| --- Log opened Thu Jun 04 00:00:31 2015 | ||
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 00:14 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 265 seconds] | 00:19 | |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 01:15 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 276 seconds] | 01:21 | |
| -!- zadock [~outsider@cthulhu.tuiasi.ro] has joined ##hplusroadmap | 01:38 | |
| archels | kanzure: haha I like how they put "peaceful purposes" on that hackerspace KickStarter | 02:06 |
|---|---|---|
| -!- Beatzebub [~beatzebub@d75-155-236-222.bchsia.telus.net] has quit [Quit: Beatzebub] | 02:12 | |
| -!- drewbot [~cinch@ec2-54-157-9-11.compute-1.amazonaws.com] has quit [Remote host closed the connection] | 02:58 | |
| -!- drewbot [~cinch@ec2-23-23-38-230.compute-1.amazonaws.com] has joined ##hplusroadmap | 02:59 | |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Ping timeout: 256 seconds] | 03:42 | |
| -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has joined ##hplusroadmap | 04:08 | |
| -!- Taek [~quassel@2001:41d0:1:472e::] has quit [Quit: No Ping reply in 180 seconds.] | 04:11 | |
| -!- Taek [~quassel@2001:41d0:1:472e::] has joined ##hplusroadmap | 04:13 | |
| -!- Adlai [~Adlai@unaffiliated/adlai] has quit [Ping timeout: 276 seconds] | 04:19 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has joined ##hplusroadmap | 04:31 | |
| -!- Adlai [~Adlai@82.166.20.29] has joined ##hplusroadmap | 04:33 | |
| -!- eudoxia [~eudoxia@r167-57-122-215.dialup.adsl.anteldata.net.uy] has joined ##hplusroadmap | 04:49 | |
| -!- zadock [~outsider@cthulhu.tuiasi.ro] has quit [Quit: Leaving] | 04:57 | |
| -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-actswyndkumbgdjn] has joined ##hplusroadmap | 05:03 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 252 seconds] | 05:07 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has quit [Remote host closed the connection] | 05:16 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 05:16 | |
| -!- Madplatypus [uid19957@gateway/web/irccloud.com/x-ocnawfkomvpqequj] has quit [Quit: Connection closed for inactivity] | 05:53 | |
| kanzure | neals: well there's also the austin robot group but it's a bunch of old retired people that only sometimes do stuff | 05:57 |
| -!- punsieve [~drandomtu@173-170-176-33.res.bhn.net] has joined ##hplusroadmap | 06:00 | |
| -!- archels [charl@unaffiliated/archels] has quit [Ping timeout: 264 seconds] | 06:00 | |
| -!- archels [charl@toad.stack.nl] has joined ##hplusroadmap | 06:04 | |
| -!- jdolan [~jdolan@208.83.72.113] has joined ##hplusroadmap | 06:06 | |
| -!- Acty [uid89656@gateway/web/irccloud.com/x-omsxdrokqwdijdib] has joined ##hplusroadmap | 06:23 | |
| -!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-hucuusyiwfavtcdv] has joined ##hplusroadmap | 06:24 | |
| -!- Betawolf [~matthew@xn--bta-yla.net] has joined ##hplusroadmap | 06:26 | |
| kanzure | according to the taq1a article, vitamin b6 is a required cofactor, so it would be interesting to look at vitamin b6 - working memory studies. | 06:28 |
| -!- justanot1eruser [~Justan@unaffiliated/justanotheruser] has joined ##hplusroadmap | 06:31 | |
| -!- EnLilaSko [~Nattzor@unaffiliated/enlilasko] has joined ##hplusroadmap | 06:33 | |
| -!- justanotheruser [~Justan@unaffiliated/justanotheruser] has quit [Quit: Reconnecting] | 06:33 | |
| kanzure | in-vivo microdialysis of working memory in freely moving pigeons http://www-brs.ub.ruhr-uni-bochum.de/netahtml/HSS/Diss/KarakuyuDilek/diss.pdf | 06:35 |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 06:40 | |
| -!- berndj [~berndj@azna.co.za] has quit [Ping timeout: 244 seconds] | 06:51 | |
| -!- berndj [~berndj@azna.co.za] has joined ##hplusroadmap | 06:53 | |
| kanzure | "I have used Transcriptic and I am looking into Arcturus. So far I think Arcturus is better in my opinion even though Transcriptic does more. Arcturus seems to be more DIY friendly, while Transcriptic caters towards more professional labs." | 06:58 |
| kanzure | https://www.arcturus.io/ | 06:58 |
| kanzure | hmm | 06:59 |
| kanzure | https://www.arcturus.io/don_andres/blue-prometheus | 06:59 |
| kanzure | https://forum.arcturus.io/t/open-source-insulin-project/32 | 07:00 |
| kanzure | https://forum.arcturus.io/users/ryan_bethencourt/activity | 07:00 |
| kanzure | <!-- ugly hack to run experiments without paying until implement credits --> | 07:01 |
| -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-actswyndkumbgdjn] has quit [Quit: Connection closed for inactivity] | 07:03 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 272 seconds] | 07:28 | |
| -!- TheoryCat [inbisible@unaffiliated/wye-naught/x-8734122] has joined ##hplusroadmap | 07:53 | |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 08:13 | |
| -!- narwh4l [~michael@unaffiliated/thesnark] has quit [Remote host closed the connection] | 08:22 | |
| maaku | "while Transcriptic caters towards more professional labs" that's the truth | 08:23 |
| maaku | that's where the money is though | 08:24 |
| -!- TheoryCat [inbisible@unaffiliated/wye-naught/x-8734122] has quit [Quit: Goodbye] | 08:24 | |
| kanzure | maaku: i think there's also money in pandering to the career types (like selling to developers) | 08:25 |
| maaku | imho i agree | 08:26 |
| maaku | at transcriptic it was more "can we make this cheaper than a grad student or postdoc or intern's hourly rate?" | 08:26 |
| kanzure | well those are pretty low rates as-is | 08:27 |
| maaku | i think more interesting things happen when you focus on the biohackers specifically | 08:27 |
| maaku | kanzure: shoot for the moon | 08:27 |
| maaku | mostly it's about scale though, running 1000 things in parallel where the grad student / postdoc / intern is physically limited | 08:28 |
| maaku | i can't seem to find how workflows are programmed on arcturus.io, am I blind? | 08:29 |
| kanzure | i don't think there are any workflows | 08:29 |
| kanzure | looked like just a few <select>s | 08:29 |
| maaku | :( | 08:31 |
| kanzure | for a while i was successfully tracking all mentions of diybio in the news and every blog post | 08:33 |
| kanzure | http://diyhpl.us/wiki/diybio/faq/news/ | 08:33 |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap | 08:33 | |
| kanzure | got distracted on other stuff though | 08:33 |
| -!- blueskin [~blueskin@unaffiliated/blueskin] has quit [Read error: Connection reset by peer] | 08:40 | |
| -!- blueskin [~blueskin@unaffiliated/blueskin] has joined ##hplusroadmap | 08:41 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 245 seconds] | 08:46 | |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 08:55 | |
| archels | is that service actually live or is it just a mockup? | 09:12 |
| kanzure | not sure | 09:19 |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Ping timeout: 264 seconds] | 09:22 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 265 seconds] | 09:38 | |
| -!- justanot1eruser is now known as justanotheruser | 09:41 | |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 09:41 | |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 272 seconds] | 09:49 | |
| -!- narwh4l [~michael@unaffiliated/thesnark] has joined ##hplusroadmap | 09:53 | |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 10:07 | |
| maaku | would there be interest in a "programming language" for specifying experiments? | 10:26 |
| eudoxia | what would that look like, more or less? | 10:27 |
| -!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 10:29 | |
| maaku | here's an example in a presumed dependently typed stack language : [H20 20ml provision Sample-X 5ml provision 2ml pipette 30deg 20min incubate] | 10:31 |
| maaku | it would be like an assembly language for specifying the experiment | 10:32 |
| kanzure | transcriptic provided one of these | 10:34 |
| maaku | kanzure: i know i wrote part of it ;) | 10:34 |
| maaku | it's got some flaws though | 10:34 |
| kanzure | .g site:groups.google.com/group/openmanufacturing recipes | 10:34 |
| yoleaux | http://ipv6.google.com/sorry/IndexRedirect?continue=http://www.google.com/search%3F%26q%3Dsite:groups.google.com/group/openmanufacturing%2520recipes%26btnI%3D&q=CGMSECoBBPgCAXIwAAAAAAAIaucY7ZjCqwUiGQDxp4NLxpHoXcnnhNHTvoA173AJDnQ8PTk | 10:34 |
| kanzure | wat | 10:34 |
| kanzure | maaku: try these? | 10:35 |
| kanzure | https://groups.google.com/d/msg/openmanufacturing/qNjuJFqq6X0/1ZSPXrSjGXoJ | 10:35 |
| kanzure | https://groups.google.com/d/msg/openmanufacturing/Wm54uayZ2FE/hWq_uXCIuKgJ | 10:36 |
| kanzure | https://groups.google.com/d/msg/diybio/8oMYA5LxbN0/8svEr4Z_SSMJ | 10:36 |
| maaku | yeah 'recipe' is the right word | 10:37 |
| kanzure | https://groups.google.com/d/msg/diybio/raIonrvAD-A/opHuqoLjyW4J | 10:37 |
| maaku | XML is maximal yuck though | 10:38 |
| kanzure | https://groups.google.com/d/msg/diybio/Gd8YfvEBRzU/7DoJuphNqhEJ | 10:38 |
| kanzure | http://diyhpl.us/~bryan/irc/pcr.xml | 10:38 |
| kanzure | yeah i agree | 10:38 |
| kanzure | english grammar parser for protocol -> program http://88proof.com/synthetic_biology/blog/archives/290 | 10:38 |
| kanzure | maaku: this was an important non-existing component of http://gnusha.org/skdb/ | 10:39 |
| maaku | *was*? are you using something now? | 10:40 |
| kanzure | no :-) | 10:40 |
| kanzure | also, it's really hard to decide whether a scripting language is appropriate or not. you get into "great now i have to analyze a program to learn things about it" problems, compared to "great now i have to define a new file format" problems. bunch of wacky sucky tradeoffs..... | 10:41 |
| eudoxia | yeah everyone hates accidentally turing-complete config formats | 10:42 |
| maaku | well it's something I've been thinking about ever since I left transcriptic. i was not happy with the protocol-description language developed there, although it is absolutely the right idea | 10:42 |
| kanzure | have you played with the python library they pumped out? | 10:42 |
| maaku | maybe. i played with a version that existed then which was python but pumped out a proprietary back-end language | 10:43 |
| kanzure | oh wait where did the repo go | 10:43 |
| kanzure | ah, "autoprotocol" | 10:43 |
| maaku | right | 10:43 |
| kanzure | http://developers.transcriptic.com/v1.0/docs/autoprotocol | 10:44 |
| kanzure | https://github.com/autoprotocol/autoprotocol-python | 10:44 |
| maaku | yeah that's it | 10:44 |
| kanzure | hard tradeoffs tho | 10:44 |
| maaku | what I worked on was the backend for compiling and scheduling autoprotocols | 10:44 |
| kanzure | i think you basically have to commit to some grammar and then always maintain backwards compatibility or something | 10:45 |
| kanzure | obv. i wanted this for all machine tools of any kind | 10:45 |
| maaku | kanzure: eh, that's the major design flaw which i hated | 10:45 |
| kanzure | and i was also working on a translation layer that dumps it into human-readable instructions | 10:45 |
| kanzure | for manual human operators | 10:45 |
| maaku | use a stack based language and move most of the grammar rules into type checking | 10:46 |
| kanzure | and then i began modeling humans as actuators and things like "maximum reasonable force exertion in which directions".. so that the planner would be able to pick when and when not to use a human. | 10:46 |
| maaku | makes it much, much more future-proofed | 10:46 |
| kanzure | and fenn has done a bunch of weirdo kinematic planning stuff | 10:46 |
| maaku | right, the encoded program is the experiment, the human readable thing is compiled from that (not the other way around) | 10:46 |
| kanzure | one major problem that skdb ran into was that nobody wanted to be a hardware package maintainer | 10:47 |
| maaku | can fenn do a low-cost industrial arm? that's the major secret sauce here | 10:47 |
| kanzure | who wants to sit around writing up machine instructions for building stuff if nobody is able to execute the instructions yet? | 10:47 |
| kanzure | so... incentives are a problem. | 10:47 |
| kanzure | er yeah i think fenn and i worked on an arm once | 10:47 |
| kanzure | http://diyhpl.us/~bryan/irc/2010-05-28_puma500.jpg (not fenn) | 10:48 |
| maaku | well you can get away without that automation by diy mechanical turking it | 10:48 |
| kanzure | yes but my point is that package maintainers wont spontaneously appear and do lots and lots of work like that | 10:48 |
| maaku | not until it is at the point of fully automated, yes | 10:49 |
| kanzure | not only that | 10:49 |
| eudoxia | maybe the incentive could be created by making packages for, say, a few OSE machines and the components they depend on? | 10:49 |
| kanzure | but they have to have the equipment themselves or something | 10:49 |
| eudoxia | then showing how you can use skdb to generate both human readable instructions and machine-readable blueprints | 10:49 |
| kanzure | because otherwise there's no reason for them to bother with all that other shit (they just wnat to make a screw or a car or whatever, why would they want to learn about this machine-readable programming stuff?) | 10:49 |
| kanzure | eudoxia: unfortunately that's unpossible because OSE machines are underspecified | 10:50 |
| eudoxia | kanzure: well, then, take a path through the possible configuration space and make a package on that | 10:50 |
| kanzure | how is that an incentive anyway | 10:50 |
| maaku | well my interest for the moment is more the diy bio hacking, which is an easier problem than generalized manufacturing | 10:51 |
| eudoxia | people would have to download skdb to produce the manual instructions for building the thing | 10:51 |
| eudoxia | (then again people could host that) | 10:51 |
| kanzure | eudoxia: that's dumb, someone can just copy-paste | 10:51 |
| kanzure | maaku: can you be more specific? you want machine-readable instructions, but no equipment? | 10:51 |
| maaku | no, no you absolutely need equipment | 10:52 |
| eudoxia | one possible incentive: make user-personalized recommendations for purchase of items that can be purchased. ie, given a budget and a location, it finds nearby shops where you can buy screws etc. within the required tolerances/buget | 10:52 |
| kanzure | maaku: there's also a set of problems here related to equipment/inventory bootstrapping. but everyone has different equipment and inventory, which makes this even harder. | 10:52 |
| eudoxia | that's something that would have to be created dynamically and won't be the same for everyone | 10:52 |
| maaku | kanzure: hrm my impression at least from transcriptic experience is that many of these machines (e.g. liquid chromatography, mass spectrometer, thermocycler, etc.) can be generalized | 10:53 |
| kanzure | maaku: you may also be interested in cubespawn | 10:54 |
| maaku | so you should be able to clone a protocol repository and have it tell you 'Needs: Tools A, B, C; Consumables X, Y, Z' | 10:54 |
| maaku | kanzure: maybe. 3d printing and/or automated CNC hasn't held my interest, perhaps because I haven't been sold on a big-picture vision | 10:56 |
| kanzure | well, i suggest constraining the problem size considerably, and writing a small specification doc, which i would be very happy to review and obliterate | 10:56 |
| kanzure | maaku: not sure if i can sell you on it right this very moment, but i will keep that in mind and construct some compelling words for you to read soon | 10:56 |
| maaku | kanzure: like a protocol specification for PCR? | 10:57 |
| kanzure | no i mean more general like a specification for the goals and requirements of an implementation of x | 10:57 |
| kanzure | (and success criteria) | 10:57 |
| kanzure | of the system you have in mind | 10:57 |
| kanzure | we did this for skdb but nobody was ever able to provide particularly great criticism http://diyhpl.us/cgit/skdb/plain/doc/package_spec.yaml | 10:58 |
| -!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-hucuusyiwfavtcdv] has quit [Quit: Connection closed for inactivity] | 11:00 | |
| @fenn | the other problem was that most machines not only do not have data on their specifications, but there isn't even an ontology in which to describe them | 11:00 |
| kanzure | hurts lots when every package maintenance activity you engage in means redefining the whole ontology or trying to not collapse the house of cards | 11:00 |
| @fenn | what are the relevant parameters of a machine tool? who knows... | 11:01 |
| kanzure | ideally there should be some way of abstraction where you don't have to define the entire universe first, but i haven't figured out how | 11:01 |
| kanzure | fenn: maaku would like to be sold on the idea of more-than-just-for-biology-projects | 11:02 |
| kanzure | and it's my lunch time so i'm off the clock | 11:02 |
| * kanzure vanishes | 11:02 | |
| @fenn | heh | 11:02 |
| @fenn | well for most people 3D printing is just a vector to get consumer crap to them, like amazon uses UPS currently | 11:03 |
| @fenn | but the real innovation is shortening the hardware compile time loop | 11:03 |
| @fenn | it's the difference between batch programming and interactive programming | 11:04 |
| maaku | fenn kanzure: so I'm sold on the nanofactory vision, sure. if I can have a little box with gas canister feedstock that can manufacture anything I care about, i get how that would be revolutionary | 11:10 |
| maaku | but with 3d printing, even generalized to non-plastic feedstocks, it's stlil a very significant supply chain network and capital requirements | 11:11 |
| -!- erasmus [~esb@unaffiliated/erasmus] has quit [Read error: Connection reset by peer] | 11:12 | |
| maaku | it's not a little box that does everything, it's a bunch of little boxes that do individual pieces which have to be integrated, and you stlil need a wide variety of feedstocks, ... | 11:12 |
| @fenn | but it's still less stuff than stocking every color of kitchen spoon | 11:13 |
| @fenn | right now every engineered product is essentially hand-coded with lots of intricate optimization for hardware cost | 11:14 |
| maaku | yeah but ... that's a productivity improvement. it makes distribution networks easier because you're just shipping raw materials around instead of finished products | 11:15 |
| @fenn | but once programmer time becomes more expensive than hardware cost i think we'll start seeing more and more standardized components being used in everyday products | 11:15 |
| -!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 11:15 | |
| maaku | i guess what I'm saying is I could be motivated to automate diy bio because synthetic biology can do things which are impossible now. diy 3d printing just means I can make something at home that I could have bought at the store | 11:15 |
| @fenn | sure, i've had this argument many times and it boils down to the fact that it's not useful unless you're a hacker | 11:16 |
| @fenn | or unless you buy hacker-derived products | 11:17 |
| -!- Viper168_ [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 11:18 | |
| * maaku goes back to working on bitcoin | 11:19 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 250 seconds] | 11:19 | |
| -!- Viper168_ is now known as Viper168 | 11:20 | |
| @fenn | i wonder why there is still no cooking robot | 11:21 |
| @fenn | seriously, mini donuts is the only thing we can figure out how to automate? | 11:22 |
| @fenn | .title http://www.amazon.com/Nostalgia-Electrics-MDF200-Automatic-Factory/dp/B005Q8XBRK/ | 11:24 |
| yoleaux | Nostalgia Electrics MDF200 Automatic Mini Donut Factory:Amazon:Kitchen & Dining | 11:24 |
| @fenn | hmm. "You do have to watch the donuts as they travel through the cooking process and flip some of them with an appropriate hot oil spatula to make sure they cook evenly on both sides. It is not a machine that you can just put the batter in and have donuts come out the other end like the description makes one believe. It does take a great deal of supervision" | 11:25 |
| @fenn | lame | 11:25 |
| * kanzure appears on a bolt of lightning | 11:26 | |
| kanzure | maaku: there are many benefits to being able to replicate all technology in a civilization, without human intervention | 11:26 |
| kanzure | maaku: perhaps the most interesting benefits are related to political neutrality | 11:27 |
| @fenn | yeah but a 3d printer is just the end point of a long chain of capital | 11:27 |
| @fenn | you arent going to make your own microchips on a 3d printer | 11:28 |
| kanzure | without the ability to just pick up and leave, you are more strongly tied to local politics | 11:28 |
| @fenn | maybe there has been some development of printable electronics since i last checked... | 11:28 |
| @fenn | polymer semiconductors etc | 11:28 |
| kanzure | additionally, the ability to automate the construction of hardware allows for other crazy things like space habitats that do interesting things, freitas-style AASM stuff, freitas-style KSRM stuff, as well as generic complex machinery necessary for various transhumanist projects. | 11:29 |
| maaku | kanzure: i was thinking more charitable uses, e.g. a shipping container machine shop that you could deploy anywhere on a moment's notice, e.g. disaster relief | 11:29 |
| maaku | kanzure: right ok, I'm on board with the AASM stuff | 11:30 |
| kanzure | from a transhumanist perspective, you are not going to get anywhere if you have to read 100 million words of technical documentation just to construct all of your transhumanist tech | 11:30 |
| maaku | but the answer there I would guess is to standardize the starting set | 11:30 |
| kanzure | possibly, yes. | 11:30 |
| @fenn | fortunately the elements are mostly standardized | 11:30 |
| maaku | with MakerBot-like open hardware designs | 11:30 |
| kanzure | you mean reprap-like | 11:31 |
| maaku | er, yeah | 11:31 |
| kanzure | makerbot is no longer open-source | 11:31 |
| kanzure | or it never was | 11:31 |
| kanzure | did they retcon their previous licenses? | 11:31 |
| @fenn | we've always been at war with MakerBot | 11:31 |
| kanzure | well we already knew that | 11:31 |
| @fenn | mumble mumble eastasia | 11:32 |
| kanzure | yes i understood the reference -_- | 11:32 |
| maaku | so http://diyhpl.us/cgit/skdb/plain/doc/package_spec.yaml is current? | 11:34 |
| @fenn | if you can consider v0.0.0 current | 11:34 |
| kanzure | (yes) | 11:34 |
| maaku | so i skimmed it, and maybe i missed it, but where are the procedure format specification? | 11:38 |
| kanzure | well i think that was cut | 11:38 |
| @fenn | is "procedure" referring to manufacturing processes? | 11:39 |
| kanzure | if you mean instructions or recipes then https://github.com/kanzure/skdb/tree/master/doc/proposals | 11:39 |
| maaku | fenn: yes the steps to take to turn the material dependencies/inputs into finished product | 11:39 |
| maaku | hrm ok. i think this can be considerably improved | 11:40 |
| kanzure | i am absolutely certain this can be improved | 11:40 |
| kanzure | and your input is very welcomed | 11:40 |
| maaku | there's some open problems here though | 11:40 |
| kanzure | or your replacements | 11:40 |
| * maaku commutes to the office, see y'all in 45min | 11:40 | |
| kanzure | see this is why i started using lightning bolts for transportation | 11:41 |
| @fenn | in http://github.com/kanzure/skdb/blob/master/doc/proposals/techniques.py i define the press fit technique, which is represented in some yaml recipe file as "!press *part1interface1 *part2interface34" | 11:43 |
| @fenn | there are a lot of possible techniques and it wasn't obvious where to begin: https://github.com/kanzure/skdb/blob/master/processes.yaml | 11:46 |
| @fenn | this is basically copied from a text book and formatted to be machine parseable | 11:47 |
| @fenn | but it doesn't really hook into the python/yaml type system | 11:48 |
| @fenn | huh i messed up... it says !process but it should be !technique | 11:49 |
| @fenn | a process is like a physical thing that just happens, but a technique is the application of that process to an end | 11:50 |
| -!- Beatzebub [~beatzebub@d75-155-236-222.bchsia.telus.net] has joined ##hplusroadmap | 12:03 | |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has quit [Ping timeout: 258 seconds] | 12:07 | |
| -!- nmz787_i [~ntmccork@192.55.54.42] has joined ##hplusroadmap | 12:12 | |
| kanzure | http://c2.com/cgi/wiki?PushDownGoalStack | 12:13 |
| maaku | kanzure: share me the secret of lightning transport someday plz :) | 12:14 |
| @fenn | "EverythingIsTerrible" there i made a metaprogramming pattern | 12:15 |
| maaku | kanzure: yeah i think we can do better than yaml / xml for the procedural definition | 12:16 |
| maaku | i think there's actually room for a proper programming language there | 12:16 |
| maaku | of course defining the types & instructions is where the ontology really is | 12:16 |
| kanzure | writing a new programming language should be out of scope | 12:17 |
| maaku | nix programming language | 12:17 |
| maaku | i mean intermediate machine language | 12:17 |
| @fenn | a huge factor in choosing yaml was the fact that most engineers suck at programming and i didn't want to use a full fledged programming language for package definitions because i wanted people to be able to contribute data | 12:17 |
| kanzure | well there's llvm intermediate representation | 12:17 |
| maaku | fenn: i want to be able to automate and recombinate procedures, and having a formal intermediate system helps there greatly | 12:18 |
| kanzure | formalizations are in-scope | 12:19 |
| @fenn | is this just a set of strong types? | 12:19 |
| -!- nmz787_i [~ntmccork@192.55.54.42] has quit [Quit: Leaving.] | 12:19 | |
| maaku | mostly | 12:21 |
| maaku | imo i think it would also be beneficial to have a stack-based execution model, but the onus is on me to show that | 12:21 |
| kanzure | choosing highly constrained scopes is also an important trick | 12:21 |
| @fenn | yes i like the stack model for recipes | 12:22 |
| maaku | yeah, frankly a automated machine shop is harder for me to reason about than a liquid handler + handful of bio tools | 12:22 |
| maaku | so i might start with a bio example | 12:22 |
| maaku | harder meaning larger ontology, not intrinsicly different | 12:23 |
| @fenn | it ends up being qualitatively different because of the extra complexity | 12:23 |
| -!- delinquentme [~delinquen@74.61.157.78] has joined ##hplusroadmap | 12:23 | |
| @fenn | like vector graphics vs pixels or something | 12:23 |
| @fenn | with tiny drops of liquid you don't have to worry about geometry or dynamics at all | 12:24 |
| kanzure | electronics is easier to reason about than biology stuff | 12:24 |
| kanzure | (pick-and-place, etc.) | 12:24 |
| @fenn | it's very similar i think | 12:24 |
| maaku | kanzure: electronics is too simple then. one of the things that I'd want to show is how pick-and-place is abstracted out of the language itself | 12:25 |
| @fenn | i've been fighting against it because i don't think it would be useful to have yet another oversimplified electronics/biology schema | 12:25 |
| maaku | transcriptic made this insight, and it is valid. pick-and-place is inferred by the execution environment, not explicit in the instructions | 12:26 |
| @fenn | implicit instructions are dangerous :( | 12:26 |
| @fenn | gah wtf http://istanbul.indymedia.org/sites/default/files/styles/colo/public/2014/01/befunky_null_3.jpg.jpg | 12:31 |
| maaku | fenn: hrm? what does it matter how a workpiece gets from position A in machine X to position B in machine Y? | 12:31 |
| @fenn | maybe you're thinking about pipetting things around when writing the recipe, but when compiled it goes through a microfluidic device and gets cross-contaminated or the sample is adsorbed or something | 12:35 |
| maaku | fenn: that would be an error in the compiler / execution environment | 12:35 |
| maaku | the compiler and/or execution interpreter should be smart enough to fix for limitations in its own equipment | 12:36 |
| @fenn | it's a slippery slope between "smart enough to solve its own limitations" and trying to do AGI | 12:36 |
| kanzure | but he is trying to do agi | 12:37 |
| kanzure | you know i'm actually proud of myself for not turning skdb into ai | 12:37 |
| maaku | :) | 12:37 |
| maaku | but in all seriousness in this particular case, i don't see how it's any different than the hoops VHDL compilers go through in making sure their circuits execute correctly | 12:38 |
| @fenn | by dangerous i meant "does things that you didn't expect" resulting in failure or equipment damage | 12:38 |
| maaku | and not having to encode explicit workpiece movement instructions (1) makes things easier to reason about (granted, iff you trust the machine), and (2) let's you do things like scale by interleaving protocols in an automated setup | 12:39 |
| kanzure | you can do kinematic planning and other motion planning, but only within reason (e.g. you have to know the capabilities of the machine) | 12:39 |
| maaku | kanzure: but the execution environment doesn't have to experience failure. you can check beforehand and say "oops, sorry, this won't work on your equipment" | 12:41 |
| -!- delinquentme [~delinquen@74.61.157.78] has quit [Ping timeout: 272 seconds] | 12:42 | |
| @fenn | ideally any recipe that has been entered into skdb package database will "just work" in reality when the artifact is created. it may be possible to have a rating system for each manufacturing process path permutation in order to work out the bugs via statistics | 12:42 |
| @fenn | if there are 2^999 possible manufacturing paths it becomes unfeasible to collect enough data to do that | 12:43 |
| @fenn | the rating system is something like "i did path XYZ and my part failed" or "i did path XZY and my part works" | 12:43 |
| @fenn | and more fine-grained testing | 12:44 |
| kanzure | the exact method of how something gets moved doesn't matter as long as that method happens to satisfy all of the requirements | 12:44 |
| kanzure | known requirements at least | 12:44 |
| @fenn | i think we are talking about different things. there are more parameters than just a series of points things get carried to | 12:45 |
| @fenn | in machining for example there is 1) how you hold the part in the machine, 2) cutter geometry with 8-10 parameters, 3) feeds and speeds, 4) lubricants and coolants, 5) machine geometry/materials, 6) machine wear and cutter wear | 12:47 |
| maaku | fenn: none of that would be inferred... | 12:48 |
| maaku | that's what the language would explicitly encode | 12:48 |
| kanzure | what was implicit then | 12:48 |
| maaku | for starters, just motion planning and resource allocation | 12:49 |
| maaku | maybe other stuff could be inferred to, but crawl before you walk | 12:49 |
| @fenn | i would like to infer which components are to be used from the available inventory, but it brings up a whole host of problems that can only be solved with huge amounts of simulations | 12:51 |
| kanzure | (he also wants a giant simulator though) | 12:51 |
| maaku | fenn: are components commodity items? | 12:52 |
| @fenn | yes | 12:52 |
| maaku | then what problems are introduced? | 12:52 |
| @fenn | "will it work?" | 12:52 |
| @fenn | will the bolts snap, will the part overheat, will it perform its function correctly | 12:52 |
| @fenn | defining the last one is hard | 12:53 |
| maaku | fenn: if you have steel block A and steel block B, what does it matter if you use A or B? | 12:54 |
| maaku | the code calls for "provision: 1 steel block" and the execution environment decides to use Block A for this purpose | 12:54 |
| @fenn | (setting aside the huge variety in steel alloys and their heat treatment) they are the same thing then, and not what i am talking about | 12:55 |
| maaku | regarding e.g. part overheat, you encode in the recipe a time delay or actual temperature requirement | 12:55 |
| @fenn | however if they are different dimensions then you have to do a simulation to make sure you can actually machine part X from block A and block B | 12:55 |
| maaku | fenn: right, i think we are talking past each other | 12:55 |
| maaku | fenn: yup, you need to type check your physical parts | 12:56 |
| maaku | although I think the process would be the reverse -- you extract from the recipe constraints on what size bars would be sufficient | 12:57 |
| @fenn | i don't think it's quite that simple | 12:57 |
| maaku | kanzure: it is important that the execution environment is not AGI-complete, although narrow AI search and planning involved | 12:57 |
| kanzure | if when you say "AI planning" you mean "planning" then yes | 12:58 |
| maaku | kanzure: in as much as anytime something in AI is successful it stops being called AI, yes | 12:58 |
| maaku | fenn: meh, this is standard practice in other fields, e.g. digital logic and high reliability software | 12:58 |
| @fenn | i figure it's not really AI if it doesn't involve learning | 12:58 |
| maaku | fenn: well this doesn't involve learning | 12:59 |
| @fenn | erk i didn't mean to bring up an AI discussion | 12:59 |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 13:00 | |
| maaku | well kanzure did, because yes it is true my long term hope is to have an AGI on top of this running experiments and learning | 13:00 |
| maaku | but I really believe that is orthogonal to this discussion | 13:00 |
| @fenn | well of course, but it shouldn't need AGI in the bottom levels of implementation | 13:01 |
| @fenn | otherwise nothing happens | 13:01 |
| kanzure | perhaps there should just be one way to do anything in the whole system. and if you don't have those parts or machines or components, then tough luck. it will be useless to almost everyone, but it would hvae a chance of actually working. | 13:01 |
| maaku | fenn: agreed | 13:02 |
| @fenn | we (humanity) have a huge amount of built up knowledge about what works and what doesn't work. all i want to do is encode this in a format that machines can use | 13:02 |
| maaku | kanzure: that's the starting point | 13:02 |
| maaku | kanzure: tha twould be sufficient. you could later build a planning engine on top that know how to evoke recipes to make the pieces it needs | 13:02 |
| punsieve | ... the system is usually setup with a list of possible components to be used for each step... then preference is put to the cheapest/most plentiful item. | 13:03 |
| maaku | again, not an AGI, just your typical planning constraint solver | 13:03 |
| kanzure | oh right, punsieve does manufacturing planning for a pulley manufacturer | 13:04 |
| punsieve | yes. Pulleys are awesome. | 13:04 |
| -!- erasmus [~esb@unaffiliated/erasmus] has quit [Ping timeout: 255 seconds] | 13:08 | |
| kanzure | punsieve: this is one of the things this sort of system would automate the fabrication and use of http://www.opentrons.com/ | 13:08 |
| kanzure | also, the problem with a list of possible components is that almost nobody wants to sit there writing down all possibilities (which can get very long) especially since not everyone has the same inventory etc | 13:09 |
| -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Quit: Leaving] | 13:09 | |
| punsieve | well, that's the only way I know how to do it... for machining I suppose instead of selecting specific metal blocks, you could have size-constraints. You'd still have to have your inventory accurately described. | 13:11 |
| kanzure | ugh inventory inaccuracies | 13:11 |
| punsieve | inventory accuracy is of primary importance in all manufacturing. If you don't have the piece you think you have ... you are screwed | 13:12 |
| -!- eudoxia [~eudoxia@r167-57-122-215.dialup.adsl.anteldata.net.uy] has quit [Quit: Leaving] | 13:12 | |
| punsieve | I'd say 70% of our emergency situations arise when someone failed to scrap out correctly. | 13:13 |
| * fenn has to go oxidize some carbohydrates... brb | 13:14 | |
| maaku | you should be able to workpiece characterization in many cases | 13:17 |
| kanzure | verb? | 13:17 |
| -!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 13:18 | |
| maaku | *do | 13:18 |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap | 13:18 | |
| -!- Beatzebub_ [~beatzebub@d75-155-236-222.bchsia.telus.net] has joined ##hplusroadmap | 13:25 | |
| -!- Beatzebub [~beatzebub@d75-155-236-222.bchsia.telus.net] has quit [Ping timeout: 276 seconds] | 13:27 | |
| -!- Beatzebub_ [~beatzebub@d75-155-236-222.bchsia.telus.net] has quit [Max SendQ exceeded] | 13:27 | |
| -!- Beatzebub_ [~beatzebub@d75-155-236-222.bchsia.telus.net] has joined ##hplusroadmap | 13:47 | |
| -!- erasmus [~esb@unaffiliated/erasmus] has quit [Quit: COQ] | 13:58 | |
| delinquentme | kanzure, | 14:08 |
| delinquentme | know what we have in common? | 14:08 |
| delinquentme | other than the level of our gyradoses ? | 14:09 |
| -!- Beatzebub_ [~beatzebub@d75-155-236-222.bchsia.telus.net] has quit [Ping timeout: 265 seconds] | 14:09 | |
| -!- nmz787_i [~ntmccork@192.55.54.40] has joined ##hplusroadmap | 14:24 | |
| -!- nmz787_i [~ntmccork@192.55.54.40] has quit [Read error: Connection reset by peer] | 14:25 | |
| -!- jdolan [~jdolan@208.83.72.113] has quit [Remote host closed the connection] | 14:25 | |
| -!- Madplatypus [uid19957@gateway/web/irccloud.com/x-unwubcwaskgxjrqr] has joined ##hplusroadmap | 14:27 | |
| kanzure | delinquentme: real ultimate power | 14:30 |
| -!- nmz787_i [ntmccork@nat/intel/x-trzshkuppaaybziw] has joined ##hplusroadmap | 14:35 | |
| -!- nmz787_i [ntmccork@nat/intel/x-trzshkuppaaybziw] has left ##hplusroadmap [] | 14:40 | |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has quit [Ping timeout: 264 seconds] | 14:42 | |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has joined ##hplusroadmap | 14:44 | |
| -!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 15:15 | |
| -!- Douhet [~Douhet@unaffiliated/douhet] has quit [Ping timeout: 245 seconds] | 15:22 | |
| -!- Douhet [~Douhet@unaffiliated/douhet] has joined ##hplusroadmap | 15:23 | |
| -!- strangewarp_ [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has joined ##hplusroadmap | 15:24 | |
| -!- night [~Adifex@unaffiliated/adifex] has quit [Ping timeout: 245 seconds] | 15:24 | |
| -!- night [~Adifex@unaffiliated/adifex] has joined ##hplusroadmap | 15:27 | |
| -!- poohbear [~poohbear@unaffiliated/tigger] has quit [Excess Flood] | 15:27 | |
| -!- strangewarp [~strangewa@c-76-25-206-3.hsd1.co.comcast.net] has quit [Ping timeout: 245 seconds] | 15:27 | |
| -!- poohbear [~poohbear@unaffiliated/tigger] has joined ##hplusroadmap | 15:28 | |
| -!- erasmus [~esb@unaffiliated/erasmus] has quit [Quit: wine] | 15:29 | |
| -!- delinquentme [~delinquen@c-50-161-66-31.hsd1.ca.comcast.net] has quit [Ping timeout: 245 seconds] | 15:38 | |
| -!- night [~Adifex@unaffiliated/adifex] has quit [Quit: Just going out for a swim...] | 15:39 | |
| -!- night [~Adifex@unaffiliated/adifex] has joined ##hplusroadmap | 15:44 | |
| -!- night [~Adifex@unaffiliated/adifex] has quit [Client Quit] | 15:46 | |
| -!- night [~Adifex@unaffiliated/adifex] has joined ##hplusroadmap | 15:50 | |
| -!- night [~Adifex@unaffiliated/adifex] has quit [Client Quit] | 15:51 | |
| -!- night [~Adifex@unaffiliated/adifex] has joined ##hplusroadmap | 15:53 | |
| -!- AmbulatoryCortex [~Ambulator@173-31-9-188.client.mchsi.com] has joined ##hplusroadmap | 15:56 | |
| -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-moxdpmpmawdtypjz] has joined ##hplusroadmap | 16:05 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 16:28 | |
| -!- delinquentme [~delinquen@c-73-202-243-243.hsd1.ca.comcast.net] has joined ##hplusroadmap | 16:35 | |
| -!- chris_99 [~chris_99@unaffiliated/chris-99/x-3062929] has quit [Remote host closed the connection] | 16:37 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has joined ##hplusroadmap | 16:48 | |
| kanzure | "Tools for downloading, converting and sharing online the data from a LiveScribe Pulse or Echo pen" https://github.com/CounterCultureLabs/dumpscribe | 16:56 |
| kanzure | juul: hi | 16:57 |
| -!- TheoryCat [user0000@unaffiliated/wye-naught/x-8734122] has joined ##hplusroadmap | 17:10 | |
| @fenn | hum i guess electronic notebooks are related to bio lab | 17:10 |
| -!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-shkwkxucporyehyq] has joined ##hplusroadmap | 17:10 | |
| -!- Acty [uid89656@gateway/web/irccloud.com/x-omsxdrokqwdijdib] has quit [Quit: Connection closed for inactivity] | 17:25 | |
| -!- delinquentme [~delinquen@c-73-202-243-243.hsd1.ca.comcast.net] has quit [Ping timeout: 264 seconds] | 17:33 | |
| kanzure | .title http://beacon-center.org/ | 17:34 |
| yoleaux | BEACON | Center for the Study of Evolution in Action | 17:34 |
| -!- delinquentme [~delinquen@c-73-202-243-243.hsd1.ca.comcast.net] has joined ##hplusroadmap | 17:35 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has quit [Remote host closed the connection] | 17:45 | |
| kanzure | http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR12.TRC2.A0.H0.Xhuman+skull.TRS0&_nkw=human+skull&_sacat=0 | 17:53 |
| kanzure | delinquentme: see link | 17:53 |
| delinquentme | damn dude. | 17:54 |
| delinquentme | expensive calcium. | 17:54 |
| kanzure | http://www.ebay.com/itm/New-Fine-Handmade-Bronze-Casting-1-1-Full-Life-Size-Gorilla-Skull-Adornment-/141669032508?pt=LH_DefaultDomain_0&hash=item20fc21e23c | 17:55 |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has joined ##hplusroadmap | 17:57 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has quit [Remote host closed the connection] | 17:58 | |
| -!- jdolan [~jdolan@c-50-188-243-111.hsd1.fl.comcast.net] has joined ##hplusroadmap | 17:59 | |
| kanzure | .title http://avida.devosoft.org/ | 17:59 |
| yoleaux | Avida by devosoft | 17:59 |
| kanzure | "Avida is a free, open source scientific software platform for conducting and analyzing experiments with self-replicating and evolving computer programs. It provides detailed control over experimental settings and protocols, a large array of measurement tools, and sophisticated methods to analyze and post-process experimental data." | 18:00 |
| kanzure | http://devolab.msu.edu/ | 18:00 |
| -!- erasmus [~esb@unaffiliated/erasmus] has joined ##hplusroadmap | 18:00 | |
| Adlai | .title http://devolab.msu.edu/ | 18:01 |
| yoleaux | Devolab | Digital Evolution Laboratory | Michigan State University | 18:01 |
| -!- Beatzebub [~beatzebub@d75-155-236-222.bchsia.telus.net] has joined ##hplusroadmap | 18:09 | |
| @fenn | i thought ebay explicitly forbid selling human remains of any kind | 18:12 |
| @fenn | also, what the hell? | 18:14 |
| @fenn | where do these come from? | 18:14 |
| kanzure | probably old retiring teachers | 18:17 |
| kanzure | er, they probably had the skulls as props, not that they are the skulls of the teachers themselves | 18:18 |
| @fenn | heh | 18:18 |
| -!- TheoryCat is now known as valleycat | 18:22 | |
| -!- valleycat is now known as TheoryCat | 18:22 | |
| maaku | Avida is awesome | 18:25 |
| kanzure | tim schmidt is working in that guy's lab now, apparently | 18:26 |
| delinquentme | austen is dead =[ | 18:47 |
| delinquentme | did you hear kanzure | 18:47 |
| delinquentme | ? | 18:47 |
| kanzure | austen hill? | 18:49 |
| kanzure | http://synbiobeta.com/remembering-austen-heinz/ | 18:51 |
| kanzure | http://igem.org/In_Memory_Of_Austen | 18:52 |
| kanzure | https://www.facebook.com/huffmantm/posts/10105111442071011 | 18:53 |
| -!- delinquentme [~delinquen@c-73-202-243-243.hsd1.ca.comcast.net] has quit [Ping timeout: 272 seconds] | 18:53 | |
| kanzure | i know this is rather morbid but the list of people who "liked" this post is like a cross-section of very interesting people | 18:53 |
| heath | http://blockchainworkshops.org/Agenda.pdf | 18:55 |
| kanzure | https://scholar.google.com/scholar?hl=en&q=lexical+fitness+testing&btnG=&as_sdt=1%2C44&as_sdtp= | 18:55 |
| kanzure | "Lexical fitness testing is when, instead of having one fitness function, you have a lexicon of fitness functions, and choose which one (or multiple) to use for each generation (or individual) semi-randomly. This results in remarkably flexible organisms." | 18:59 |
| kanzure | cc nsh | 19:13 |
| @fenn | woah how did i not know about this | 19:13 |
| kanzure | i'm sure there's variations of this like "distribution of scores across multiple fitness tests" | 19:14 |
| kanzure | oh you mean.... yeah. | 19:15 |
| @fenn | austen | 19:15 |
| kanzure | /win 8 | 19:19 |
| kanzure | jeroiqjerqeqjrq | 19:19 |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has quit [Ping timeout: 240 seconds] | 19:35 | |
| @fenn | .title http://youtu.be/VWMOSviTF4Y&t=1m15s | 19:39 |
| yoleaux | 404 Not Found | 19:39 |
| @fenn | hm | 19:39 |
| -!- PatrickRobotham [uid18270@gateway/web/irccloud.com/x-shkwkxucporyehyq] has quit [Quit: Connection closed for inactivity] | 19:40 | |
| kanzure | your url foo broked | 19:40 |
| @fenn | .title http://youtube.com/watch?v=VWMOSviTF4Y&t=1m15s | 19:40 |
| yoleaux | Austen Heinz, Part 2: DNA laser printing demo & exploration of the dark side - YouTube | 19:40 |
| kanzure | (you wanted ? not & in the first one) | 19:40 |
| @fenn | oh | 19:42 |
| @fenn | i always thought that syntax was dumb anyway | 19:42 |
| -!- wrldpc [~ben@182.72.135.27.ap.yournet.ne.jp] has quit [Quit: wrldpc] | 19:43 | |
| kanzure | you can also use an anchor #t=1m20s | 19:43 |
| kanzure | heh 18m 30s | 19:45 |
| kanzure | okay i didn't see anything else in there | 19:48 |
| kanzure | ah i guess https://www.youtube.com/watch?v=VWMOSviTF4Y&t=44m30s | 19:48 |
| @fenn | weird he is actually advocating a sort of "cloud conspiracy" for DNA synthesis, where it's impossible to actually order a custom test tube of DNA in the mail | 19:52 |
| -!- CheckDavid [uid14990@gateway/web/irccloud.com/x-moxdpmpmawdtypjz] has quit [Quit: Connection closed for inactivity] | 19:53 | |
| kanzure | this is why equipment should be local | 19:53 |
| @fenn | this interviewer is really bad | 19:58 |
| @fenn | kelp mining w00t | 19:59 |
| -!- punsieve [~drandomtu@173-170-176-33.res.bhn.net] has quit [Ping timeout: 255 seconds] | 19:59 | |
| -!- wrldpc [~ben@162.131.169.216.client.dyn.strong-sf94.as22781.net] has joined ##hplusroadmap | 20:15 | |
| -!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has quit [Quit: Leaving...] | 20:26 | |
| -!- punsieve [~drandomtu@173-170-176-33.res.bhn.net] has joined ##hplusroadmap | 20:32 | |
| -!- TheoryCat [user0000@unaffiliated/wye-naught/x-8734122] has quit [Quit: Goodbye] | 20:43 | |
| juul | kanzure: hallo :) | 20:58 |
| juul | I'm also | 20:58 |
| juul | I'm working on a beagle bone black sd card image (nearly done) and trying to get someone to make a 3d printable docking station for it | 21:00 |
| juul | I developed that stuff because the Real Vegan Cheese project was having an issue: The people at BioCurious were doing experiments, but people would come in after their 9-5 in the evening, run experiments in the wet lab and take notes with pen and paper, but then be too tired to write it up on the wiki after-wards. | 21:01 |
| juul | This often meant a 1-2 week lag between experiments happening and people at Counter Culture Labs being able to read about them | 21:02 |
| juul | I'm working on a raspberry-pi based, wide-angle, lab-table mountable, one-button webrtc video recorder/broadcaster | 21:05 |
| juul | motivated primarily by incidents where someone who is not super experienced in bio does an experiment, the experiment fails and no-one else was present so we have no clue what went wrong | 21:06 |
| juul | majority of video will probably be recorded and never looked at | 21:06 |
| juul | I'm using a modified ikea architect lamp (lamp-head replaced with raspi+cam) as the mount | 21:07 |
| @fenn | hey i had the same idea | 21:09 |
| -!- AmbulatoryCortex [~Ambulator@173-31-9-188.client.mchsi.com] has quit [Read error: Connection reset by peer] | 21:09 | |
| @fenn | except i'm using a playstation eye as the camera | 21:10 |
| juul | i'm using the raspicam because the raspberry pi has built-in 1080p 30fps H.264 encoding and you can get $25 fisheye cameras for it | 21:10 |
| juul | but the webrtc code to pull it off is really new and still buggy | 21:11 |
| juul | trying to find the bug that causes frame drops | 21:11 |
| @fenn | what is webrtc? | 21:11 |
| juul | webrtc is the new standard for low latency audio/video/data peer to peer connections between browsers | 21:12 |
| juul | built into firefox and chrome | 21:12 |
| juul | it works well in browsers | 21:12 |
| @fenn | hum ok | 21:12 |
| juul | but the server side code is still a bit new | 21:12 |
| juul | and you don't want to run a full web browser on a raspi just for the purposes of sending video | 21:13 |
| -!- punsieve [~drandomtu@173-170-176-33.res.bhn.net] has quit [Ping timeout: 272 seconds] | 21:16 | |
| @fenn | juul: so do you think you'll switch to the wifi version of the livescribe pen? so nobody has to manually sync it up | 21:18 |
| juul | fenn: i thought about it, but it might not be possible, since it's cloud based and it might verify the server side ssl certificate + i'd have to reverse-engineer another protocol + it's more expensive + you really want people to dock it anyway since it needs to have its battery charged up | 21:19 |
| juul | but if someone were to give me one i'd give it a try | 21:19 |
| @fenn | neat, "It syncs the audio to your writing so that when you go back and tap on a word you will hear the audio that transpired at that moment in time" | 21:21 |
| juul | yeah, I don't have support for that in dumpscribe yet though. I can only associate audio to a page, not a word. I'd also have to figure out how to link audio into a pdf, or convert the pdf to html and link the words. | 21:23 |
| juul | probably not suuuper hard to reverse-engineer | 21:23 |
| juul | but it seemed less important than other features | 21:23 |
| @fenn | yeah i don't think pdf should have audio in it | 21:23 |
| juul | yeah it seems dirty | 21:23 |
| @fenn | maybe a movie of the words being drawn and a subtitle track :P | 21:24 |
| juul | feature creeeeeep | 21:26 |
| @fenn | so there are other devices that work by various sensor technologies but don't require special paper. i think i am probably more likely to actually use these | 21:29 |
| @fenn | one kind uses an infrared camera that watches the pen, another uses resonant coil coupling like in a wacom tablet | 21:30 |
| juul | i haven't seen anything that doesn't require a third thing other han pen and paper | 21:38 |
| juul | which kinda kills the idea for me | 21:38 |
| -!- juri__ [~juri@pool-72-83-63-96.washdc.east.verizon.net] has joined ##hplusroadmap | 21:40 | |
| -!- juri_ [~juri@vpn166.sdf.org] has quit [Ping timeout: 258 seconds] | 21:41 | |
| -!- archels [charl@toad.stack.nl] has quit [Read error: Connection reset by peer] | 21:41 | |
| -!- archels [~neuralnet@toad.stack.nl] has joined ##hplusroadmap | 21:42 | |
| -!- sheena [~home@S0106c8be196316d1.ok.shawcable.net] has joined ##hplusroadmap | 21:44 | |
| -!- wrldpc [~ben@162.131.169.216.client.dyn.strong-sf94.as22781.net] has quit [Ping timeout: 255 seconds] | 21:47 | |
| -!- wrldpc [~ben@214.131.169.216.client.dyn.strong-sf94.as22781.net] has joined ##hplusroadmap | 21:48 | |
| juul | kanzure: any good ideas for where to spam our kickstarter? | 21:50 |
| juul | obvious first answer is "this channel" | 21:51 |
| juul | https://www.kickstarter.com/projects/1836537355/counter-culture-labs-your-biohacking-and-citizen-s | 21:51 |
| -!- neals [~lolwut@99-91-0-194.lightspeed.austtx.sbcglobal.net] has quit [Remote host closed the connection] | 22:14 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has quit [Ping timeout: 272 seconds] | 22:25 | |
| -!- Viper168 [~Viper@unaffiliated/viper168] has joined ##hplusroadmap | 22:26 | |
| sheena | http://gnusha.org/logs/2014-08-09.log ... logged and searchable? | 22:49 |
| -!- erasmus [~esb@unaffiliated/erasmus] has quit [Ping timeout: 255 seconds] | 23:13 | |
| -!- augur [~augur@c-73-46-94-9.hsd1.fl.comcast.net] has joined ##hplusroadmap | 23:24 | |
| -!- Acty [uid89656@gateway/web/irccloud.com/x-yhbjtythbpxvwsti] has joined ##hplusroadmap | 23:34 | |
| -!- zadock [~outsider@cthulhu.tuiasi.ro] has joined ##hplusroadmap | 23:49 | |
| --- Log closed Fri Jun 05 00:00:32 2015 | ||
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!
