| --- Day changed Sun Jan 18 2009 | ||
| kanzure3 | fenn, did I send you the Saul Griffith calculations? | 00:00 |
|---|---|---|
| fenn | dunno, i dont remember seeing it | 00:00 |
| kanzure3 | something about needing an olympic-sized stadium of algae every 2 seconds to maintain growing rates of energy demand of human civilization. | 00:00 |
| genehacker | well I got a use for algae | 00:01 |
| kanzure3 | "Half a terawatt of biofuels? Something like one Olympic swimming pools of genetically engineered algae, installed every second. (About 15,250 square miles a year, times 25.)" | 00:01 |
| kanzure3 | ^ at that rate you have to worry about the amount of phosphorous being redirected out of the local environment | 00:01 |
| kanzure3 | s/phosphorous/phosphorous and other elements and their systems/ | 00:02 |
| genehacker | biofuel is a temporary solution | 00:02 |
| kanzure3 | (since you're dealing with exponentials) | 00:02 |
| genehacker | the killer app for algae is feeding cows | 00:02 |
| genehacker | or maybe synth meat | 00:03 |
| kanzure3 | fenn: fwd'd. | 00:03 |
| fenn | ah yes i saw the soda bottle calculations in make magazine | 00:04 |
| fenn | bottle contains more energy than the soda | 00:04 |
| genehacker | damn WD 40 isn't fish oil | 00:04 |
| fenn | snake skin isn't alligator hide! | 00:05 |
| genehacker | I wanted to sell algae for WD 40 | 00:05 |
| kanzure3 | speaking of make, did you see that I got a hold of Zach Hoecken (thingiverse)? and that the ponoko CEO wants to do packaging formats too? | 00:05 |
| genehacker | unrelated | 00:05 |
| fenn | havent got to OM yet | 00:05 |
| kanzure3 | oomlaut was a bit hostile to my approach though. | 00:05 |
| kanzure3 | alec mentioned some passing interest in my correspondence with him.. | 00:06 |
| kanzure3 | well, a bit more than passing. | 00:06 |
| genehacker | so I did some calculations a while back too, it would take less energy to make an equal amount of diamond than it would of plastic | 00:06 |
| fenn | wtf is wrong with nuclear power? | 00:07 |
| fenn | you never EVER see a detailed analysis of the environmental effects of nuclear power in popular print | 00:07 |
| genehacker | nothing, chernobly just freaked people out | 00:07 |
| genehacker | did you hear about that neighborhood nuclear reactor company fenn, they're sold out for 3 years | 00:08 |
| genehacker | they can't keep up with demand | 00:08 |
| fenn | is it a beta battery? or some kind of thermoelectric thing? | 00:09 |
| fenn | RTG | 00:09 |
| genehacker | a bona fide nuclear reactor I think | 00:09 |
| kanzure3 | wtf, keeping up with demand is easy | 00:09 |
| kanzure3 | sounds to me like mostly an architectural issue with their business model .. | 00:10 |
| kanzure3 | not scaling fast enough.. | 00:10 |
| fenn | "building nuclear reactors is easy" really | 00:10 |
| genehacker | http://www.technovelgy.com/ct/Science-Fiction-News.asp?NewsNum=1993 | 00:10 |
| genehacker | 25 mW for 40 years | 00:10 |
| kanzure3 | didn't say it was easy | 00:10 |
| kanzure3 | but they are also not sprinkling magic into it | 00:10 |
| kanzure3 | nor should they be making custom designs each time, so .. | 00:11 |
| genehacker | building nuclear reactors is easy, building safe ones on the other hand... | 00:11 |
| kanzure3 | http://books.slashdot.org/comments.pl?sid=1094497&cid=26489733 "That's what SOA aims at: interchangeable components in systems. You're not crafting one big program, or complex of programs, from end-to-end, making it up as you go. You're building uniformly-structured and interchangeable components, and assembling them." | 00:12 |
| kanzure3 | "You mean... like Unix?" | 00:12 |
| kanzure3 | *cough* http://heybryan.org/shell.html *cough* not quite implemented anywhere to my knowledge .. not even SOA, WSDL, B2B, shells, debs, .. | 00:12 |
| genehacker | speaking of uniform components I've been having second thoughts about making a replicator out of legos | 00:13 |
| genehacker | fuck legos, I can make my own blocks | 00:13 |
| genehacker | designing the blocks is the hard part | 00:14 |
| kanzure3 | genehacker: so, you remember the recent two-actor ribozyme replication duo, right? | 00:15 |
| genehacker | yes, yes | 00:15 |
| kanzure3 | in the article it said there was something like 30 bits of information that could be channeled into each of the designs. | 00:15 |
| genehacker | that's why I want to use blocks | 00:15 |
| kanzure3 | hold on a sec though | 00:15 |
| kanzure3 | this is a two-actor system and the set of things that they can make is going to be physically limited | 00:15 |
| kanzure3 | it's a very big possibility space though, I'd give you that | 00:16 |
| fenn | two-actor system? | 00:16 |
| kanzure3 | that's the "minimal cell" or "minimal program" approach | 00:16 |
| genehacker | yeah, I want to make a build system though with parts like pipes, linear bearings | 00:16 |
| kanzure3 | fenn: there were two complementary replication agents working together. polymerase and | 00:16 |
| genehacker | two actor system, one does the replication, one provides the information to replicate | 00:17 |
| kanzure3 | well. my brain just shut off apparently | 00:17 |
| fenn | ok, von neumann cellular automata basically | 00:17 |
| genehacker | yeah | 00:17 |
| genehacker | something like that I guess | 00:17 |
| kanzure3 | anyway, my point was going to be about the alternative to starting from the bottom | 00:18 |
| genehacker | I'm not familar with von neumann cellular automata | 00:18 |
| kanzure3 | from the small starting set of those 30 bits of information (actually it's not 2-bit, and it's not 4-bit either because there are some strings of 4-bits that are practically chemically equivalent) | 00:18 |
| kanzure3 | you could try to extract and draw out, say, a human, or a giraffe from those bits | 00:19 |
| kanzure3 | and try to stuff as much non-directly-replication-design-stuff as possible, trying to figure out how to make it work in that model | 00:19 |
| kanzure3 | or | 00:19 |
| genehacker | well the advantage of designing a build system, is that we can sell it as a toy to fund research | 00:19 |
| fenn | it's 2d, there's a machine sort of like a turing machine that moves around the digital space in response to commands on a tape (also represented in the digital space) and it can change the value of a state at will (sort of like creating/destroying legos) | 00:19 |
| kanzure3 | you go the other way: you already have your machine shop prototyped and can just trace the dependencies | 00:20 |
| kanzure3 | fenn: maybe you can see the point that I was going to highlight | 00:20 |
| kanzure3 | but it's not an important one at the moment, just thought like mentioning it. | 00:20 |
| genehacker | the Charles Collins replicator is a physical manifestation of that, a way too literal physical manifestation | 00:20 |
| genehacker | yeah I know | 00:21 |
| fenn | DNA is quad, even if the conversion to amino acids is redundant | 00:21 |
| kanzure3 | blah, fine. | 00:21 |
| fenn | 3-quad code | 00:21 |
| genehacker | I wouldn't make that comparison | 00:21 |
| genehacker | heh have you heard about tricore chips? | 00:21 |
| * kanzure3 is running a quadcore. | 00:22 | |
| genehacker | you know what a tricore chip is? | 00:22 |
| fenn | can they vote on the answer? :) | 00:22 |
| kanzure3 | we call that a race condition | 00:22 |
| fenn | super duals or sucky quad-cores? | 00:22 |
| genehacker | it's a quadcore with a defective core cut out | 00:22 |
| genehacker | really | 00:22 |
| genehacker | that's what it is | 00:22 |
| fenn | ok now they just need to scale that concept | 00:22 |
| fenn | by about ten orders of magnitude | 00:23 |
| fenn | "i'll take a wafer, thank you" | 00:23 |
| genehacker | you can fenn, just ask them, some hand em out for free | 00:23 |
| genehacker | I got one in my room | 00:23 |
| fenn | eh? like the pope? | 00:23 |
| fenn | i have some defective large silicon dies (dice?) from electronic goldmine.. | 00:24 |
| genehacker | heh, I'm sure in the church of saint turing they will | 00:24 |
| fenn | the probability of a defect goes up with area, so that's why large chips are much more expensive | 00:24 |
| genehacker | I have a silicon wafer with the chips made on it, makes a cool diffraction pattern when you shine a laser on it | 00:24 |
| genehacker | a laser which I happened to break :( | 00:25 |
| fenn | but if yo have a large number of identical/commodity elements on the chip that you can route around, you might as well use the whole wafer as a supercomputer | 00:25 |
| genehacker | well you see fenn, you have to cut them, then you have to package them, then you have to do a little dance | 00:26 |
| genehacker | btw, does anyone have any HD-DVD drives they are willing to give me? | 00:27 |
| genehacker | cutting them out could be expensive | 00:29 |
| fenn | the wafer's already cut, into a wafer | 00:29 |
| fenn | so you just have to bond wires to the edges and do some magic re-routing foo in the silicon | 00:29 |
| fenn | flip chip packaging would probably be best | 00:30 |
| genehacker | you gotta package them fenn | 00:30 |
| fenn | i dont understand | 00:30 |
| genehacker | you gotta get rid of heat | 00:30 |
| fenn | of course | 00:30 |
| fenn | however, if you arent using sucky x86 processors... there is much less heat | 00:30 |
| genehacker | interesting idea though | 00:30 |
| genehacker | you gotta get rid of heat | 00:31 |
| fenn | ok well I dont see what "packaging" has to do with that | 00:31 |
| fenn | for example, why can't I just use aluminum substrate PCB and conduct heat through the solder connections? | 00:32 |
| genehacker | give it a try, if you can solder to them | 00:32 |
| genehacker | do you have a really accurate CNC machine? | 00:33 |
| fenn | i think it has copper on the top layer, "metal core" | 00:33 |
| fenn | no, i dont have a functioning CNC at all right now | 00:33 |
| genehacker | oh packaging also helps prevent tiny bits of metal from shorting your chip out | 00:33 |
| fenn | chip surface would be pressed up against the circuit board (and probably sealed around edges with epoxy) | 00:35 |
| genehacker | solder a chip to a circuit board, then we can talk | 00:35 |
| fenn | i dont have any chips with solder | 00:36 |
| fenn | i'm talking mass production, not DIY stuff | 00:36 |
| genehacker | the connections are very very small | 00:37 |
| genehacker | oh | 00:37 |
| genehacker | ok | 00:37 |
| genehacker | well it looks like it might be interesting to try | 00:37 |
| fenn | it's basically the same idea I had for self assembling DNA-tile FPGA's | 00:38 |
| genehacker | haven't heard that one | 00:38 |
| genehacker | hey we could form a company around this | 00:38 |
| kanzure | *facepalm* | 00:39 |
| kanzure | sometimes I feel like a ghost around here. | 00:39 |
| genehacker | maybe you are | 00:40 |
| genehacker | how do you know you aren't? | 00:40 |
| -!- fenn changed the topic of #hplusroadmap to: no philosophy, please | 00:40 | |
| genehacker | oh shoot forgot about that | 00:40 |
| genehacker | so how can we design replicatable super-blocks? | 00:41 |
| fenn | start by defining what you mean by replicatable | 00:43 |
| genehacker | easy to replicate | 00:44 |
| fenn | -_- | 00:46 |
| genehacker | capable of replication, particulary fast replication | 00:48 |
| genehacker | I'm thinking something along the lines of modular robots | 00:55 |
| fenn | bitcoin: promoting the largest waste of CPU power, ever! | 00:55 |
| kanzure | welp, there goes my entire bubble of safety. /home/bryan/dad/Dad's Files/Brett's Bar and Grill/Apogee Publishing/Projects/Nascency/artificial-intelligence-risk[1].pdf | 01:02 |
| kanzure | that's a paper written by Eli :( | 01:02 |
| genehacker | ??? | 01:03 |
| genehacker | http://www.roblocks.org/ | 01:06 |
| genehacker | if only magnets weren't so hard to replicate | 01:06 |
| genehacker | it'd be interesting if neodymium magnets occurred in nature, be interesting to see what would evolve | 01:08 |
| kanzure3 | http://heybryan.org/dad/Biography.html <- found in his archives. | 01:08 |
| genehacker | ok | 01:10 |
| genehacker | AI probably won't be that dangerous | 01:10 |
| kanzure | orly? | 01:10 |
| genehacker | if we can get it to think it's a human too | 01:11 |
| kanzure | http://heybryan.org/eli.html <- within his system of thought, he's absolutely completely correct. But read that page for the actual objections. | 01:11 |
| kanzure | no. | 01:11 |
| genehacker | I detect philosophy | 01:12 |
| fenn | hmm roblocks looks cool but i dont know why they tried to make "programming" blocks | 01:16 |
| genehacker | for the kids | 01:17 |
| fenn | there was some open source project to make swivel blocks like in that hod lipson video | 01:17 |
| fenn | with cad files etc | 01:18 |
| genehacker | I think that is or was roblocks | 01:18 |
| -!- wrldpc [n=ben@173.48.207.37] has joined #hplusroadmap | 01:18 | |
| kanzure | I haven't seen that one. | 01:18 |
| fenn | they were making the blocks with some sort of RP machine | 01:18 |
| fenn | it was university based, they definitely didnt have a "catalog" | 01:18 |
| fenn | hmm CMU maybe it was the same site | 01:19 |
| genehacker | well roblocks are made with an RP machine | 01:19 |
| kanzure | *cough* LCD on the third image.. | 01:19 |
| kanzure | unless you rapid prototype LCDs now? | 01:20 |
| genehacker | ??? | 01:20 |
| genehacker | the plastic is | 01:20 |
| genehacker | http://www.molecubes.org/ | 01:20 |
| genehacker | is this it? | 01:20 |
| fenn | no it was a different site.. after looking at archive.org | 01:20 |
| fenn | yes molecubes.org | 01:21 |
| kanzure | say, I remember this one | 01:21 |
| kanzure | it's been mentioned in here before methinks | 01:21 |
| genehacker | Hod Lipson is involved with that | 01:21 |
| fenn | ah cool it's wei-min shen's group | 01:23 |
| genehacker | so what we really need is a way to connect cubes together, and stick without magnets | 01:23 |
| kanzure | .. that's what the interop stuff has been about | 01:23 |
| kanzure | but I'm merely a ghost | 01:23 |
| kanzure | don't bother listening to me. | 01:23 |
| kanzure | ~~~~~~ | 01:23 |
| genehacker | interop? | 01:23 |
| kanzure | interoperability | 01:24 |
| genehacker | oh | 01:24 |
| kanzure | intercompatibility | 01:24 |
| fenn | hunting of the snark | 01:24 |
| kanzure | huh? | 01:24 |
| genehacker | how do you achieve it? | 01:24 |
| kanzure | http://heybryan.org/shell.html has some notes. | 01:24 |
| kanzure | debtags is one partial solution, but they wimped out kinda | 01:24 |
| fenn | genehacker: dovetails with locking tabs works | 01:25 |
| fenn | genehacker: i have these tiny boxes for storing SMD components | 01:25 |
| kanzure | I must be thinking of a different level here | 01:25 |
| kanzure | surely there is more than just mechanical linkages to consider | 01:25 |
| kanzure | unless you're interested in doing a totally restrictive regime (ex: legos) | 01:25 |
| genehacker | think something that a blind person wearing mitts could assemble | 01:26 |
| kanzure | ("all things must have these dots to this precision") | 01:26 |
| fenn | heh anders always shows up in my image searches: http://www.nada.kth.se/~asa/bilder/box.jpg | 01:26 |
| genehacker | yeah we need something ajdjustable | 01:26 |
| fenn | why adjustable? | 01:26 |
| kanzure | lego = "planned economy" | 01:26 |
| kanzure | *legos | 01:26 |
| fenn | nothing wrong with planned economies | 01:27 |
| genehacker | so you aren't restricted to the dots | 01:27 |
| kanzure | right, legos are good, but we suddelny realize at some point that we want more than plastic lego bricks to connect together, and we realize that we can't replan it all (since we're not lego) | 01:27 |
| fenn | how often does that issue really come up? | 01:27 |
| kanzure | recipe substitution. | 01:28 |
| fenn | but you dont have to replan it all, you just need the source code and build capability | 01:28 |
| genehacker | what do you mean we aren't lego? | 01:28 |
| kanzure | take the "100% fractal automation" bullshit article from the other day | 01:28 |
| fenn | and then you make a DIP-14 brick or whatever | 01:28 |
| kanzure | they were doing top down centralized planning for everything | 01:28 |
| kanzure | their bricks incorporated the pipes, wires, etc., built into the bricks | 01:28 |
| kanzure | now if you wanted something new | 01:29 |
| kanzure | the entire standards would have to be updated | 01:29 |
| kanzure | and probably the old stuff too.. | 01:29 |
| fenn | no way | 01:29 |
| fenn | look at NXT | 01:29 |
| fenn | it's still backwards compatible | 01:29 |
| genehacker | well we can make lego "monster bricks" we have the technology | 01:29 |
| genehacker | we can remake lego in our own image | 01:30 |
| fenn | I don't | 01:30 |
| kanzure | backwards compatibility is not reusability. it's legacy maintenance.. | 01:30 |
| fenn | most people don't realize the amount of work that goes into Lego injection mould dies | 01:30 |
| fenn | the die makers are treated like priests I hear | 01:30 |
| genehacker | fenn we have free 3d printer acess | 01:30 |
| fenn | is that good enough to make legos? | 01:31 |
| kanzure | fenn: why doesn't Google have their own lego line? | 01:31 |
| fenn | because they are software people? | 01:31 |
| kanzure | they're the ones that played with legos though | 01:31 |
| fenn | huh? | 01:31 |
| genehacker | the SLS printer can print a very detailed locket | 01:31 |
| kanzure | isn't lego-fandom a requirement to employability at Google? | 01:31 |
| fenn | possibly.. | 01:31 |
| fenn | i think you also have to use a macbook and wear thick rimmed glasses | 01:32 |
| kanzure | ew. | 01:32 |
| genehacker | think about it, if we make legos grow on tree... | 01:32 |
| kanzure | genehacker: are you having a separate conversation right now? | 01:32 |
| kanzure | heh | 01:32 |
| genehacker | cool I have two qualifications for google employment | 01:32 |
| kanzure | fenn: would debian count as a planned economy? | 01:33 |
| fenn | yes | 01:33 |
| kanzure | yet individual packages aren't really. | 01:34 |
| fenn | it doesn't fit the metaphor perfectly | 01:34 |
| kanzure | in legos, it seems to be top-down novelty stuff going on | 01:34 |
| kanzure | (top-down as in corporate management bosses stuffs) | 01:34 |
| fenn | debian is an aggregator.. there isnt really any equivalent in economics except maybe feudalism | 01:34 |
| genehacker | well, I'm going to jack out now guys, so good night | 01:35 |
| fenn | ciao babe | 01:35 |
| kanzure | one day we'll all have the same conversation at once, and it will be good | 01:35 |
| genehacker | chatrooms aren't very productive | 01:36 |
| kanzure | meh, that's questionable | 01:36 |
| kanzure | fenn: "how often does that issue come up?" <- I think that could also be asked as "how often do projects break their debian packages, or don't quite fit the model" | 01:37 |
| fenn | ah finally found the configurable boxes http://www.dealextreme.com/feedbacks/BrowseReviews.dx/sku.4580 | 01:40 |
| kanzure | constructal theory: all of the most inefficient decisions are made the most frequently in a flow system (i..e, in human transportation, it's individual decisions, rather than mass transit as a norm), same with packaging, and tools to help make better decisions for packaging even in the absence of all out 'instruction formats' and interoperability of parts, would be nice. | 01:40 |
| fenn | they snap together, hold firmly, and come apart easily with the proper massaging | 01:40 |
| fenn | and there are other sizes | 01:40 |
| fenn | seems the concept could be extended to other stuff than just boxes | 01:40 |
| fenn | it's hard to see from the pictures but there's a tapered dovetail | 01:41 |
| fenn | the taper is slightly an interference fit, so the plastic flexes a tiny bit and makes it a snug connection | 01:41 |
| kanzure3 | nice. | 01:42 |
| fenn | tools to help make better decisions for packaging... well, that sounds like skdb | 01:42 |
| fenn | to be compatible with the widest variety of parts requires comparing to an inhumanly large set of other things you might want to connect to | 01:43 |
| fenn | on the other hand i just want to bomb a bunch of chinese factories that make single purpose consumer crap that isnt compatible with anything | 01:43 |
| fenn | and just force everyone to use standard parts | 01:44 |
| kanzure3 | with the 'instruction format' though, you don't need to "help" make decisions, that can be all automated away mostly | 01:44 |
| fenn | instruction format == recipe? | 01:44 |
| kanzure3 | but I'm wondering about what specifically in the absence of that since the instructions/recipe-substitution stuff (ability to snap different abstract parts together, whatever standards they might beu sing) | 01:44 |
| kanzure3 | yes | 01:44 |
| kanzure3 | sorry | 01:44 |
| kanzure3 | I'm kind of talking about three things but it's the same thing | 01:44 |
| kanzure3 | instructions, recipes, and the interoperability all reference the same shell.html stuff (the dselect select-your-own-usage-of-the-program-or-part/tool) | 01:45 |
| kanzure3 | yn | 01:46 |
| fenn | software has gotten away with a lot of cruft because it's so easy to replicate and storage space is expanding faster than we can fill it with code | 01:46 |
| fenn | but with hardware we can't assume that you'll have some archaic editor from the 70's | 01:46 |
| kanzure3 | um, absence of that (since the .. doesn't exist nor can I hack out a demo) | 01:46 |
| kanzure3 | ^ add a * somewhere in that last message. | 01:46 |
| kanzure3 | (beginning) (this is only making my attempt to correct an error even worse) | 01:47 |
| fenn | so, what you're saying is.. ? | 01:47 |
| fenn | I liked that kanzure guy better | 01:47 |
| fenn | he seemed to make sense | 01:47 |
| fenn | some of the time at least | 01:47 |
| fenn | perhaps you need meta-instructions to explain how to use the instructions, like DTD | 01:48 |
| kanzure3 | given no instruction/recipe format or ability to do interoperability/intercompatibility at the moment, and nothing better than the original dimensional analysis of the ability of things to interconnect, | 01:48 |
| kanzure3 | what tools for helping making sound packaging decisions could we come up with? | 01:48 |
| fenn | packaging decisions meaning what? | 01:49 |
| kanzure3 | *what tools for helping to make these parts/tools datasets useful could we come up with, from the point of view of a package maintainer who doesn't necessarily see "the larger picture"? | 01:51 |
| kanzure3 | I don't think that helps though, hrm. | 01:51 |
| fenn | you're familiar with providing a comparison function for sorting lists right? | 01:51 |
| kanzure3 | list1 > list2? | 01:52 |
| fenn | i think we need to provide comparison functions for interface compatibility | 01:52 |
| kanzure3 | so, what, "this part accepts IETF standard RJ45" | 01:52 |
| fenn | like, if you wanted to sort colors you would lookup the color name in some table, find the RGB values, convert to HSV, then compare the H values of two colors | 01:52 |
| fenn | i think there has to be some physical simulation involveed | 01:53 |
| kanzure3 | this just seems to be "tagging the inputs and outputs, and then checking if the tags of a possible machine to connect to overlaps with the input/output of the other machine" | 01:53 |
| fenn | if the blackbox data already fits we wouldnt need to provide a function.. this is like how we dont need to write a comparison function for sorting numbers | 01:54 |
| kanzure3 | machine 1: inputs (a, b, c) outputs (x, y, z) | 01:54 |
| kanzure3 | machine 2: inputs (x, y, z) outputs (a, b, c) | 01:54 |
| kanzure3 | machine1.out matches machine2.in | 01:54 |
| kanzure3 | where a,b,c,x,y,z would be tags for standards | 01:54 |
| kanzure3 | but doesn't this inherent the fundamental suckness of debtags? | 01:54 |
| fenn | no that's blackboxing, and requires lots of coordination and standards | 01:54 |
| fenn | (and work) | 01:54 |
| fenn | tags are useful, for some things | 01:55 |
| kanzure3 | I need you to elaborate on the alternative in this situation then, because tags don't solve it. | 01:55 |
| fenn | there must be some low level standard representation for physical quantities that everyone can agree on | 01:56 |
| fenn | everything else is just semantics | 01:56 |
| fenn | by which i mean "incompatible choice of words" | 01:56 |
| kanzure3 | but but then you just get to my crappy code where I was saying "input: 7 m^2" (of space) | 01:56 |
| fenn | if you really only need some arbitrary 7m^2 of area, then what's the problem? | 01:57 |
| fenn | if it must be planar, then you need to describe the plane with more constraints | 01:57 |
| kanzure3 | hrm, I suppose taking a hint from django and GNU autoproject would be nifty there .. deploy the common measurement utilities to make sure you have your workbench setup and stuff | 01:57 |
| kanzure3 | ok, blackbox that and assume that's fine | 01:58 |
| fenn | "tabletop" can be decomposed into a number of attributes, like normal vector (up) surface hardness, rigidity, impact resistance etc | 01:58 |
| kanzure3 | didn't we have some GUID funkyness with the materials class? | 01:58 |
| fenn | so then you can look at other surfaces and see if they fit the requirements of a tabletop | 01:58 |
| kanzure3 | right | 01:59 |
| fenn | but you only have to specify "tabletop" in the package requirements | 01:59 |
| fenn | all the magic happens invisibly | 01:59 |
| fenn | bleh | 01:59 |
| -!- parodyoflanguage [n=kevin@mmds-216-19-34-153.twm.az.commspeed.net] has joined #hplusroadmap | 01:59 | |
| fenn | do you see how this is different from tags? | 01:59 |
| parodyoflanguage | No philosophy? Aw.. :) | 02:00 |
| kanzure3 | well, yes, but no | 02:00 |
| kanzure3 | IEEE 1994 input isn't specified by dimensional units.. | 02:01 |
| kanzure3 | I am not being coherent am I? | 02:01 |
| fenn | it gets around the standards problem by encoding the standard in dimensional units, and comparing the units | 02:01 |
| kanzure3 | 59.2 IEEE 1994s | 02:01 |
| fenn | i believe signals can be represented by units and code | 02:01 |
| fenn | what is IEEE 1994? | 02:01 |
| kanzure3 | firewire? | 02:02 |
| kanzure3 | 1394? | 02:02 |
| fenn | ya | 02:02 |
| kanzure3 | yep, 1394. sorry about that. | 02:02 |
| fenn | so, firewire is both a protocol and an electronic signal and a set of connectors and cabling | 02:02 |
| fenn | the protocol is software and can be represented with BNF? | 02:03 |
| fenn | surely there must be some universal protocol representation language | 02:03 |
| kanzure3 | when you go buy a plastic-wrapped device off of Amazon, you get a list of tags of what it supports (IEEE 1394, RCJ 45, ...) | 02:03 |
| kanzure3 | hm. | 02:03 |
| fenn | you get tags from amazon? | 02:04 |
| kanzure3 | well, no, | 02:04 |
| kanzure3 | but it's in the product descriptions | 02:04 |
| kanzure3 | usually in a list or something | 02:04 |
| fenn | oh | 02:04 |
| kanzure3 | and thus how stuff shows up in search results | 02:04 |
| kanzure3 | a list on the side of the box, etc. like food labels. | 02:04 |
| fenn | almost useful info.. sigh | 02:05 |
| fenn | so, i hope whoever is making packages for representing protocols and electrical signals will look at what software libraries exist already before they start reinventing the wheel | 02:08 |
| fenn | oh, there has to be some base geometry classes too | 02:08 |
| kanzure3 | don't know if you took a look at ECIX when I was ranting about it the other day | 02:09 |
| kanzure3 | electronic component packaging XML standard stuff. | 02:09 |
| kanzure3 | (electronics being the golden standard of everything, all hail the electronics software people) | 02:09 |
| fenn | i suppose they already have a signal format/ontology | 02:10 |
| kanzure3 | when you're pushing things you can't see, you go hire thinkers to think abstractly for you anyway | 02:11 |
| fenn | wait, this is b2b stuff | 02:11 |
| kanzure3 | so maybe that's why that happened | 02:11 |
| kanzure3 | oh, is it? | 02:11 |
| kanzure3 | I thought I saw stuff like pin geometries, timing diagrams, | 02:12 |
| kanzure3 | Electronic Component Information Exchange | 02:12 |
| kanzure3 | http://archives.si2.org/si2_publications/#ECIX_Electronic_Component_Information | 02:12 |
| fenn | yeah digging through that | 02:12 |
| kanzure3 | http://archives.si2.org/si2_publications/ecix/pdf/ecixquickvcdemo.pdf | 02:14 |
| fenn | i think they just reference the relevant standard | 02:14 |
| fenn | it's not a standards format | 02:14 |
| fenn | god that's barely even a relational database | 02:15 |
| fenn | this is more the sort of thing i'm interested in http://archives.si2.org/si2_publications/pinpak/Specs/ECPinPak1.0.pdf | 02:16 |
| kanzure3 | yep | 02:17 |
| fenn | it really sucks that standards orgs are so closed | 02:21 |
| -!- parodyoflanguage [n=kevin@mmds-216-19-34-153.twm.az.commspeed.net] has quit ["Leaving"] | 03:56 | |
| -!- jm|afk [n=jm@p57B9FDD9.dip.t-dialin.net] has joined #hplusroadmap | 05:42 | |
| -!- jm|earth [n=jm@p57B9DC4F.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)] | 06:00 | |
| -!- genehacker_ [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has joined #hplusroadmap | 08:39 | |
| -!- genehacker [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has quit [Read error: 60 (Operation timed out)] | 08:45 | |
| -!- jk4930 [n=jk@p54BCCA93.dip.t-dialin.net] has joined #hplusroadmap | 09:05 | |
| -!- samrose [n=samrose@75-134-171-156.dhcp.bycy.mi.charter.com] has joined #hplusroadmap | 09:06 | |
| -!- samrose [n=samrose@75-134-171-156.dhcp.bycy.mi.charter.com] has quit [Remote closed the connection] | 11:02 | |
| -!- genehacker_ [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has quit [Read error: 60 (Operation timed out)] | 11:30 | |
| kanzure3 | http://www.boddie.org.uk/david/Projects/Python/ldraw/ | 11:51 |
| -!- PeerInfinity [n=someone@216.36.180.162] has joined #hplusroadmap | 12:31 | |
| -!- jm|afk [n=jm@p57B9FDD9.dip.t-dialin.net] has quit ["quit"] | 13:07 | |
| -!- jk4930 [n=jk@p54BCCA93.dip.t-dialin.net] has left #hplusroadmap ["Konversation terminated!"] | 13:23 | |
| -!- genehacker [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has joined #hplusroadmap | 13:33 | |
| genehacker | you there kanzure? | 14:47 |
| kanzure | yes | 14:48 |
| genehacker | want to make a reprap sometime this year? | 14:58 |
| kanzure | sure. | 15:05 |
| kanzure | yay | 15:46 |
| kanzure | ponoko ceo wants to do apt-get for hardware | 15:46 |
| genehacker | ok what do you think about forming an Austin Reprap User Group(RUG) | 16:01 |
| kanzure | why only reprap | 16:01 |
| kanzure | I mean, that's what the Austin Robot Group is for really | 16:01 |
| kanzure | they're the guys that have been working on repraps anyway.. | 16:01 |
| genehacker | they're working on repraps? | 16:02 |
| kanzure | a few of their members, yes. | 16:02 |
| genehacker | good point | 16:02 |
| kanzure | like I said, they commonly bring up desktop cnc machines to the meetings.. | 16:02 |
| genehacker | I was thinking of working with the actlab people | 16:02 |
| kanzure | so they're on the same wavelength. | 16:02 |
| kanzure | nothing wrong with the actlab people | 16:03 |
| genehacker | they aren't engineers | 16:03 |
| -!- genehacker [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has left #hplusroadmap [] | 16:04 | |
| -!- genehacker [n=chatzill@wireless-128-62-174-156.public.utexas.edu] has joined #hplusroadmap | 16:04 | |
| genehacker | how did I leave? | 16:04 |
| genehacker | so how small can a CNC be and still be useful? | 16:05 |
| genehacker | now who to ask in actlab about building a reprap? | 16:08 |
| kanzure | sandy and brandon | 16:08 |
| genehacker | thanks | 16:08 |
| genehacker | I think I'll just email the whole actlab | 16:11 |
| kanzure | mention my name and cc me please :) | 16:11 |
| genehacker | how do you want to be mentioned? | 16:13 |
| kanzure | positively? | 16:13 |
| kanzure | what are you asking them anyway | 16:13 |
| genehacker | if they'd be interested in finishing the reprap they tried to build | 16:16 |
| kanzure | ponoko is *really really* excited about this. | 16:18 |
| genehacker | about what? | 16:20 |
| kanzure | skdb | 16:20 |
| genehacker | about your SKDB? | 16:20 |
| genehacker | so I'll say you're interested in the project too | 16:29 |
| genehacker | I can't think of a way to mention you | 16:33 |
| genehacker | actlab email is broken | 16:42 |
| -!- wrldpc [n=ben@173.48.207.37] has quit [] | 17:03 | |
| genehacker | emailed Sandy | 17:12 |
| -!- parodyoflanguage [n=kevin@mmds-216-19-34-153.twm.az.commspeed.net] has joined #hplusroadmap | 17:59 | |
| -!- wrldpc [n=ben@c-98-217-181-39.hsd1.ma.comcast.net] has joined #hplusroadmap | 19:05 | |
| genehacker | hey you know of any facial recognition software kanzure? | 19:40 |
| kanzure | opencv or opencp or openmp | 19:40 |
| kanzure | it's one of those heh' | 19:40 |
| -!- wrldpc_ [n=ben@c-98-217-181-39.hsd1.ma.comcast.net] has joined #hplusroadmap | 20:01 | |
| -!- wrldpc [n=ben@c-98-217-181-39.hsd1.ma.comcast.net] has quit [Read error: 110 (Connection timed out)] | 20:15 | |
| kanzure3 | oh god. | 21:48 |
| genehacker | ??? | 21:48 |
| kanzure3 | http://bugs.kde.org/show_bug.cgi?format=multiple&id=132480 | 21:49 |
| genehacker | centimeter sized sheets of graphene | 21:49 |
| kanzure3 | you'll note that I submitted a patch on 2009-01-03 | 21:49 |
| kanzure3 | http://heybryan.org/projects/browsehack/tabtabtab.html | 21:49 |
| kanzure3 | you'll note that I apparently made a patch on 2008-05-26 | 21:49 |
| kanzure3 | completely forgot about that .. | 21:49 |
| kanzure3 | <-- has repeated his own work :( | 21:49 |
| kanzure3 | compare: (2008-05-26 patch) http://heybryan.org/projects/browsehack/2008-05-26.diff to (2009-01-03 patch) http://bugs.kde.org/attachment.cgi?id=29847 | 21:50 |
| genehacker | http://science.slashdot.org/article.pl?sid=09/01/17/2341232 | 21:50 |
| genehacker | shit happens, I recently burnt out an expensive laser that I almost broke from shorting a nine volt batter across it and a trimpot | 21:52 |
| genehacker | broke it when I forgot to short out the capacitor in the driver before hooking it up | 21:53 |
| -!- PeerInfinity [n=someone@216.36.180.162] has quit [] | 23:24 | |
| -!- wrldpc_ [n=ben@c-98-217-181-39.hsd1.ma.comcast.net] has quit [] | 23:41 | |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!
