenn | so there's no real advantage to charging for access to bio information (yet) | 18:46 |
* nsh nods | 18:47 | |
nsh | looking for some stats | 18:47 |
fenn | "99% of statistics never lie" | 18:48 |
nsh | (finding it hard to think of sensible search terms for this kind of query) | 18:48 |
kanzure | nsh: Classes are going well. Chem prof is trying to teach his class how to solve Schrodinger's wave equation for polyelectronic atoms and the poor kids haven't even gone through diff-eq. "Building Brains" is ok, but I failed when I couldn't recite ASCII 'A' off the top of my head. | 18:49 |
fenn | open access bioinformatics | 18:49 |
nsh | heh: http://www.ncbi.nlm.nih.gov/pubmed/15657104 | 18:49 |
kanzure | Does anyone remember the Word DOC to HTML converter that I was using in 2006? | 18:49 |
nsh | as in, the ascii char number for 'A'? | 18:49 |
kanzure | nsh: Re: that link. Marcotte Lab was doing just the same for GRNs. They're doing supermassive correlation studies. This is the lab with the full yeast knockout library in their attic. | 18:50 |
kanzure | nsh: yes | 18:50 |
kanzure | I said 39, and it was 48 | 18:50 |
nsh | meh | 18:50 |
nsh | what would it serve anyone to brainclutter with that? | 18:50 |
fenn | word doc to html? | 18:50 |
nsh | the point of the human computer symbiosis is that we can easily check simple things like that | 18:51 |
fenn | like `strings *.doc` >> foo.html | 18:51 |
kanzure | fenn: Somebody shoved a laptop under my fingers during the 'building brains' class and so I typed up the entire lecture in real time, but then it was in DOCX format and I cried. | 18:51 |
nsh | and so dedicate more brainspace to abstract, non-verbal conceptualisation | 18:51 |
fenn | why didnt you just save as some other format | 18:51 |
nsh | kanzure | 18:51 |
kanzure | nsh: No, I think that after all these years that's something I should be able to recall | 18:51 |
kanzure | so much debugging etc | 18:51 |
* nsh smiles | 18:51 | |
kanzure | fenn: because I must not have been thinking. | 18:51 |
nsh | but yeah, kanzure, i'd suggest audiorecording every lecture, then sharing the transcription load between members of the class | 18:52 |
kanzure | wvware | 18:52 |
kanzure | wtf, I would have never remembered that | 18:52 |
nsh | only one person needs to transcribe something for everyone to have a copy | 18:52 |
kanzure | nsh: my fingers are quick enough for it to not matter much | 18:53 |
kanzure | but yes, I agree | 18:53 |
* nsh nods | 18:53 | |
nsh | ideally, you'd share the work out, and have a multi-layered collaborative note-system | 18:53 |
nsh | so, audio/visual recordinging, transcript, wikilinked/referenced version, etc. | 18:53 |
fenn | ascii A is 65? | 18:54 |
nsh | then share comments and resolve queries on the material | 18:54 |
kanzure | fenn: crap. | 18:54 |
nsh | ultimately, this would feedback into a free (CC/whatever) textbook being generated for the next year's class | 18:54 |
fenn | why would you need to know that anyway? | 18:54 |
kanzure | right right | 18:54 |
kanzure | uhm. | 18:54 |
nsh | the incumbant year could accept donations from the succeeding year's students for efforts and costs incurred | 18:55 |
kanzure | How many times have you wondered wtf your program was doing, so printed out a char, and then had to go look up on the ascii tables to check? Those clicks can be reduced. | 18:55 |
fenn | more important is to know how to find the value if you need it | 18:55 |
nsh | actually | 18:55 |
nsh | it'd be nice to go meta on this | 18:55 |
kanzure | of couse, it's quicker to just do a quick script to check for you | 18:55 |
kanzure | nsh: the notetaking? | 18:55 |
fenn | printf(%X", char c) | 18:56 |
kanzure | rawr, yes fenn | 18:56 |
nsh | no, the information access question. | 18:56 |
fenn | or %d if you prefer | 18:56 |
kanzure | nsh: to what extent? | 18:56 |
nsh | have a 'watcher' program, that keeps tabs on how often you are required to look up a certain thing | 18:56 |
nsh | when it reaches a predetermined threshold, have it suggest memorisation | 18:56 |
nsh | or locally mirror for a resource that you query often | 18:56 |
nsh | etc. | 18:56 |
kanzure | well, I was intending to start mirroring every page I ever view since I have the capacity for it | 18:57 |
nsh | so information that's needed often is more 'cached' than more rarely accessed info | 18:57 |
kanzure | and then "sift" and "filter" stuff to the top or bottom of the stacks | 18:57 |
nsh | right | 18:57 |
nsh | keep meta-data on correlations between use of various parts | 18:57 |
nsh | so automate semantic linking | 18:57 |
fenn | this is starting to sound like vannevar bush | 18:57 |
kanzure | my point is that we've already been over this | 18:58 |
kanzure | I'll get to it "eventually" | 18:58 |
nsh | fenn, the guy who came up with the microfilm internet idea? | 18:58 |
fenn | right | 18:58 |
kanzure | also, Phreedom, when he comes back, is doing some semantic desktop stuff that he's proud of | 18:58 |
nsh | Memex, right | 18:58 |
* nsh nods | 18:58 | |
fenn | the idea was that you could store different paths through the web | 18:58 |
kanzure | huh? | 18:58 |
nsh | right, it makes perfect sense | 18:58 |
kanzure | I don't remember this | 18:58 |
kanzure | I thought I knew Memex though | 18:58 |
nsh | every trail is data that should feed back into the system | 18:58 |
nsh | the internet needs to be activated | 18:58 |
kanzure | right, like that old 2006 comment on my front page on the iste | 18:59 |
kanzure | *site | 18:59 |
nsh | interaction should enrich it | 18:59 |
nsh | kanzure, can you paste it? | 18:59 |
kanzure | suggesting http://trexy.com/ and http://prefound.com/ | 18:59 |
nsh | mmm | 18:59 |
kanzure | recording search/click paths through the web | 18:59 |
nsh | right | 18:59 |
fenn | these sorts of things can be easily spambotted or hacked with social engineering | 18:59 |
kanzure | semantics fails | 18:59 |
kanzure | that's why we need 'realistics' people! | 19:00 |
nsh | well, i think the wikipedia effect might mitigate that enough | 19:00 |
* kanzure is ashamed of himself. | 19:00 | |
fenn | you need trust networks, and additionally you need smart difficult-to-hack people to trust | 19:00 |
nsh | (i.e. enough 'good' people to negate the spam) | 19:00 |
nsh | trust networks and metrics are a quagmire that no-one has been able to escape yet | 19:00 |
nsh | i've yet to see anything workable | 19:00 |
* nsh not touching with a long stick until some light is seen | 19:01 | |
fenn | no, not to negate the spam, just to provide enough good content that there's something worth paying attention to | 19:01 |
nsh | right | 19:01 |
nsh | in a sense, it's an inertia problem | 19:01 |
fenn | spam is automatically removed by the trust network's filtering | 19:01 |
nsh | have to get enough good people rolling the ball before such obstacals are encountered | 19:01 |
fenn | and to keep it rolling once the unwashed masses jump on the bandwagon | 19:01 |
nsh | perhaps gmail's long-term beta + invitation strategy would work well | 19:01 |
nsh | aye... | 19:02 |
nsh | that's a whole nother kettle of fish | 19:02 |
nsh | perhaps tiering | 19:02 |
fenn | gmail invitation seemed more like reverse psychology to me than anything | 19:02 |
nsh | have an "inner party", "outer party" type thing | 19:02 |
fenn | 'you cant have it' so it must be good | 19:02 |
nsh | fenn, it certainly had that effect too | 19:02 |
nsh | to be fair though, at the time, it was objectively a great service | 19:02 |
nsh | o/ gene | 19:03 |
nsh | bye then | 19:03 |
nsh | so, when i was thinking about this trust stuff a few years back, i decided one good approach would be something you might call 'successive cementing', each person that follows a user's trail of contributions without reverting or forking them adds confidence to them | 19:04 |
fenn | An associative trail as conceived by Bush would be a way to create a new linear sequence of microfilm frames across any arbitrary sequence of microfilm frames by creating a chained sequence of links in the way just described, along with personal comments and side trails. | 19:04 |
* kanzure doesn't understand tust | 19:04 | |
kanzure | *trust | 19:04 |
kanzure | fenn: just a scaffold/annotation/tracking dealy | 19:05 |
kanzure | Xanaduguys wanted it to be implemented into the backbone of the web | 19:05 |
kanzure | web-server-side | 19:05 |
fenn | yeah, it would be very easy to implement | 19:05 |
fenn | right now i mean | 19:05 |
kanzure | don't think so, they would have done it by now | 19:05 |
fenn | you have referrer urls, but nobody shares them, why? | 19:06 |
kanzure | all of the guys like us wanting it, I mean | 19:06 |
kanzure | is that all it would take? | 19:06 |
fenn | for a fragmented subset of trails, yes mostly i think | 19:06 |
fenn | of course there's spamming/spoofing of referrer url's, but that's a different problem | 19:07 |
kanzure | OpenOffice can do DFX / "AutoCAD interchange format" apparently | 19:07 |
kanzure | Can't quite find docx though | 19:08 |
fenn | i think this is what google analytics does sorta | 19:08 |
fenn | DXF compatibility is a .. um, chinese horse (cant think of the metaphor) | 19:09 |
fenn | DXF isnt a standard, it's just "whatever loads into autocad version 1.2.3.4 when the author wrote it) | 19:10 |
kanzure | http://heybryan.org/school/buildingbrains/2008-09-12.html | 19:11 |
kanzure | finally | 19:11 |
kanzure | that was a pain :) | 19:12 |
kanzure | also, the class is kind of a pain since it's going over basics at the moment | 19:12 |
fenn | is there any reason to believe it will ever progress beyond basics? | 19:13 |
kanzure | the drop out date has passed, so yes | 19:15 |
fenn | hey that's halfway decent translation to html | 19:15 |
* kanzure just opened up a paper from the "Knowledge Based Engineering Laboratory" .. what else would it be ? | 19:17 | |
kanzure | fantasy engineering? | 19:18 |
fenn | i think they mean explicit knowledge rather than implicit | 19:18 |
fenn | i.e. stuff you can put in a computer and frottage in various ways | 19:19 |
kanzure | this paper, "Function-to-form mapping - model, representation and applications in design synthesis" and http://heybryan.org/~bbishop/docs/manufacturing/An%20intelligent%20agent%20for%20conceptual%20design%20-%20informed%20search%20using%20a%20mapping%20of%20abstract%20qualities%20to%20physical%20form%20-%20Koile%20-%20good.pdf | 19:19 |
kanzure | both take the approach of calling the user's idea _wrong_ | 19:20 |
kanzure | that's why they're using a computer for design anyway, right? | 19:20 |
kanzure | if they knew the design they would just write it out completely | 19:20 |
kanzure | but since they don't, the user is wrong and the computer is there to repair the user's misconceptions | 19:20 |
kanzure | (with their help, of course) | 19:20 |
fenn | i think its just the user doesnt know what to do next | 19:21 |
kanzure | like the times I become completely confused and wonder what it was that I was supposed to be doing during the time that I am being confused | 19:22 |
kanzure | here's that other paper: http://heybryan.org/~bbishop/docs/manufacturing/Function-to-form%20mapping%20-%20model,%20representation%20and%20applications%20in%20design%20synthesis.pdf | 19:22 |
kanzure | 'What to do next: Using problem status to determine the course of action - Get this article | 19:24 |
kanzure | DG Ullman, D Herling, BD Ambrosio - Research in Engineering Design, 1997 - Springer' | 19:24 |
kanzure | HEH | 19:24 |
fenn | stuff like this is so much easier to understand by example (rather than technical paper) | 19:25 |
kanzure | which 'this' are we talking about? "what to do next" or the "fixing how the user is wrong" paper? | 19:25 |
fenn | um, "model, representation, and applications in design synthesis" | 19:26 |
kanzure | Hey spookact. | 19:27 |
fenn | crap like this makes my eyes glaze over: " Let us assume that an artifact, Artjk has been found in the | 19:27 |
fenn | design stage j with some elements of Artjk.Inp are in PSj.Inp | 19:27 |
spookact | hey kanzure | 19:27 |
fenn | <- REALISTIC | 19:28 |
kanzure | fenn: the professor was doing that the other day on the whiteboard in his office .. I had to chuckle since I'm coming from a perspective where I'm thinking only in terms of implementation, and there he was going off with all of this "fancy" abstract algebra :p | 19:29 |
kanzure | so much more simpler to just say "here's the code, fool" | 19:29 |
fenn | no shit | 19:30 |
fenn | at least i can play with the code to figure out how it works | 19:30 |
fenn | equations are useless | 19:30 |
kanzure | fenn: so, "what to do next" is a big question methinks | 19:31 |
fenn | somehow i think it's related to keeping their job, for if anyone saw how simple their ideas were they wouldnt seem worth paying attention to | 19:31 |
kanzure | I think the solutions are something like (1) try to retrace your steps back to the todo list you had in working mem, or (2) go sleep (which doesn't help as much). | 19:32 |
kanzure | right | 19:32 |
fenn | what to do next assumes you had a goal in the first place, which might not have been the case | 19:32 |
kanzure | but weren't you designing something? | 19:32 |
fenn | if i'm an architect fooling around in my studio (computer) i'm not necessarily working on a job for someone | 19:33 |
fenn | many times artists will come up with something on their own, and then later sell it as part of a commissioned project | 19:33 |
fenn | s/artists/engineers/ | 19:34 |
kanzure | but let's say you're big business boss sitting in front of CAD (our type of cad, real design stuffs) and you want to make the most ultimate hamburger ever, you want x y z features, blah blah blah, now what. | 19:34 |
kanzure | the concept of "repairing" their specs into usable information is kinda nice but you claim it's not enough | 19:34 |
kanzure | or that it's the wrong approach | 19:34 |
fenn | are you talking about the difficulty of making an algorithm to satisfy given constraints? | 19:35 |
fenn | or just the human-prodding aspect of working on a project | 19:35 |
kanzure | sort of .. except the constraints need to be in terms of the system in the first place | 19:35 |
kanzure | and if they already knew that, then they wouldn't have troubles beginning to talk about it in terms that the system can work with | 19:35 |
kanzure | erm, knew that -> knew how to write it in that way | 19:36 |
fenn | ah, well, that's an interface problem | 19:36 |
fenn | all the usual power vs simplicity arguments apply | 19:36 |
fenn | (not that i necessarily agree with them) | 19:36 |
kanzure | guess I should make a trip over to c2 | 19:36 |
fenn | what's c2? | 19:37 |
kanzure | c2.com, the oldest wiki ever? | 19:37 |
fenn | ah | 19:37 |
kanzure | http://c2.com/xp/BigDesignUpFront.html | 19:37 |
fenn | most of my c2 trails are related to philosophy and cat herding, strangely enough | 19:37 |
fenn | at least not explicitly programming stuffs | 19:37 |
kanzure | cat herding literally? | 19:38 |
kanzure | herding developers or something? | 19:38 |
fenn | crap like http://c2.com/cgi/wiki/quickDiff?CarFree | 19:39 |
kanzure | fenn: did you look at the idea of "repairing the user's specifications" ? if that's an interface problem, I don't see how it maps on to the typical interface arguments/debates like wysiwyg, do what I mean not what I say, etc. | 19:39 |
fenn | oddly enough that page references BigDesignUpFront in multiple places | 19:39 |
fenn | no, i didnt find "repairing the users specifications" in either paper | 19:40 |
fenn | wolfram seems to think that iterative design strategies suck | 19:41 |
fenn | design / evolution | 19:42 |
fenn | basically he showed how in a vast majority of cases, iterative algorithms will never be able to satisfy even a simple set of constraints (user specified goals) | 19:43 |
nsh | hmmm | 19:44 |
kanzure | fenn: from that page, wtf: "Bicycles have been used on the battlefield to great success, though I forget by whom. Basically, cars don't do very well in forested terrain, and men on foot as always move relatively slowly. " :) | 19:44 |
kanzure | fenn: methinks wolfram is right, iterative design does tend to suck | 19:44 |
kanzure | interesting | 19:45 |
fenn | it's true, bicycles were heavily used in WWII by paratroopers | 19:45 |
fenn | i saw some us army stuff about repurposing consumer mountain bikes for paratroopers | 19:45 |
kanzure | get those troopers a lightweight folding bike and they'll be happy .. always carry your mode of escape (or attack) | 19:46 |
kanzure | god I feel like a little kid racing down the street again | 19:46 |
kanzure | "charge!" | 19:46 |
fenn | wolfram on constraints: http://www.wolframscience.com/nksonline/page-342 | 19:49 |
fenn | he doesn't really provide any other way to satisfy constraints though, so i can only guess that he prefers rule-based built from the ground up systems (which meshes nicely with c2) | 19:49 |
fenn | ok enough pdf's for me for now | 19:51 |
* nsh thinks he ought watch a film | 19:51 | |
kanzure | "So why does the procedure not work better? The problem turns out to be a rather general one. And as a simple example, consider a line of black and white squares, together with the constraint that each square should have the same color as its right-hand neighbor. This constraint will be satisfied only if every square has the same color--either black or white. But to what extent will an iterative procedure succeed in finding this solution?" | 19:53 |
kanzure | http://www.wolframscience.com/nksonline/page-345-text | 19:53 |
kanzure | "How can one avoid this? One general strategy is to add randomness, so that in essence one continually shakes the system to prevent it from getting stuck. But the details of how one does this tend to have a great effect on the results one gets." | 19:54 |
kanzure | evolution gets in a rut sometimes | 19:54 |
kanzure | that's where mutation comes in | 19:54 |
kanzure | in case recombination incests itself or something | 19:55 |
kanzure | where did I put my nks pdf | 19:58 |
fenn | http://fennetic.net/pub/ebooks/bookwarez/Science popularization books/Stephen Wolfram - A New Kind Of Science.pdf | 20:00 |
fenn | 82MB tho | 20:00 |
fenn | just images of the text from website | 20:00 |
fenn | will the miracles never cease: http://www.teafromtaiwan.com/index.php?main_page=product_info&cPath=1_6&products_id=10&zenid=036286f9c945d77a0de19966fb4bc398 | 21:44 |
* nsh had a good NKS pdf at one point | 21:52 | |
nsh | could only ever find a half-buggered copy of GEB though | 21:53 |
* kanzure found a physical GEB for $5 | 22:20 | |
kanzure | fenn: GABA tea? :-) | 22:20 |
kanzure | Can't you just go buy MSG in a store and be happy? | 22:20 |
kanzure | Not quite the same thing, nevermind. | 22:21 |
fenn | glutamic acid = MSG + vinegar or whatever, right? | 22:21 |
fenn | kanzure is "time line for distributed manufacturing" from some list or just bouncing around? | 22:23 |
nsh | don't eat MSG | 22:24 |
nsh | (if you can avoid it) | 22:24 |
fenn | why? | 22:24 |
nsh | because it's [excito]toxic | 22:25 |
fenn | please explain how it's worse than, say, soy sauce | 22:25 |
nsh | because the active chemical in soy sauce does not cause wreak havoc with your neuroreceptors? | 22:27 |
fenn | but it's the same chemical | 22:27 |
nsh | depends on the sauce i guess | 22:29 |
nsh | apparantly the levels are comparable | 22:29 |
fenn | i've never seen a good argument why MSG is bad, it might just be my poor researching skills, but i'm skeptical of most crazed foodie claims | 22:29 |
* nsh would avoid using too much soy sauce also | 22:29 | |
fenn | i think i remember reading something about contamination at an msg factory in the 70's, but i might be making tha tup | 22:29 |
nsh | http://en.wikipedia.org/wiki/Monosodium_glutamate#Health_controversy | 22:30 |
nsh | and http://en.wikipedia.org/wiki/Glutamic_acid_(flavor)#Health_concerns | 22:30 |
fenn | but really it looks like a general irrational fear of "chemicals" in food, and having no fucking clue what any chemical actually is | 22:30 |
* nsh smiles | 22:30 | |
nsh | MSG is implicated in a lot of disorders | 22:31 |
nsh | http://www.jpands.org/hacienda/article27.html | 22:31 |
nsh | and it's the ubiquity as well as the toxicity that has to be taken into consideration | 22:31 |
nsh | effects of neurodegeneration are cumulative | 22:32 |
nsh | and highly compensated | 22:32 |
nsh | you can be way past fucked-up before it's apparant | 22:32 |
fenn | lyme borreliosis? how is that caused by MSG! | 22:33 |
* nsh has no idea | 22:33 | |
* nsh suspects someone just searched for papers which mentioned MSG and lyme borreliosis came up in the results | 22:34 | |
spookact | ticks love flavor | 22:34 |
spookact | MSG makes your blood yummy :) | 22:35 |
fenn | "an unknown percentage of the population may react to monosodium glutamate and develop a monosodium glutamate symptom complex when consuming more than 3 grams of monosodium glutamate alone" who in their right mind is going to eat 3 grams of MSG | 22:39 |
fenn | that would be like snorting caffeine or chugging vinegar | 22:39 |
nsh | heh | 22:40 |
fenn | i'm all for accuracy in food labeling, but after reading all that it still just sounds like mass hysteria due to ignorance | 22:43 |
nsh | romans had their lead pipes.. time will tell :-) | 22:43 |
* nsh tries to err on the side of not getting brainspazed | 22:44 | |
fenn | but there's so many real poisons out there that nobody seems to give a shit about | 22:44 |
fenn | what about the stuff they dump on crops from airplanes by the ton | 22:45 |
nsh | :-/ | 22:45 |
nsh | consumers have very little walletvote power on shit like that though | 22:46 |
nsh | which is probably why it gets less talktime | 22:46 |
kanzure | fenn: just bouncing around | 22:54 |
kanzure | nsh: I've had, bad experiences with MSG. | 22:54 |
fenn | what specifically? and what made you think it was MSG? | 22:55 |
kanzure | I remember when I was about seven or eight having some intense experiences in a Chinese restuarant that is still around | 22:55 |
kanzure | Trial and error to tell that it was the MSG on the lemon chicken dish. | 22:55 |
kanzure | We went out to eat often, so it didnt' take long for the parents to catch on. | 22:56 |
kanzure | *didn't | 22:56 |
nsh | what happened during these experiences? | 22:57 |
kanzure | Symptoms included some sort of intense headache or migrane, closing-in-of-vision kind of like when you're fading off to sleep and you jolt awake at the last moment, and general loss of activity 10 to 15 minutes after consumption begins. This probably means it wasn't the MSG since I don't think it gets into the blood that quickly. | 22:58 |
kanzure | I've actually been meaning to try some of it again to see how repeatable it is. | 22:59 |
fenn | sounds like a fun experiment :\ | 22:59 |
fenn | "lets see if this gives me a seizure.. no? next!" | 22:59 |
kanzure | fenn: Mostly because of my fascination with GABA, glutamate receptors, and so on re: attention links. | 22:59 |
kanzure | 'Airway Effects of Monosodium Glutamate in Subjects with Chronic Stable Asthma' <- Hm, this would have been me at the time. | 23:00 |
kanzure | but I don't recall airway effects | 23:00 |
kanzure | 'This article reviews the literature from the past 40 years of research related to monosodium glutamate (MSG) and its ability to trigger a migraine headache, induce an asthma exacerbation, or evoke a constellation of symptoms described as the "Chinese restaurant syndrome."' | 23:01 |
fenn | my working hypothesis is that there is some contaminant in synthetic MSG that some people have an immune sensitivity to | 23:01 |
kanzure | http://www3.interscience.wiley.com/journal/118565688/abstract?CRETRY=1&SRETRY=0 | 23:01 |
fenn | do you have any unusual allergies? | 23:01 |
kanzure | No. | 23:03 |
kanzure | None to begin with. | 23:03 |
* nsh doubts it's contamination | 23:03 | |
kanzure | Although I was terribly sick child .. asthma + hyperactivity ftw ;-) | 23:03 |
nsh | glutamate's purpose, as a food additive, is to overexcite nerves | 23:03 |
kanzure | that wasn't the cause of course | 23:03 |
nsh | it's what it does... | 23:03 |
kanzure | nsh: Anxiety? I don't recall that being an enhanced symptom after consumption. | 23:04 |
kanzure | Too long ago to remember. | 23:04 |
kanzure | although I am interested in substances that increase anxiety | 23:04 |
nsh | neural nerves, not emotional nerves :-) | 23:04 |
fenn | it's not to overexcite nerves, it's to make it taste like beef broth ( = hydrolyzed protein = amino acids) | 23:04 |
kanzure | $100 to anybody who can get me anxious enough to stop telling everybody all me sekr9ts or whatever | 23:05 |
fenn | and why isnt there some equivalent controversy over nitrous oxide in Guinness beer? | 23:05 |
fenn | or whipped cream | 23:05 |
spookact | increase anxiety? I'm more interested in decreasing :) I like my GABA, this big bottle of Ativan here says so | 23:05 |
kanzure | 'Is the Inattentive Type of ADHD a Separate Disorder?' <-- Haha. | 23:05 |
fenn | heh kanzure, your self-publicity is your greatest asset | 23:06 |
kanzure | "biohacking toolkit! bryan bishop!" <-- yeah, greatest asset | 23:06 |
kanzure | pfft | 23:06 |
kanzure | spookact: I can sometimes arbitrarily modulate anxiety. I also notice interesting correlations with those with abnormally high levels of anxiety. | 23:07 |
* nsh has a somewhat dissociational reactionm to anxiety | 23:08 | |
nsh | i am aware that i am physiologically anxious, but my conscious 'ego' (for want of a better term) is not overly affected | 23:09 |
fenn | headache, flushing, sweating, swelling of the throat, chest pain, heart palpitations, and shortness of breath. <- sounds like immune reactions | 23:09 |
biopunk | kanz: between themselelves as a group or between you and them? | 23:09 |
kanzure | biopunk: themselves? | 23:09 |
kanzure | Who? | 23:09 |
kanzure | oh | 23:09 |
biopunk | you talked about correlations | 23:09 |
kanzure | Correlations as in, they have other interesting side-effects. | 23:09 |
biopunk | you don't feel anxiety? | 23:10 |
biopunk | cool | 23:10 |
nsh | it's interesting | 23:10 |
nsh | for example, if i'm asking a question in an important conference | 23:10 |
nsh | i can think quite rationally, though my heart races and voice starts to falter | 23:11 |
nsh | i think it's a relic of a close encounter with a psychotic break a few years ago | 23:11 |
nsh | during which i was constantly anxious over a period of a few months | 23:12 |
biopunk | it sounds like something very normal to me | 23:12 |
kanzure | 'The Center for Rapid Automated Fabrication Technologies intends to | 23:12 |
kanzure | construct "a building a custom-designed house in a day" by 2012. Does | 23:12 |
kanzure | that seem feasible?' | 23:12 |
fenn | not many people would describe that as 'close encounter with a psychotic break' | 23:12 |
nsh | that was just one aspect, fenn... | 23:12 |
kanzure | biopunk: No, I do experience anxiety. | 23:12 |
fenn | is "building a custom-designed house in a day" supposed to be an exhibit or something? | 23:13 |
biopunk | kanz... It's a good thing I think | 23:13 |
nsh | most classically symptomatic was an almost complete epistemological catastrophe, religious experience, delusions of grandeur, magical thinking, etc. | 23:13 |
nsh | *more | 23:14 |
kanzure | fenn: he goes on to mention hexayurt in the next paragraph, so take it with a grain of salt | 23:14 |
fenn | say.. seeing how this is #hplusroadmap, er, how do we engineer our way around psychotic episodes and MSG-attacks? | 23:14 |
fenn | kanzure: you dont like hexayurt? | 23:14 |
nsh | good early detection and aversion systems | 23:14 |
fenn | its not exactly a house.. but it's rapid fabrication | 23:15 |
kanzure | fenn: I thought you don't :p | 23:15 |
kanzure | 'Given the current industrial paradigm, we're all screwed. IKEA might have it right for the big places, but for the smaller places we really need to fix the broken dependencies.' (talking about material dependencies) | 23:15 |
kanzure | the latest post by Smuri on http://groups.google.com/group/openmanufacturing/ is interesting | 23:16 |
fenn | did i say that? | 23:16 |
kanzure | he's on some fucked up island south of Africa | 23:16 |
kanzure | and he can hardly get 4x8 plywood | 23:16 |
kanzure | so what's he supposed to do ? :) | 23:16 |
kanzure | fenn: methinks you did, I'll double check / not hold you to it (one of these) | 23:16 |
kanzure | I'll be right back .. need to steal some food from the cafeteria | 23:16 |
fenn | wow 20 messages in a day? | 23:16 |
kanzure | hm? | 23:16 |
fenn | openmanufacturing | 23:16 |
kanzure | meh | 23:16 |
kanzure | I've seen worse ;-) | 23:17 |
fenn | yeah but it's brand new | 23:17 |
kanzure | p2pfoundation.net is kind of seeding it | 23:17 |
fenn | oh.. did nathan cravens do anything else besides freeconomy? | 23:17 |
kanzure | not sure | 23:17 |
kanzure | rawr, bbl | 23:17 |
kanzure | re: material dependency brokenness -- this is why he's indeed fucked .. you shouldn't colonize unless you're damned sure you have the available materials, and if not, then you should be damned prepared to trade until the end of oblivvion | 23:19 |
fenn | trade? if he cant get what he needs, what to trade for? | 23:20 |
biopunk | the hexayurt was cool.. it was new to me | 23:20 |
fenn | i liked the inflatable radio antenna sitting next to it even better :{ | 23:21 |
biopunk | ... i watched a youtube clip | 23:21 |
ybit | anyone know of say... a neuroengineering graduate school in spain? :) | 23:51 |
kanzure | in spain? | 23:51 |
kanzure | who the hell lives in spain? | 23:51 |
ybit | s? se?or | 23:51 |
ybit | ha | 23:51 |
kanzure | ybit: but seriously, the comp-neuro mailing list is full of europeans | 23:52 |
ybit | yep, but not seeing many positions posted for a spanish uni | 23:52 |
ybit | i think this may be the place, but catalan is annoying http://www.upc.es/ | 23:52 |
ybit | while i'm at it, might as well diss the galician language as well :) | 23:53 |
ybit | if i live in spain, i want everyone to speak spanish! :P | 23:54 |
ybit | they don't have techshop, but they do have a fablab in barcelona | 23:55 |
ybit | speaking of, is your school going to get one kanzure? | 23:55 |
kanzure | Working On It. | 23:55 |
kanzure | ybit: Don't look for positions being posted. | 23:56 |
ybit | i'm not | 23:56 |
ybit | my email box is bombarded by them occasionally though, so i take a look | 23:56 |
kanzure | It shouldn't be too hard to retrieve a full list of universities in Spain. | 23:56 |
ybit | ah, and then spider for keywords, riight | 23:56 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!