| --- Day changed Wed Aug 27 2008 | ||
| ybit | from earlier... http://www.improve-eu.info/develop.html | 00:16 |
|---|---|---|
| kanzure | hurray for 8 am classes | 00:18 |
| * kanzure crashes | 00:19 | |
| willPow3r | i find that it's easier to just stay up all night than try to get up in time for an early class | 00:30 |
| fenn | i find that it's easier to drop out of school and be a bum.. | 00:36 |
| fenn | ybit: why would you use retinal implants when we already have RGB lasers and direct laser retinal displays? | 00:37 |
| ybit | good question, i dunno | 00:38 |
| ybit | i will in a few minutes though :) | 00:39 |
| fenn | for that matter, why don't we see any DLP-based HMD's? | 00:39 |
| fenn | DLP = high res MEMS mirror chips found in new projectors | 00:39 |
| bkero | you mean old projectors | 00:39 |
| bkero | I have one :P | 00:39 |
| fenn | send it on over :0 | 00:39 |
| fenn | i want to shine it in my eye! | 00:39 |
| bkero | Fuck off sir :P | 00:39 |
| bkero | You want 2200 lumens in your eye? | 00:40 |
| fenn | well, no, i was planning on substituting one of thoe RGB LED's | 00:40 |
| bkero | It uses halogens :P | 00:40 |
| bkero | or a halogen is hould say | 00:40 |
| fenn | color wheel is stupid.. someone give me a multinational corporation so i can make stuff that doesn't suck | 00:41 |
| bkero | fenn: I have bkero investments llc | 00:41 |
| fenn | i have fenn-wants-money.com | 00:42 |
| bkero | Based out of the hrp. We have offices in Oregon and a mailbox in the hrp | 00:42 |
| fenn | what is hrp? | 00:42 |
| bkero | Hood River Province | 00:42 |
| bkero | Sorry | 00:42 |
| bkero | Hutt | 00:42 |
| bkero | http://en.wikipedia.org/wiki/Hutt_River_Province_Principality | 00:43 |
| fenn | HRPP! | 00:43 |
| fenn | well, alrighty then | 00:43 |
| fenn | today, DLP HMD's, tomorrow the galaxy! | 00:44 |
| * fenn goes back to writing his personal inventory software | 00:45 | |
| fenn | someone i'm helping out in #emc.. i liked the name of his website: http://www.civilizationfromscratch.org/blog/ | 00:51 |
| kanzure | is this Alex? | 00:51 |
| kanzure | civilization-in-a-box? | 00:51 |
| fenn | tom owad | 00:51 |
| kanzure | cool | 00:52 |
| kanzure | a new guy | 00:52 |
| kanzure | abduct him | 00:52 |
| fenn | uh, wat do i say? | 00:52 |
| kanzure | civilization from scratch guys | 00:53 |
| bkero | Fuck I really need a subscription to make | 00:54 |
| fenn | make sucks | 00:54 |
| fenn | all they do is hot-glue pez dispensers together | 00:54 |
| bkero | Why? | 00:54 |
| bkero | lol | 00:54 |
| bkero | About once per issue they do some cool shit | 00:54 |
| fenn | seriously, why dont we see any 'how to build your own mass spectrometer' articles | 00:55 |
| bkero | They ran an article about building an electrophoresis and one about extracting your dna, selecting enes, and pcring them | 00:56 |
| bkero | *genes | 00:56 |
| fenn | meh | 00:56 |
| bkero | They give von slatt props and I have to respect that. | 00:56 |
| fenn | the steampunk guy? | 00:56 |
| kanzure | mass specs ? | 00:56 |
| kanzure | argh | 00:56 |
| kanzure | opera is down | 00:56 |
| kanzure | but I have bookmarks on that | 00:56 |
| bkero | I did use their article to make myself a wideband o2 sensor | 00:57 |
| fenn | make has a lot of good stuff, sure, but it just gets drowned out by the frilly 'me too' wannabe knitting projects | 00:57 |
| kanzure | on the other side of things, | 00:58 |
| bkero | You mean the 'diy community' | 00:58 |
| fenn | right, whatever that means | 00:58 |
| kanzure | it leaves little of a lasting community in the areas that the Maker Faire goes to | 00:58 |
| kanzure | there's no "diy community" left behind ;-) | 00:58 |
| bkero | I did go to the first maker faire | 01:00 |
| bkero | and I knew the people there converting a prius to full electric | 01:00 |
| bkero | kanzure: I'm part of a diy community :) | 01:00 |
| kanzure | It was funny today. I was going from table to table of the student organizations that are recruiting freshmen. Experimental rocketry groups, engineering + underdeveloped infrastructures in the third world, robotics, solar powered cars. So I walked up to each, talked with everyone, and at each one it seemed I had something to offer. | 01:00 |
| kanzure | To the third world country guys, I mentioned cheap diy biotech kits, to the rocketry guys I mentioned the SEDS conference and some CFD solvers, and with the library people I apologized and told them I'm the guy doing all of the automated interlibrary loan requests. One group already knew me ... it was odd. | 01:00 |
| kanzure | bkero: yeah, we have those groups too like http://therobotgroup.org/ and http://austinev.org/ or something | 01:00 |
| kanzure | bah, just check my http://austinbrains.org/ | 01:01 |
| bkero | heh austinev | 01:01 |
| bkero | I know a lot of cars from he austinev | 01:01 |
| bkero | Great site | 01:01 |
| kanzure | It's kind of weird talking about these groups irl because usually the other person I'm talking to is a /member/ of these groups | 01:02 |
| kanzure | and I start looking like an ass. | 01:02 |
| kanzure | without my knowing their membership status | 01:02 |
| kanzure | thus not displaying the proper amount of parsing of the information available about each group, you see | 01:02 |
| fenn | because you didnt memorize all the names on the membership list? | 01:06 |
| kanzure | more or less. | 01:06 |
| fenn | ah well, tell them their shoes look nice | 01:07 |
| fenn | or something | 01:07 |
| fenn | humans.. | 01:07 |
| * bkero isn't really a member of any clubs on campus | 01:08 | |
| bkero | I was sort of a member of the solar challenge team, but didn't show up because they didn't give me anything to do, and planning is bullshit. | 01:08 |
| kanzure | http://www.engr.utexas.edu/current/studentorgs/ <--- This is a terrible page. I need a better calendar. | 01:08 |
| kanzure | membership doesn't matter to me | 01:09 |
| kanzure | but I still do stuff | 01:09 |
| kanzure | most of the stuff that these organizations are trying to do | 01:09 |
| kanzure | i.e., "solar cells!" huh, I'm pretty sure I've dived into that fairly deeply on occassion | 01:09 |
| bkero | Most of these groups don't do shit | 01:10 |
| kanzure | Meh. | 01:10 |
| bkero | The 'environmental group' here ignores my ideas on solar cells and oil extraction from algae | 01:11 |
| kanzure | what do you mean by ignore | 01:11 |
| bkero | They just want to collect money to buy wind power from Pacific Power | 01:11 |
| kanzure | what do you do to tell them, are you doing a project and do they not come to see it ? | 01:11 |
| fenn | i had that problem with IU robo club.. nobody knew anything and on top of that they didnt even want to learn on their own what was possible | 01:11 |
| fenn | and that's like the entirety of engineering activity in bloomington | 01:11 |
| kanzure | I heard a few group members behind the tables saying similar things | 01:11 |
| bkero | Most robot clubs are a joke. :/ | 01:11 |
| kanzure | "there's really only one or two guys here that actually know what they're doing ..." | 01:12 |
| kanzure | bkero: The Austin robot clubs are fairly impressive. | 01:12 |
| kanzure | we even had a lab | 01:12 |
| kanzure | workshop thingy | 01:12 |
| fenn | kanzure: i think its true of every group of humans | 01:12 |
| kanzure | a thousand square feet | 01:12 |
| bkero | Nice | 01:12 |
| kanzure | until they all lost their jobs in the ~2002 engineering industry crashes | 01:12 |
| kanzure | and it was only one guy paying the bill .. | 01:12 |
| kanzure | :p | 01:12 |
| kanzure | poor Vern. | 01:13 |
| kanzure | he's been telling me that he's the only guy really doing anything for the group | 01:13 |
| fenn | 2002 engineering industry crash? what caused that? | 01:13 |
| * bkero is probably going to texas next spring to have dell suck my dick. | 01:13 | |
| kanzure | other members do stuff on occassion, but it's obvious that he's hyperoverloaded | 01:13 |
| kanzure | fenn: uhm, maybe 2000? the one where suddenly Mr. MBA+EE+CS finds himself flipping burgers | 01:13 |
| fenn | due to outsourcing or what? | 01:17 |
| kanzure | dunno, | 01:18 |
| kanzure | http://www.thinkandask.com/news/ibm.html IBM's 2003 round of layoffs | 01:18 |
| kanzure | http://www.ecommercetimes.com/perl/board/mboard.pl?board=ecttalkback&thread=5679&id=5681&display=1 50k layoffs @ IBM, per 2002 | 01:18 |
| bkero | IBM was also spiraling the shitter for a long time and eventually settled on selling off it's most prized possession--the Thinkpad line | 01:19 |
| kanzure | http://www.google.com/search?num=100&hl=en&q=semiconductor+layoffs+2000..2004&btnG=Search | 01:20 |
| kanzure | http://www.forbes.com/2004/05/19/cx_jp_2003layoffs_print.html layoff tracker | 01:20 |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has quit [Read error: 110 (Connection timed out)] | 02:23 | |
| kanzure | lynx question | 02:37 |
| kanzure | I've had a page up for a month | 02:37 |
| kanzure | I need to view the post/get data | 02:37 |
| kanzure | how? | 02:37 |
| bkero | Some firefox extensions will do that. | 02:39 |
| bkero | That's how I do it. | 02:39 |
| kanzure | doesn't help me.. | 02:40 |
| kanzure | I realized that I'm not paying for my internet connections | 02:40 |
| kanzure | so I'm retracing the forms I submitted to their server | 02:40 |
| kanzure | http://www.backyardmonorail.com/ | 03:13 |
| bkero | 07:15 <zarathustra_23> god damn that site is loading slower than shit | 03:16 |
| bkero | 07:15 <bkero> oh i thought it was me | 03:16 |
| bkero | 07:16 <zarathustra_23> I feel like I'm building the monorail, with him, right now. | 03:16 |
| bkero | 07:16 <zarathustra_23> each pixel at a time | 03:16 |
| kanzure | who's zarathustra? | 03:16 |
| bkero | Friend of mine | 03:17 |
| -!- elias` [n=me@unaffiliated/elias/x-342423] has quit [Read error: 110 (Connection timed out)] | 03:44 | |
| -!- elias` [n=me@unaffiliated/elias/x-342423] has joined #hplusroadmap | 03:45 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has joined #hplusroadmap | 03:53 | |
| faceface | Kanzure - is it normal to feel a hatred for "counselors" and "academic advisors" that don't actually do what they should be doing? | 04:06 |
| faceface | yes, but you need to learn how to negotiate people like any other obstacle | 04:06 |
| faceface | you can get hung up on idiots, hipocrits and bastards, but at the end of the day you need to just deal with them. | 04:06 |
| faceface | sad but true. | 04:06 |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 04:29 | |
| -!- Phreedom [n=freedom@ip-194-50-167-184.mir.dn.ua] has joined #hplusroadmap | 04:35 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has joined #hplusroadmap | 05:03 | |
| -!- Nade [n=lixasd@cpc2-clif5-0-0-cust516.nott.cable.ntl.com] has joined #hplusroadmap | 05:49 | |
| Nade | wow, i'm overwhelmed by the broadness of the group | 05:54 |
| Nade | is there anything you guys don't cover? | 05:55 |
| bkero | If there is--tell us. | 05:56 |
| bkero | (see todays xkcd) | 05:56 |
| bkero | http://www.xkcd.com/ | 05:56 |
| Nade | :) | 05:57 |
| Nade | so what exactly do most of you do here? | 05:57 |
| bkero | Talk | 05:59 |
| bkero | about SCIENCE! | 05:59 |
| Nade | :) | 05:59 |
| Nade | does anyone do any science? | 05:59 |
| Nade | i want to build an open source project to communicate with the brain | 06:01 |
| Nade | something that anyone can read about and can go out and build for just a few dollars | 06:01 |
| bkero | the inductive circuits for that can get expensive--$50 at least | 06:03 |
| bkero | Ping me tomorrow when I'm awake, good night | 06:03 |
| Nade | ok, night night | 06:04 |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 06:12 | |
| -!- willPow3r_ [n=will@cpe-66-75-6-181.san.res.rr.com] has joined #hplusroadmap | 06:14 | |
| -!- nsh [n=nsh@wikipedia/nsh] has quit [Read error: 110 (Connection timed out)] | 06:20 | |
| -!- willPow3r [n=will@cpe-66-75-6-181.san.res.rr.com] has quit [Read error: 110 (Connection timed out)] | 06:30 | |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has quit [Read error: 110 (Connection timed out)] | 06:46 | |
| -!- marainein [n=marainei@220.253-207-176.VIC.netspace.net.au] has quit ["That is not dead Which can eternal lie, And with strange eons Even death may die."] | 07:23 | |
| -!- nsh [n=nsh@eduroam-52.uta.fi] has joined #hplusroadmap | 07:44 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 07:52 | |
| -!- faceface [n=dbolser@bioinformatics.org] has quit [Remote closed the connection] | 08:16 | |
| kanzure | hm | 08:59 |
| kanzure | how late is it okay to show up for a class? | 08:59 |
| ybit | speaking of class... | 09:03 |
| -!- ybit [n=h@unaffiliated/ybit] has quit [Read error: 104 (Connection reset by peer)] | 09:03 | |
| -!- kanzure [i=bryan@66.112.232.134] has quit [Read error: 110 (Connection timed out)] | 09:17 | |
| nsh | people go to classes still? | 09:17 |
| nsh | i thought we were in the 21st century now | 09:17 |
| nsh | the classes come to *you* | 09:17 |
| -!- dgarr [n=dgarrett@66-81-72-169.bayarea.dialup.o1.com] has joined #hplusroadmap | 09:58 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has joined #hplusroadmap | 10:52 | |
| -!- dgarr [n=dgarrett@66-81-72-169.bayarea.dialup.o1.com] has left #hplusroadmap [] | 11:01 | |
| kanzure_ | nsh: apparently not | 11:13 |
| kanzure_ | Hey dgarr. | 11:14 |
| kanzure_ | Bay Area? | 11:14 |
| kanzure_ | D Garrett. Hm. | 11:14 |
| * nsh smiles | 11:14 | |
| kanzure_ | No, I don't think I met anyone there with dialup. | 11:14 |
| kanzure_ | There was a ridiculous lack of laptops. | 11:15 |
| -!- kanzure [i=bryan@66.112.232.134] has joined #hplusroadmap | 11:18 | |
| nsh | mm | 11:32 |
| procto | anyone here live real close to burlingame, ca, and wants to host me in october? | 11:51 |
| * procto is going to the seasteading conference | 11:51 | |
| fenn | this is dgarr's page, not sure why he ended up here: http://www.panix.com/~dgarrett/phpv1/index.php | 11:55 |
| kanzure | ooh | 11:56 |
| kanzure | I saw a cool shirt yesterday | 11:56 |
| kanzure | "Sustainability for security" | 11:56 |
| kanzure | procto, you listening? | 11:56 |
| procto | listening to what? | 11:56 |
| kanzure | the quote :p | 11:56 |
| kanzure | you do security work, no? | 11:56 |
| fenn | lol | 11:57 |
| fenn | information obscurity work | 11:57 |
| procto | some, yes | 11:57 |
| procto | ah | 11:57 |
| kanzure | whois chizu | 11:58 |
| kanzure | ohloh? so must be bkero's double | 11:58 |
| fenn | trevor hardcastle | 12:00 |
| procto | kanzure: I'm not quite sure what means. speficially, what type of sustainability and what type of security it's refering to | 12:00 |
| procto | kanzure: just fyi, my main work occupations right now have to do with information extraction and video processing :) | 12:00 |
| kanzure | procto: I have a video from a conference with very good information on slides that are a bit too blurry because of the mpeg encoder used, could you take a look at it for me? | 12:01 |
| kanzure | :p | 12:01 |
| kanzure | Steve tells me that there's some good algorithms out there | 12:01 |
| procto | kanzure: we're actually doing quite a different type of video processing system | 12:01 |
| kanzure | with keywords like 'super' and 'image' and 'extraction' | 12:01 |
| procto | (I do information extraction for my dayjob in the hedge fund industry, and I moonlight at a startup I co-founded, where we do video summarization) | 12:02 |
| procto | if you had a long lecture and needed only the interesting bits | 12:02 |
| procto | then I could help you | 12:02 |
| Nofaris | How should I kill the next 90 minutes of my life? | 12:02 |
| fenn | sounds useful for classes | 12:02 |
| procto | fenn: that's one potential use | 12:02 |
| procto | our audio-salience is so-so right now, so we're working on it | 12:03 |
| fenn | sounds like a Hard Problem | 12:03 |
| procto | spent all weekend hacking a webcam client | 12:03 |
| procto | several have been running since, seems it is working robustly | 12:03 |
| procto | webcams streams into the system, and the vidoe gets summarized on the fly | 12:04 |
| procto | video* | 12:04 |
| procto | I like to think of it as sousveillance on crack | 12:04 |
| fenn | is this like 'motion' where it just records when something is moving? | 12:04 |
| procto | not quite | 12:05 |
| nsh | WHUT PLS | 12:05 |
| procto | it's time-dilation proportional to various salience measures | 12:05 |
| nsh | summarise conversation subjects | 12:05 |
| procto | inversely proportional that is | 12:05 |
| * nsh is allergic to parsing buffers | 12:05 | |
| nsh | ____SUMMARY_____ | 12:05 |
| fenn | irc is a parsing buffer | 12:06 |
| fenn | that's why i use it | 12:06 |
| procto | nsh: in that case you might be interested in the topic extractions algorithms I write for work | 12:06 |
| procto | alas, they're still proprietary | 12:06 |
| * nsh suspects, in the words of the eminant 20th century philosophy Peter Griffin, that they'd be shallow, and pedantic | 12:07 | |
| procto | the algorithms? | 12:08 |
| nsh | yeah | 12:08 |
| procto | they are pretty shallow, but not pedantic | 12:08 |
| nsh | as in, only looking at superficial features and easily confused by minor changes in spelling that doesn't affect the sense | 12:08 |
| * nsh isn't convinced :-) | 12:08 | |
| procto | right. in fact, one of the main things I work is reducing its pedantic nature | 12:09 |
| nsh | s/doesn't/don't/ | 12:09 |
| nsh | procto, cool | 12:09 |
| procto | with fancy disambiguation and such | 12:09 |
| nsh | asplain pls? | 12:09 |
| procto | shallowness isn't a problem for our applications | 12:09 |
| nsh | why? | 12:09 |
| procto | because we're not doing free-form natural language processing | 12:10 |
| procto | but rather specific extractions | 12:10 |
| nsh | specific to what? | 12:10 |
| procto | we need to extract a limited set of topic from specific types of documents | 12:10 |
| procto | hedge funds | 12:10 |
| procto | I work for a company that makes software for hedge funds | 12:10 |
| nsh | oh, right; i remember | 12:10 |
| nsh | you're one of "them" | 12:10 |
| nsh | ;-) | 12:10 |
| procto | I fight the system from within :> | 12:11 |
| nsh | same with me and illegal drugs culture | 12:12 |
| * nsh smiles | 12:12 | |
| nsh | but i'm interested, what kinds of topics do your softwares extract from what types of documents? | 12:12 |
| nsh | SEC filings? | 12:12 |
| nsh | media articles mentioning companies and market-influencing phenomena? | 12:13 |
| nsh | like a form of economic traffic analysis, or what? | 12:13 |
| procto | some sec filings, internal emails, pretty much anything | 12:14 |
| procto | but we have a closed corpus | 12:14 |
| procto | i.e. companies and invdividuals relating to them | 12:14 |
| procto | and events that occur to those | 12:14 |
| procto | we make a research management system | 12:16 |
| procto | so as data pours in, it needs to be sorted and categorized as quickly and accurately as possible | 12:16 |
| nsh | interesting | 12:16 |
| nsh | and this indirectly affects the hedge-funds' market artivities | 12:16 |
| nsh | *activities | 12:16 |
| nsh | so by reverse engineering your algorithms, someone could feasably apply forcing functions to the hedge funds' interest, and by extension buy/sell probabilities | 12:17 |
| nsh | with enough resources, that might be an interesting exploit | 12:17 |
| procto | not quite | 12:18 |
| * nsh nods - the input documents would be very hard to manipulate | 12:19 | |
| procto | our product gets used by research-based funds | 12:19 |
| procto | the way the fund managers operate is that they make certain models in their heads | 12:19 |
| nsh | not sure i know what research-based fund means | 12:19 |
| procto | about what they expect to happen | 12:19 |
| procto | and they use research to verify and perhaps correct those models | 12:19 |
| procto | as well as develop new ones | 12:19 |
| nsh | ok | 12:20 |
| procto | in quant funds, fancy algorithms are used to decide on trading | 12:20 |
| procto | in research funds the algorithms are all in the heads of the managers | 12:20 |
| procto | who usually concentrate on specific asset targets | 12:20 |
| procto | rather than general trading strategies | 12:20 |
| nsh | but the topic extraction algorithms give differential weighting to certain matters | 12:20 |
| procto | which is what quants often do | 12:20 |
| procto | maybe if you had 100 years worth of our extraction data | 12:21 |
| procto | you could realize some info leak | 12:21 |
| procto | but there's very little skewing | 12:21 |
| procto | that's sorta the point of reliable disambiguation | 12:21 |
| nsh | ah ok, i see | 12:22 |
| nsh | so procto, you could make me a good algorithm for extracting statistically improbably phrases? | 12:23 |
| nsh | actually, to be fair, google makes that pretty trivial | 12:23 |
| procto | define statistically improbable? | 12:23 |
| procto | phrases that don't appear very often in a corpus? | 12:24 |
| nsh | like amazon uses | 12:24 |
| nsh | each book gets a list of phrases deemed important or specific to that book by virtue of appearing a higher than would be expected frequencies | 12:24 |
| procto | ah, gotcha | 12:25 |
| procto | there are plenty such algorithms which are quite efficient | 12:25 |
| procto | I wrote, for fun, a tool that analyzed the american presidential debates | 12:25 |
| nsh | i'd apply it to things like irc chats. for example, some years ago i thought it'd be fun to have google results float past behind the (translucent) irc window for things being discussed | 12:25 |
| procto | and then could tell you whether a particular segment from a future debate was from a republican or a democrat | 12:25 |
| nsh | so a wikipedia intro paragraph, picture result for something clearly pictorial, etc. | 12:26 |
| procto | it did this by analyzing collocations, that is, words that appear together more frequently than other words, then I trained a naive bayesian on the top 20 collocations of democrats, and top 20 collocations of republicans | 12:26 |
| nsh | interesting | 12:26 |
| procto | identification was quite reliable | 12:26 |
| procto | there are some "non sequitors" of course | 12:26 |
| procto | for example, republicans almost invariably mention ronald raegan | 12:26 |
| procto | hence the words "ronald" and "raegan" are a collocation | 12:27 |
| procto | while democrats almost never do | 12:27 |
| procto | and if they do, it's at most once or twice | 12:27 |
| nsh | what was the average period of collocation? once every N words: N = 20, 50, 100? | 12:27 |
| procto | don't remember | 12:27 |
| nsh | it would be interesting to give a novelty score to political speeches | 12:27 |
| nsh | that is, determine the degree of phrase reuse | 12:28 |
| nsh | then build up, using some hidden markov model or analogue, a citation tree | 12:28 |
| nsh | as in rhetoric a lot of stock phrases have quite tracable lineage | 12:28 |
| nsh | then one could determine rhetorical hereditry | 12:28 |
| nsh | *heredity | 12:29 |
| procto | one of my alpha users on the video summarization startup just did the thriller video, looks pretty nice summarized | 12:29 |
| procto | (we also finished direct-from-youtube summarization this weekend) | 12:29 |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 12:29 | |
| nsh | i can't even imagine how you'd do that... | 12:29 |
| nsh | are you using feature recognition? | 12:29 |
| procto | nope | 12:30 |
| nsh | asplain! | 12:30 |
| procto | I can't. our algorithms are currently secret. | 12:30 |
| procto | that may change later. | 12:30 |
| nsh | meh at your secrets | 12:30 |
| nsh | you can't give me an overview even? | 12:30 |
| procto | but in general, we are doing time-dilation that's inversly proportional to salience | 12:31 |
| procto | and we have multiple salience measures | 12:31 |
| procto | one prominent salience measure is motion | 12:31 |
| nsh | oh, this is what you were talking about earlier | 12:31 |
| procto | more motion, time moves slower | 12:31 |
| procto | yes | 12:31 |
| procto | less motion, time moves faster | 12:31 |
| nsh | ok, that | 12:32 |
| nsh | 's a good idea | 12:32 |
| nsh | but that just assigns a rapidity index to the video | 12:32 |
| procto | it's somewhat fancier than that | 12:32 |
| nsh | i assume summarisation would be verbal to some degree... | 12:32 |
| procto | but that's the general idea | 12:32 |
| nsh | now, perhaps in conjunction with tagging, | 12:33 |
| procto | video summarization meaning that you take a video of X minutes, and you get a video of Y minutes, where Y < X | 12:33 |
| nsh | for example, people can add events and they'll stick to certain portions of the video | 12:33 |
| procto | sorry if that wasn't clear | 12:33 |
| nsh | oh | 12:33 |
| nsh | automated editing then | 12:33 |
| procto | right | 12:33 |
| nsh | sorry, i was thinking verbal synopsis | 12:33 |
| nsh | that's why i was liek: JOHN TITOR, IS TAHT U?!?! | 12:33 |
| procto | understandable | 12:33 |
| * nsh smiles | 12:34 | |
| procto | for example, if you're familiar with justin.tv and ustream.tv | 12:34 |
| procto | there have been multiple cases now where folks got robbed | 12:34 |
| nsh | hmm | 12:34 |
| procto | then they went unto the sites, and looked through their own webcam archives | 12:34 |
| nsh | (not familiar with either) | 12:34 |
| nsh | i se | 12:34 |
| procto | the blog posts about these cases always include phrases like "after wading through hundreds of hours of footage" or "4 hours later I found the culprit" | 12:34 |
| nsh | right right | 12:34 |
| nsh | so you want to ultimately provide a way to make video fingerprints | 12:35 |
| procto | not just | 12:35 |
| procto | that's just one usecase | 12:35 |
| nsh | sure | 12:35 |
| nsh | but a good one, nonetheless | 12:35 |
| procto | where most of the time, nothing happens in your room | 12:35 |
| procto | but when the robber's in, that's the main part you'll see | 12:35 |
| procto | it's like it fast forwards to the good parts for you | 12:35 |
| nsh | ok, but that's largely an extant technology | 12:35 |
| nsh | though maybe not very mature yet | 12:36 |
| nsh | i haven't ever checked | 12:36 |
| procto | there are some existing implementations of similar things | 12:36 |
| procto | none of them are quite doing what we are (we checked) | 12:36 |
| * nsh nods | 12:36 | |
| procto | and the ones that exist, are almost all in the high-end security business | 12:36 |
| nsh | meh :-) | 12:36 |
| procto | like the israeli company briefcam | 12:36 |
| procto | http://www.briefcam.com/demo.html | 12:37 |
| procto | check those out | 12:37 |
| * nsh on it | 12:37 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has joined #hplusroadmap | 12:37 | |
| nsh | what are they doing with the colours? or is that natural? | 12:37 |
| procto | they do object tracking | 12:38 |
| procto | so they track all the objects | 12:38 |
| nsh | ok, but they're nonlinearising the shot | 12:38 |
| procto | and then they have all the objects "play" at the same time, but also resolve collisions | 12:38 |
| procto | right | 12:38 |
| procto | they are | 12:38 |
| nsh | pretty novel | 12:38 |
| nsh | i like | 12:38 |
| nsh | but you'd have to then include a way to unwrap it in real time | 12:38 |
| procto | so that's not quite what we're doing, but it's also quite high end, and requires substantially more processing than we do | 12:38 |
| nsh | which doesn't seem too hard, but perhaps difficult to UI well | 12:38 |
| * nsh nods | 12:39 | |
| procto | since we're going for a consumer market | 12:39 |
| procto | our design is based on that | 12:39 |
| procto | if you're into, say, golf | 12:39 |
| procto | imagine watching the summary, but when you catch a part you want to watch fully realtime, well, just clik the button, and it will play realtime from that exact point | 12:39 |
| kanzure | http://revminds.seedmagazine.com/revminds/member/aleksandra_m_walczak/ | 12:39 |
| kanzure | podcast on noise in genetic regulatory networks | 12:40 |
| nsh | procto, right | 12:40 |
| nsh | now... | 12:40 |
| * nsh thinks about interesting things regarding non-linear video compression and causality regression | 12:40 | |
| procto | I'll give out invites once we're in alpha2 | 12:41 |
| procto | (waiting on the bangladeshi designers we've hired to finish working) | 12:41 |
| * nsh smiles | 12:42 | |
| nsh | kanzure, don't forget to bookmark | 12:42 |
| procto | woo globalization | 12:42 |
| nsh | meh @ that kanzure | 12:44 |
| nsh | i hate that people are revolutionary for doing stuff that should have been obvious decades ago | 12:45 |
| kanzure | no argument there | 12:45 |
| procto | my business partner on the video summarization startup is a biophysicist | 12:48 |
| procto | works for the broad institute | 12:48 |
| nsh | her publications look less sophomoric than the explanation for revminds | 12:48 |
| nsh | procto, interesting | 12:48 |
| procto | usually people are like "I work doing web development as a day job, but what I'd really like to do is cure cancer!" | 12:48 |
| * nsh likes multidisciplinarians | 12:48 | |
| procto | he is "I work curing cancer, but what I'd really like to do is some web development!" | 12:48 |
| * nsh doesn't like the term web development | 12:49 | |
| nsh | it usually means: do something a million people have done before, and suboptimally, and to no general benefit of humanity | 12:49 |
| procto | I rather agree | 12:50 |
| procto | I just think "writing a web app" sounds even more cumbursome | 12:50 |
| procto | or "an app that people access on the web" | 12:50 |
| * nsh smiles | 12:50 | |
| procto | which better describes it | 12:50 |
| procto | because it isn't a *web* app | 12:50 |
| nsh | i'd just call it "dayjob" | 12:50 |
| nsh | there's dayjob, and there's work | 12:51 |
| procto | it just so happens I'm giving an HTTP, HTML interface to it... | 12:51 |
| * nsh nods | 12:51 | |
| procto | well, I enjoy both my occupations | 12:51 |
| procto | professional occupations that is | 12:51 |
| procto | obviously I'm actually occupied by much more | 12:51 |
| * nsh is occupied by 3-4 spirits on a daily basis | 12:51 | |
| kanzure | holy shit, what's wrong with you people? http://heybryan.org/buildingbrains.html ' (more axons than dendrites).' | 12:53 |
| nsh | whatnow? | 12:54 |
| kanzure | there's more dendrites than axons on a neuron | 12:54 |
| kanzure | you fools | 12:54 |
| kanzure | although there's one paper in Blackwell Synergy saying 'When stained for axons and dendrites, some interneurons in the hippocampus are found to have more axons than dendrites, unlike usual neurons. ...' | 12:54 |
| * kanzure fixes | 12:55 | |
| procto | kanzure: didn't you write the page? I'm not sure who is "you people" | 12:57 |
| procto | you're saying that people should have noticed? | 12:57 |
| nsh | apparantly, we're proofreaders :-) | 12:57 |
| procto | I see... | 12:57 |
| procto | I've never read the page, but I probably would've remarked on it if I had | 12:57 |
| nsh | though that might have been pasted from somewhere | 12:57 |
| kanzure | if I'm wrong I expect people to yell at me | 12:57 |
| kanzure | and throw stuff | 12:58 |
| kanzure | like rocks | 12:58 |
| * nsh has a good collection of rocks | 12:58 | |
| kanzure | smooth rocks? | 12:58 |
| * procto spent a year in highschool cutting up brains | 12:58 | |
| procto | lab was during lunch period | 12:58 |
| nsh | anyway, gonna head home | 12:58 |
| -!- nsh [n=nsh@eduroam-52.uta.fi] has quit ["bai"] | 12:58 | |
| procto | so now the mere sight of brains makes me salivate | 12:58 |
| kanzure | I've done a ridiculous number of dissections, though I think I've only dissected one brain before | 12:58 |
| procto | I've only ever dissected brains | 12:58 |
| kanzure | you're my hero. | 12:58 |
| procto | I think I've done at least a hundred brains | 12:59 |
| Nade | i haven't dissected a single thing, I feel like a weirdo | 12:59 |
| procto | mostly sheep | 12:59 |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 14:11 | |
| -!- elias` [n=me@unaffiliated/elias/x-342423] has quit [Read error: 145 (Connection timed out)] | 14:20 | |
| -!- elias` [n=me@unaffiliated/elias/x-342423] has joined #hplusroadmap | 14:32 | |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 14:34 | |
| -!- Nade [n=lixasd@cpc2-clif5-0-0-cust516.nott.cable.ntl.com] has quit [] | 14:52 | |
| kanzure | "we'll be learning you how to convert yer inches to moles" | 15:09 |
| kanzure | why am I in a chem 301 class? | 15:09 |
| kanzure | procto: re: Caspi. Is the Weizmann Institute of some importance? I've been noticing it from particularly good papers and projects. | 15:22 |
| procto | yes, it is | 15:23 |
| procto | it's the premier theoretical science institution in Israel | 15:23 |
| kanzure | hm | 15:23 |
| procto | and Israel does have the highest concentration of post-secondary degrees per capita in the world | 15:23 |
| procto | (at least partially because of the large influx of ex-soviet scientists. my father was a physicist in the soviet union, for example) | 15:24 |
| kanzure | it appears most of my time now is checking the calendars and schedules to make sure I can make various events around campus or, more accurately, trying to remember long-term projections for possible stuff to go to | 15:26 |
| kanzure | that's it .. time for me to get coding | 15:26 |
| kanzure | procto: looks like Caspi was doing image segmentation stuff. Know anything other than pset for page segmentation algorithms and toolkits? Preferably something that is already software. | 15:26 |
| procto | I don't | 15:28 |
| nsh | kanzure, code a personal assistant | 15:36 |
| nsh | if i ever had a schedule, i sure as hell wouldn't spend time thinking about it :-) | 15:37 |
| kanzure | right | 15:40 |
| procto | asksunday.com | 15:40 |
| procto | if I lived in NYC, I would be an asksunday user | 15:43 |
| kanzure | looks stupid | 15:43 |
| kanzure | 20 minutes | 15:43 |
| kanzure | I'll give you a task that I know, by theoretical computer science, will take you 40 | 15:43 |
| kanzure | thus your service sucks | 15:43 |
| kanzure | etc. | 15:43 |
| procto | huh? | 15:44 |
| procto | I'm not sure that's the point... | 15:44 |
| procto | their messgenger/time service would be the most useful thing for me | 15:45 |
| kanzure | http://heybryan.org/music2/radio/sky.fm%20-%20the80s.pls | 15:51 |
| kanzure | http://www.perlmonks.org/?node_id=620076 perl+yaml stuff, since I've been doing more work with perl lately | 15:51 |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has quit [Read error: 104 (Connection reset by peer)] | 16:32 | |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 16:32 | |
| kanzure | what's the best way to do factorial checks? | 16:39 |
| kanzure | should each object keep track of who it has been checked with? | 16:40 |
| kanzure | guess that's not really the right way to be doing it anyway | 16:40 |
| kanzure | since the goal is to recursively generate all possible final schedules | 16:40 |
| kanzure | thereby not necessitating any management of the checks ? | 16:40 |
| kanzure | I think conflict-check-management would still be needed | 16:40 |
| -!- ybit [n=h@unaffiliated/ybit] has joined #hplusroadmap | 17:12 | |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 17:22 | |
| -!- nsh [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has quit [Connection timed out] | 17:31 | |
| -!- Nade [n=lixasd@cpc2-clif5-0-0-cust516.nott.cable.ntl.com] has joined #hplusroadmap | 18:20 | |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has joined #hplusroadmap | 18:32 | |
| kanzure | Hey Nade. | 19:23 |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: elias` | 19:34 | |
| -!- Netsplit over, joins: elias` | 19:35 | |
| ybit | hmm, definitely going to stick with contacts | 19:52 |
| ybit | glasses are okay though, but they make me dizzy | 19:53 |
| ybit | i should stick to the topic | 19:53 |
| * ybit shuts up | 19:53 | |
| kanzure | I lose track of my fork stack | 19:58 |
| -!- marainein [n=marainei@220.253-207-176.VIC.netspace.net.au] has joined #hplusroadmap | 20:09 | |
| bkero | forks? | 20:19 |
| bkero | http://www.theonion.com/content/radio_news/new_flavored_fork_adds_taste | 20:19 |
| kanzure | hm, it wouldn't be impossible to make flavored forks | 20:20 |
| kanzure | inner capillary system maybe | 20:20 |
| kanzure | and recharge it over night at a dispenser | 20:20 |
| kanzure | hm | 20:22 |
| kanzure | there's 60 people on facebook in the biotech toolkit group | 20:22 |
| kanzure | didn't notice .. | 20:22 |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: elias` | 20:22 | |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: ybit | 20:22 | |
| -!- Netsplit over, joins: ybit | 20:23 | |
| bkero | It would be rather easy to make flavored forks | 20:24 |
| kanzure | I wonder if you can convince somebody that a texture is flavor | 20:25 |
| kanzure | "metallic flavor!" | 20:25 |
| kanzure | comes in plastic too | 20:25 |
| -!- elias` [n=me@unaffiliated/elias/x-342423] has joined #hplusroadmap | 20:26 | |
| bkero | lol | 20:26 |
| bkero | You could just put something in the plastic mold | 20:26 |
| bkero | like bbq sauce | 20:26 |
| bkero | mm bbq sauce | 20:26 |
| Nofaris | BBQ flavored anythingyoufuckingeat | 20:27 |
| Nofaris | awesome | 20:27 |
| bkero | 'eh, I'll steer clear of bbq flavored ice cream | 20:28 |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: elias` | 20:36 | |
| -!- Netsplit over, joins: elias` | 20:42 | |
| kanzure | http://blog.wired.com/wiredscience/2007/12/macgyver-scienc.html Hack: young prof makes lab-on-a-chip with shrinky dink and toaster oven | 20:44 |
| kanzure | well | 20:45 |
| kanzure | there we go. | 20:45 |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: elias` | 20:47 | |
| -!- biopunk [n=p@h246n1c1o261.bredband.skanova.com] has joined #hplusroadmap | 20:50 | |
| -!- Netsplit over, joins: elias` | 20:55 | |
| kanzure | http://en.wikipedia.org/wiki/Venus'_Flower_Basket <-- grows glass? | 20:56 |
| Nade | hi | 21:05 |
| -!- Netsplit niven.freenode.net <-> irc.freenode.net quits: elias` | 21:05 | |
| Nade | that shrinkydink thing's quite interesting | 21:06 |
| Nade | hi kanzure | 21:07 |
| -!- Netsplit over, joins: elias` | 21:07 | |
| kanzure | Nade: You were ranting in ##neuroscience yesterday about brain augmentation, implants, our open rTMS project, and a few other things -- you can carry on now :-) | 21:08 |
| Nade | :) | 21:09 |
| Nade | well i was curious about the current state of diy brain interfaces | 21:09 |
| kanzure | http://heybryan.org/2008-08-15.html contains a few links to some papers on my server re: how to make electrodes by hand. | 21:10 |
| kanzure | You could send a PCB board design off to one of the pcbfarms, or use your own pcb inkjet printer hack, and then wire up the electrodes to a PIC/microcontroller. But the heat of these things might kill your brain cells. You might want to look into this. To manufacture your own custom chips it'll be about ~$10k IIRC. | 21:11 |
| Nade | to be more specific, I wondered if there were any diy version of those eeg headset thingys that are starting to emerge commercially | 21:11 |
| bkero | Yea | 21:12 |
| bkero | There's one coming out aimed for gamers | 21:12 |
| bkero | I forget what it's called | 21:12 |
| Nade | ultimately I would want to end up with a simple set of instructions that anyone with an arduino and a couple of dollars could put together | 21:12 |
| -!- Nofaris [n=Nofaris@adsl-75-42-94-197.dsl.scrm01.sbcglobal.net] has quit [Read error: 110 (Connection timed out)] | 21:12 | |
| Nade | then focus on the software to interpret the data | 21:12 |
| Nade | similarly, I want to see if one could communicate language through a series of vibrating motors placed around the body to represent all of the letters of the alphabet | 21:13 |
| kanzure | bkero - you want to take this one? | 21:14 |
| kanzure | this is kind of your project that you were mentioning | 21:14 |
| bkero | Yea sure. | 21:15 |
| Nade | project? :) | 21:15 |
| * kanzure is not a big fan of eeg .. too many people (DoYouKnow in ##neuroscience is an excellent example, not that I'm calling out names or anything) think that EEG is the answer to brain interfacing ... look, you're just getting less than 1 cm deep and just some vague electrical readings, nothing localized really. But there is some neurofeedback that can be done with it, yes, of course. | 21:16 | |
| bkero | I was looking at the openeeg project bcause that's where the defcon talk I saw linked me to, but I found some EEG-style systems to monitor Mu-rhythms in the brain. | 21:16 |
| bkero | kanzure: With 2 electrodes on each side of the cortex, and a body reference, signals become pretty clear. | 21:17 |
| bkero | Invasive brain reading scares the shit out of most people. | 21:17 |
| bkero | Mainly because fucking up means you're dead/a vegetable. | 21:17 |
| bkero | Keep in mind these are the same people soldering wires to pennies and using those as electrodes with shampoo as a contact gel. | 21:18 |
| bkero | You can get as ghetto as you want with this kind of thing. :) | 21:18 |
| kanzure | You have a reference on that one ? | 21:18 |
| kanzure | I've been wondering what to do with these damn pennies. | 21:18 |
| bkero | Hahaha | 21:18 |
| Nade | lol | 21:19 |
| bkero | http://www.eng.utah.edu/~jnguyen/ecg/instructions.html | 21:19 |
| bkero | I guess those were used for an ekg | 21:19 |
| Nade | that's the kind of thing i was thinking about | 21:20 |
| Nade | i'm sure invasive gets much better results, but it's not something i'd want to do without a lot of research and planning | 21:21 |
| Nade | and i wouldn't stick anything in my head | 21:21 |
| Nade | although i'd have a kevin warwick style implant | 21:21 |
| kanzure | "Captain Cyborg" | 21:22 |
| Nade | i've had lectures with him | 21:22 |
| bkero | Most of the implants involve coming through the back of your head where there's a breach in the skull, drilling, or coming through the temple. ;) | 21:22 |
| Nade | a bit of a nutter | 21:22 |
| bkero | By implants I mean subdermal electrodes. | 21:22 |
| kanzure | Screw wires. | 21:23 |
| kanzure | Just use wireless communication and power transmission. | 21:23 |
| kanzure | http://heybryan.org/docs/neuro/ has the papers on how to do this. | 21:23 |
| bkero | kanzure: There was a talk at defcon about security vulnerabilities in pacemakers. | 21:23 |
| bkero | You can overload them, run them out of batteries, or even disable some routines and get them to induce a heart attack(test mode) | 21:23 |
| bkero | Imagine having something insecure tied to your central nervous system. | 21:24 |
| kanzure | Like a woman? | 21:24 |
| kanzure | har har har, a joke. A joke. | 21:24 |
| Nade | lets be honest though, it's not hard to kill old people | 21:24 |
| kanzure | I'm not arguing against noninvasive interfacing. | 21:24 |
| bkero | kanzure: None of us have ever had a women tied to us. We're on IRC for gods sake. | 21:25 |
| bkero | Nade: how about teenagers with a gimped heart? :P | 21:25 |
| Nade | you got me there | 21:25 |
| Nade | :p | 21:25 |
| Nade | so what exactly is your project then bkero | 21:26 |
| bkero | My project? | 21:26 |
| bkero | Reading Mu rhythms for cursor control | 21:26 |
| Nade | using diy equipment? | 21:26 |
| bkero | Yes | 21:26 |
| bkero | I'm trying to be so lazy I don't hav eto move my arms or speak to use a computer. | 21:27 |
| Nade | a noble goal | 21:27 |
| kanzure | don't forget the subvocalization web browser paper | 21:27 |
| kanzure | Nade, we're all very lazy people | 21:27 |
| bkero | kanzure: subvocalization equipment is expensive. | 21:27 |
| Nade | so have you got it working? | 21:27 |
| kanzure | supposedly so are electrodes, but you just showed me a really cheap two bit, no, less than two bit way of doing it | 21:27 |
| bkero | Although did you see that jaw vibration speaker and subvocal mic that some french guy got implanted to use as a cell phone? | 21:27 |
| bkero | kanzure: electrodes are about $40 from canada :) | 21:28 |
| bkero | I don't have the link up anymore, but it's about 3rd degree from the openeeg site--so good luck. :P | 21:28 |
| Nade | has anyone tried using haptics to communicate language? for example using several wearable feedback devices (e.g a vibrating motor or small electric shock) to represent the letters of the alphabet? | 21:30 |
| bkero | You mean like Hawkings mouthpiece? | 21:31 |
| Nade | nope | 21:32 |
| Nade | having a computer read you a message by activating several feedback devices at different times for each letter | 21:33 |
| Nade | so a small vibration on your upper arm and shoulder might represent the letters b and c etc... | 21:34 |
| bkero | So what's stopping you from using a little microcontroller and 26 little motors and some tape? | 21:35 |
| Nade | i don't have one yet :) | 21:35 |
| Nade | but that's what i want to try | 21:35 |
| Nade | do you think I'd be able to learn how to interpret the messages with enough practice? | 21:36 |
| kanzure | Interpretive dance? | 21:37 |
| bkero | Sure | 21:38 |
| bkero | You can remember 26 positions on your body | 21:38 |
| nsh- | semaphor is a slow code :-) | 21:39 |
| nsh- | imagine is sign language speakers only used the alphabet | 21:39 |
| nsh- | *if | 21:40 |
| nsh- | i suspect it'd be more efficient to learn a pidgin | 21:40 |
| Nade | i think if you practiced you could get it up to a reasonable speed | 21:40 |
| kanzure | Simultaneous movements -- can you pat your stomach and rub your, wait | 21:40 |
| kanzure | shit | 21:40 |
| kanzure | rub your stomach and pat your head :) | 21:40 |
| kanzure | Gah. Can't even do it in text. | 21:40 |
| * nsh- smiles | 21:41 | |
| nsh- | anyone read that Asimov short story, "Spell My Name With An S"? | 21:43 |
| Nade | nope | 21:44 |
| kanzure | Vomiza | 21:44 |
| kanzure | rawr | 21:45 |
| nsh- | should read it | 21:45 |
| kanzure | fenn: you have it? | 21:46 |
| Nade | hmmn, i wonder what it'd be like to constantly be aware of your heartbeat.. like if you connected it to a speaker or vibrating motor | 21:47 |
| kanzure | There's somebody who posted on Slashdot once about a completely silent chamber. Nobody lasts more than 30 minutes in the thing. You hear everything your body is doing. | 21:48 |
| bkero | Nade: I spend half the way with my fingers on my neck or wrist for my pulse. | 21:49 |
| bkero | *way = day | 21:49 |
| Nade | why's that? | 21:50 |
| nsh- | people do quite well in sensory deprivation tanks... | 21:51 |
| kanzure | I have my hand pressed to my face a good portion of the time too | 21:51 |
| kanzure | hm | 21:51 |
| nsh- | u has gmail | 21:51 |
| kanzure | kk | 21:51 |
| kanzure | yep | 21:51 |
| Nade | yeah but i reckon it might influence you if you could hear it all the time | 21:52 |
| kanzure | reckon? really? | 21:53 |
| Nade | slightly | 21:53 |
| Nade | if you're trying to get fit it might be useful | 21:53 |
| Nade | maybe not | 21:54 |
| kanzure | No, I mean the word choice. | 21:54 |
| nsh- | you'd stop hearing it | 21:54 |
| nsh- | it would take zen training to not automatically filter it | 21:54 |
| * kanzure was thinking of checking out the local zen center | 21:55 | |
| Nade | maybe you could control it :o | 21:55 |
| * nsh- doesn't hear the annoying noise that comes out of his bathroom constantly | 21:55 | |
| nsh- | except now of course, that i'm thinking about it | 21:55 |
| bkero | grunting? | 21:55 |
| kanzure | nsh-: the noise saying "help, I need toilet paper!" | 21:55 |
| Nade | lol | 21:56 |
| kanzure | this is why they invented the fan | 21:56 |
| kanzure | always use the fan, nobody wants to hear your liquids | 21:56 |
| * nsh- smiels | 21:56 | |
| bkero | Unless you have a lewd girlfriend. :P | 21:57 |
| kanzure | bkero: IRC, remember? | 21:57 |
| bkero | ...lewd mother? | 21:57 |
| Nade | another idea would be to make the user constantly aware of the time of day, by vibrating a tiny motor on their arm every minute or every ten mins | 21:57 |
| bkero | Your roommate has a lewd girlfriend? | 21:57 |
| kanzure | bkero: when irc guys can't get a date, I think this includes their mother | 21:57 |
| kanzure | bkero: irc peeps don't necessarily have roomies | 21:57 |
| bkero | kanzure: bullshit, you're in a dorm | 21:57 |
| kanzure | I don't have a room mate | 21:57 |
| bkero | o | 21:57 |
| bkero | Yea i did that too | 21:57 |
| bkero | fuck dormmates | 21:58 |
| kanzure | Nade: I don't know if that would tell you much. | 21:58 |
| kanzure | What are you trying to do ? | 21:58 |
| kanzure | It seems you're kind of ad hoc :) | 21:58 |
| bkero | Nade: You could always use small electrical pulses to contract muscles instead of motors. | 21:58 |
| Nade | i'm throwing a few ideas in the air :) | 21:58 |
| Nade | i'm curious, because i think simple things like that might change the users behavior | 21:58 |
| Nade | it might alter how long they spend on certain tasks or what time they do them | 21:59 |
| kanzure | so you're interested in task scheduling and so on? | 21:59 |
| Nade | or it might just annoy them | 21:59 |
| * kanzure is an extreme multitasker | 21:59 | |
| kanzure | I think I'm presently doing a few thousand things. | 21:59 |
| kanzure | I've lost count. | 21:59 |
| kanzure | but they're being "done", mysteriously | 21:59 |
| * bkero has his computers do things for him. | 21:59 | |
| fenn | hmm. asimov wrote too much, i can't find it | 22:01 |
| kanzure | fenn: nsh got to it | 22:02 |
| kanzure | gmail ref | 22:02 |
| fenn | the diy microfluidics is nifty | 22:03 |
| -!- Manion [n=chatzill@resnet-47-165.dorm.utexas.edu] has joined #hplusroadmap | 22:03 | |
| Manion | hello | 22:03 |
| -!- Manion is now known as Gene | 22:04 | |
| kanzure | Hey Gene. | 22:04 |
| -!- Gene is now known as newgenome | 22:04 | |
| fenn | hey, the paper is even free, you only have to pay 24ukp for it | 22:04 |
| newgenome | yes | 22:04 |
| * fenn wonders if wired editors know the diff between a paper and an abstract | 22:05 | |
| kanzure | We were just talking about some wearable computers, newgenome. | 22:05 |
| newgenome | that's cool | 22:05 |
| newgenome | augmented reality stuff? | 22:05 |
| kanzure | Nah, Nade was ranting in ##neuroscience yesterday about brain augmentation stuff so I started yelling at him to come in here. | 22:06 |
| kanzure | mostly augmented interfaces more than anything | 22:06 |
| newgenome | so neural interfaces, much less likely to give you eyestrain than display goggles | 22:07 |
| bkero | goggles strain eye, and throw off your inner ear wickedly bad. | 22:08 |
| kanzure | which one of us was mentioning retinal implants? | 22:08 |
| bkero | Your eye tries to put the HUD in your field of vision, but can't place the third dimension of it | 22:08 |
| bkero | So it just fucks your shit up | 22:08 |
| bkero | and you don't adapt | 22:08 |
| kanzure | oh, there was also the log from the other day about making scouters I guess | 22:09 |
| newgenome | scouters? | 22:09 |
| kanzure | something about projecting or lasering on to the eyeball for screens | 22:09 |
| kanzure | newgenome: Dragon Ball Z reference. | 22:09 |
| newgenome | I get it | 22:09 |
| fenn | head up display computer | 22:09 |
| kanzure | it's over nine thousand, etc. | 22:09 |
| newgenome | monocle though | 22:10 |
| kanzure | you going to complain? | 22:10 |
| fenn | bkero: why not just add some parallax to the HUD so that doesn't happen? | 22:10 |
| kanzure | fenn: http://heybryan.org/~bbishop/docs/Shrinky-Dink microfluidics: rapid generation of deep and rounded patterns.pdf <-- paper. | 22:10 |
| kanzure | erm | 22:10 |
| kanzure | but with %20's | 22:11 |
| fenn | we don need no steenking %20's | 22:11 |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has left #hplusroadmap [] | 22:11 | |
| -!- nsh- [n=nsh@87-94-146-186.tampere.customers.dnainternet.fi] has joined #hplusroadmap | 22:11 | |
| fenn | everything should just be underscores instead | 22:11 |
| kanzure | mm | 22:12 |
| fenn | pdms stamps can go down to nanometer resolution, so why dont we see semiconductors being etched with this process? (or am i just ignorant?) | 22:13 |
| kanzure | http://en.wikipedia.org/wiki/PDMS_stamp | 22:15 |
| kanzure | checking Google Scholar .. | 22:15 |
| kanzure | 'Non-Photolithographic Methods for Fabrication of Elastomeric Stamps for Use in Microcontact Printing -' | 22:16 |
| kanzure | 'Micropatterned Surfaces for Control of Cell Shape, Position, and Function ' | 22:16 |
| kanzure | 'Submicrometer Patterning of Charge in Thin-Film Electrets ' | 22:16 |
| kanzure | 'DNA microarray synthesis by using PDMS molecular stamp (II)' | 22:16 |
| kanzure | Hm. | 22:16 |
| fenn | spell my name with an S is in here somewhere: http://fennetic.net/pub/irc/Isaac Asimov - The Complete Stories Volume 1.pdf | 22:16 |
| fenn | with %20's | 22:16 |
| kanzure | it's also in your inbox | 22:16 |
| fenn | oh, oh well | 22:17 |
| fenn | i cheated and read the summary on wikipedia | 22:17 |
| newgenome | don't we all | 22:17 |
| kanzure | 'Patterning Self-Assembled Monolayers: Applications in Materials Science' | 22:18 |
| kanzure | fenn: "PDMS stamps" has led to a ridiculously large number of well-titled articles | 22:18 |
| kanzure | I wonder if the content is as good as the titles | 22:18 |
| fenn | it seems promising | 22:18 |
| newgenome | speaking of semiconductors | 22:18 |
| kanzure | 'A modified microstamping technique enhances polylysine transfer and neuronal cell patterning ' | 22:18 |
| kanzure | newgenome: http://heybryan.org/semiconductor.html | 22:18 |
| kanzure | http://heybryan.org/graphene.html is the graphene page I mentioned. | 22:19 |
| newgenome | does anyone know how they etch graphene? | 22:19 |
| kanzure | yes | 22:19 |
| kanzure | http://heybryan.org/instrumentation/instru.html for the AFMs and STMs | 22:19 |
| kanzure | http://heybryan.org/mediawiki/index.php/AFM_nanolithography (or something) | 22:19 |
| newgenome | that's very impracticle | 22:19 |
| kanzure | fine, | 22:19 |
| kanzure | http://heybryan.org/notes.html for etchants | 22:19 |
| newgenome | well thanks | 22:20 |
| kanzure | bwahah | 22:20 |
| newgenome | recently read they made some working qubits in graphene | 22:21 |
| kanzure | etching: | 22:21 |
| kanzure | http://forums.bit-tech.net/showthread.php?t=31766 | 22:21 |
| kanzure | http://www.nrgrecording.de/html/pcb.html | 22:21 |
| kanzure | http://web.media.mit.edu/~ladyada/resources/inhouseetch.html | 22:21 |
| kanzure | http://machinetools.com/MT/machines/index.tmpl?page=group_country&groupid=8160&countryID=CT9984321042268 | 22:21 |
| kanzure | http://www.fullnet.com/u/tomg/gooteepc.htm | 22:21 |
| kanzure | newgenome: yeah, that's on the graphene.html page I think | 22:21 |
| fenn | um, wrong kind of etching | 22:21 |
| kanzure | oops | 22:21 |
| kanzure | owhat do you mean? | 22:21 |
| kanzure | drawing? | 22:21 |
| kanzure | or chemical? | 22:21 |
| fenn | something that dissolves copper won't necessarily dissolve graphene | 22:21 |
| newgenome | I mean making lots and lots of 1 nm transistors | 22:22 |
| fenn | and those are macro scale processes, i'm not convinced it will work at very small dimensions | 22:22 |
| newgenome | yeah | 22:22 |
| newgenome | I don't think you can dissolve graphene | 22:22 |
| fenn | sure you can | 22:22 |
| fenn | shoot a beam of ozone at it | 22:22 |
| kanzure | uh | 22:22 |
| newgenome | maybe eat it away with electron beams | 22:22 |
| kanzure | where's percent_? | 22:22 |
| kanzure | he's the carbon nanotube guy | 22:22 |
| kanzure | he claims he can get us some CNTs if we ever need them | 22:22 |
| kanzure | let me see if he's on | 22:23 |
| newgenome | what I really want are some long CNTs, like a foot long or so | 22:23 |
| kanzure | hm | 22:24 |
| kanzure | why? | 22:24 |
| newgenome | of course, current technology can't make them | 22:24 |
| fenn | i think i'd settle for a couple hundred kilometers | 22:24 |
| newgenome | why? | 22:24 |
| newgenome | same here | 22:24 |
| -!- percent [n=percent@66.158.193.32] has joined #hplusroadmap | 22:24 | |
| kanzure | hi percent | 22:24 |
| kanzure | so, etchants for graphene? | 22:24 |
| kanzure | I know you work with C60, so maybe also graphene knowledge? | 22:24 |
| percent | I don't work with C60 (much). | 22:25 |
| percent | Heh, good luck, especially with nanotubes | 22:25 |
| percent | try ammonia | 22:25 |
| percent | It'll work if they're not inert. (They're inert). | 22:25 |
| fenn | percent: what about plasma etching? | 22:25 |
| percent | other than that, enjoy your piranha etch | 22:25 |
| percent | fenn: Depending on conditions, that may just create more nanoparticles | 22:26 |
| fenn | shoot a reactive ion at a couple kV and it should tear up anything | 22:26 |
| percent | Probably not graphene, but don't be surprised if you see nanotubes | 22:26 |
| percent | oh, you want to use a focused ion beam for graphene? | 22:26 |
| newgenome | good luck etching with ammonia when your trying to make 1 nm features | 22:26 |
| percent | That'll do it. | 22:26 |
| percent | FIB is a great technique. | 22:26 |
| percent | However, you might find it hard to find one. | 22:26 |
| kanzure | isn't plasma etching kinda like chemical vapor disposition processes? | 22:27 |
| newgenome | so you could make graphen microchips | 22:27 |
| kanzure | percent: we'll make one ... http://heybryan.org/instrumentation/instru.html should have some notes on that | 22:27 |
| percent | kanzure: yes | 22:27 |
| kanzure | hm, maybe not | 22:28 |
| percent | Remember, heat + anything ferrous + carbon = nanotubes | 22:28 |
| percent | also, plasma | 22:28 |
| percent | you need that too | 22:28 |
| newgenome | isn't there a better way to grow nanotubes? | 22:28 |
| percent | kanzure: I am usually all for homemade lab equipment, as you well know | 22:28 |
| kanzure | there was something in the news about radiowavelength plasma being applicable | 22:28 |
| percent | However, where will you find gallium? | 22:28 |
| kanzure | hm? | 22:28 |
| fenn | percent: another idea, not sure what it's called.. you have the graphene under some solvent that isnt quite nasty enough to react, and then shine patterned UV laser light on it (bounced off a MEMS mirror array) | 22:28 |
| newgenome | I have some | 22:28 |
| bkero | http://unitednuclear.com/chem.htm | 22:29 |
| percent | I happen to have MEMS mirror arrays :-) | 22:29 |
| bkero | $10 a gram | 22:29 |
| newgenome | in a tv? | 22:29 |
| newgenome | or seperate? | 22:29 |
| percent | separate | 22:29 |
| newgenome | dang | 22:29 |
| percent | kanzure: I switched labs. | 22:29 |
| kanzure | something better? | 22:29 |
| fenn | how much do they cost, ballpark? | 22:29 |
| percent | Well, carbon nanomaterials will always have my heart | 22:30 |
| percent | but my new project is actually MOVING FORWARD | 22:30 |
| -!- jk_ [n=jk@p57B746CE.dip.t-dialin.net] has joined #hplusroadmap | 22:30 | |
| * fenn gasps. someone said a dirty word | 22:30 | |
| percent | Because the professor is mean enough that all the bureaucrats shit themselves when he asks for something. | 22:30 |
| newgenome | http://www.goldmine-elec-products.com/prodinfo.asp?number=G15426 | 22:31 |
| newgenome | mems assembly | 22:31 |
| newgenome | $3 | 22:31 |
| fenn | i have some of those | 22:31 |
| kanzure | hm | 22:31 |
| newgenome | slightly damaged | 22:31 |
| percent | What exactly are you gentlemen trying to do? | 22:32 |
| newgenome | dna synthesis | 22:32 |
| percent | Also, may I remind you that nanoparticles are probably quite cytotoxic? | 22:32 |
| fenn | several things, most interesting being self replicating robots | 22:32 |
| percent | Oh, I know that | 22:32 |
| percent | I just wanted to know why you wanted to screw with graphene. | 22:32 |
| fenn | so, the graphene is an alternate process to create large numbers of small transistors | 22:32 |
| kanzure | cheap computers | 22:33 |
| kanzure | or cheap storage space | 22:33 |
| newgenome | with a crapload of computational power | 22:33 |
| fenn | go team | 22:33 |
| percent | And you think you can do better than 45nm? | 22:34 |
| newgenome | graphene is also cool because it is very strong | 22:34 |
| percent | Make no mistake, I'm not saying you can't. | 22:34 |
| fenn | it doesn't have to be anywhere near 45nm | 22:34 |
| newgenome | probably not | 22:34 |
| newgenome | just musing really | 22:34 |
| percent | I'm not shitting on your ideas here, I'm down for biohacking from here to milwaukee. | 22:34 |
| -!- jk_ [n=jk@p57B746CE.dip.t-dialin.net] has quit [Remote closed the connection] | 22:34 | |
| * fenn isnt quite sure how biohacking got into the conversation yet | 22:35 | |
| kanzure | uhm, 45 nm is ok? | 22:35 |
| kanzure | because the diy stuff gets down to 100 nm sometimes | 22:35 |
| fenn | 45nm is state of the art in Si fab | 22:35 |
| kanzure | no, 45 nm graphene transistors | 22:35 |
| kanzure | because the documentation says 10 nm | 22:35 |
| fenn | kanzure: nfw DIY is getting 100nm | 22:35 |
| percent | yeah, i'm also not too sure whether we really know how to design a processor | 22:35 |
| kanzure | nfw? | 22:35 |
| kanzure | percent: I do. | 22:36 |
| newgenome | good point | 22:36 |
| percent | you sure? | 22:36 |
| fenn | neer a fine way | 22:36 |
| percent | What's pipelining? No googling. | 22:36 |
| kanzure | you mean data path pipelining? | 22:36 |
| percent | Yes. | 22:36 |
| kanzure | like of instructions and so on? | 22:36 |
| kanzure | yeah | 22:36 |
| fenn | that's where you do a bunch of extra work for no good reason :) | 22:36 |
| percent | wrong | 22:36 |
| kanzure | excuse me? | 22:36 |
| kanzure | I'm fairly certain of this | 22:36 |
| percent | talking to fenn | 22:36 |
| kanzure | oh | 22:36 |
| kanzure | :) | 22:36 |
| bkero | Pipelining is rather important t o performance. | 22:36 |
| Nade | it's very important | 22:37 |
| kanzure | throw a few ALUs in and such | 22:37 |
| * fenn googles | 22:37 | |
| -!- JanK [n=jk@p57B746CE.dip.t-dialin.net] has joined #hplusroadmap | 22:37 | |
| kanzure | hobbyists commonly design their own 4 bit processors | 22:37 |
| percent | So you're not looking to build some badass monster from hell | 22:37 |
| kanzure | it's of course scalable if you actually care | 22:37 |
| percent | Just a simple graphene processor, to see if you can? | 22:37 |
| kanzure | if it works for small stuff then it's easily expanded | 22:37 |
| newgenome | dude a microprocessor is really complex | 22:37 |
| * kanzure has drawn graphene circuits with pencil markings | 22:37 | |
| percent | btw, I now have access to a focused ion beam | 22:37 |
| kanzure | newgenome: not 4 bit processors | 22:37 |
| percent | In addition to all my other shit | 22:38 |
| kanzure | the complexity is due to your automated generators | 22:38 |
| newgenome | I don't think you have any idea how complex | 22:38 |
| bkero | Uh | 22:38 |
| fenn | percent: the idea is to reduce the infrastructure required to build a processor | 22:38 |
| newgenome | well not 4 bit | 22:38 |
| bkero | I've built them | 22:38 |
| kanzure | you write out the circuits in Verilog or something | 22:38 |
| kanzure | bkero: hurray | 22:38 |
| bkero | Dude | 22:38 |
| kanzure | then you agree with me? | 22:38 |
| percent | newgenome: Tell us how complex they are. | 22:38 |
| bkero | I work for intel, my landlord is on the validation team for Nehalem. | 22:38 |
| percent | I KNOW it's possible to build a very, very simple processor. | 22:38 |
| bkero | He works with crazy ass versions of verilog all day. | 22:38 |
| kanzure | hahah | 22:38 |
| kanzure | awesome | 22:38 |
| kanzure | :) | 22:38 |
| kanzure | So, automatic gate/wiring generators and such right? | 22:38 |
| newgenome | 47 layers of several different materials in some of the least complex versions | 22:38 |
| percent | I didn't say microprocessor, did I? | 22:39 |
| kanzure | newgenome: ignore the Intel CISC stuff | 22:39 |
| newgenome | ok | 22:39 |
| percent | Photo litho is much harder. | 22:39 |
| bkero | Intel does stupid shit because they're intel | 22:39 |
| newgenome | well yeah | 22:39 |
| percent | fucking faggots | 22:39 |
| kanzure | percent: harder than what? | 22:39 |
| bkero | Look at ARM processors | 22:39 |
| kanzure | how did photo litho get in on this | 22:39 |
| kanzure | bkero: sure, anything RISC is probably okay. | 22:39 |
| percent | newgenome referenced it | 22:39 |
| kanzure | yeah, he was probably thinking Intel's stuff | 22:39 |
| kanzure | and as bkero says, | 22:40 |
| kanzure | they do crazy shit because they are Intel :-) | 22:40 |
| newgenome | photo lith is hard because photons are general bigger than the features they make | 22:40 |
| percent | DESTINATION COMES BEFORE SOURCE LOL | 22:40 |
| percent | Photo litho is hard because it requires special equipment | 22:40 |
| newgenome | really | 22:40 |
| percent | things we just can't do in our bathrooms. | 22:40 |
| newgenome | that too | 22:40 |
| kanzure | please remember who started photo lithography for processors back in their garages, ahem | 22:40 |
| kanzure | some people put bathrooms in garages, IIRC | 22:41 |
| percent | Okay, YOU align a mask by hand, then strip an oxide layer from an Si wafer | 22:41 |
| bkero | Is it me, or ahs the industry gotten so pathetic, we in our bathrooms are competing with them? | 22:41 |
| kanzure | anyway | 22:41 |
| kanzure | back to the 100 nm question | 22:41 |
| percent | i for one refuse to fuck with HF sans fume hood | 22:41 |
| fenn | kanzure: i thought it started in NASA for spy satellites or something like that | 22:41 |
| percent | really, really easy way to get dead | 22:41 |
| newgenome | yeah, my old school's semiconductor fab used halogen lights to expose wafers | 22:41 |
| kanzure | fenn: quite possibly ... I'd like a ref eventually | 22:41 |
| bkero | I've always been a big fan of galvanic etching to get circuits. :) | 22:41 |
| kanzure | is 100 nm resolution enough to make graphene transistors? | 22:42 |
| fenn | yes, but, how are you going to get 100nm resolution? | 22:42 |
| kanzure | apparently people have gotten down that far | 22:42 |
| kanzure | hold on | 22:42 |
| newgenome | maybe | 22:42 |
| percent | may i remind you that nanotech is defined by the govt as <100nm? | 22:43 |
| fenn | yay government | 22:43 |
| newgenome | so what is anything below 100 nm? | 22:43 |
| newgenome | femtotech? | 22:43 |
| newgenome | chemistry? | 22:43 |
| fenn | newgenome: um, maybe you should spend some time with wikipedia | 22:44 |
| kanzure | who cares that nanotech is less than 100 nm? | 22:45 |
| kanzure | by the govt? | 22:45 |
| newgenome | facepalm | 22:45 |
| kanzure | am I missing something? | 22:45 |
| bkero | God damn it mythtv can eat my dick. WTF is with this qt crap? | 22:45 |
| kanzure | qt? | 22:45 |
| fenn | percent isn't branch prediction essential to getting pipelining to work? | 22:45 |
| kanzure | :) | 22:45 |
| bkero | like gtk vs qt | 22:45 |
| kanzure | oh | 22:45 |
| * kanzure read "myth busters" | 22:45 | |
| percent | fenn: That's above my knowledge right now. | 22:45 |
| kanzure | MythBusters on gtk vs qt would be awesome. | 22:45 |
| kanzure | fenn: yes, it is | 22:45 |
| kanzure | for good pipelining at least | 22:46 |
| bkero | kanzure: I use neither. | 22:46 |
| fenn | ok so i was right *gives self a cookie* | 22:46 |
| bkero | Well, I use gtk for firefox, but thats it. | 22:46 |
| kanzure | grr, | 22:46 |
| kanzure | I've lost track of the amateur stm project that had 100 nm resolution | 22:47 |
| fenn | stm != photolitho | 22:47 |
| newgenome | yeah | 22:47 |
| kanzure | photolitho would mean lenses, right? | 22:47 |
| newgenome | yep | 22:47 |
| fenn | or mirrors | 22:47 |
| newgenome | I believe so | 22:47 |
| fenn | i suppose you could use holograms if you had a way of making them fine enough | 22:48 |
| kanzure | 1:1 correspondence with fineness? | 22:48 |
| fenn | like a diffraction grating | 22:48 |
| kanzure | you use lasers to make holograms | 22:48 |
| kanzure | wait, why can't we use a laser again? | 22:48 |
| kanzure | I forget. | 22:48 |
| fenn | i dont think there are any 100nm lasers | 22:48 |
| percent | WRONG | 22:48 |
| fenn | meh | 22:48 |
| percent | WRONG WRONG WRONG | 22:49 |
| fenn | please explain | 22:49 |
| percent | FENN IS WRON | 22:49 |
| percent | HEY EVERYONE FENN IS WRONG | 22:49 |
| kanzure | this is a good thing to be wrong about | 22:49 |
| newgenome | as I said before they use uv light to make features much smaller than uv light | 22:49 |
| percent | UV lasers are quite common | 22:49 |
| newgenome | why don't we stick to biohacking | 22:50 |
| newgenome | chip production seems a bit useless for biohacking | 22:50 |
| percent | Why don't we stick to your mom | 22:50 |
| newgenome | lol | 22:50 |
| kanzure | mostly because we're trying to close the loop on manufacturing and bootstrap it ourselves | 22:52 |
| kanzure | i.e., a von Neumann probe | 22:52 |
| kanzure | http://heybryan.org/projects/atoms/ has a good description of a von Neumann probe | 22:52 |
| fenn | fluorine excimer is 157nm, and while i guess free electron lasers could go higher i'm not really sure how they work or whether it would be practical for etching graphene | 22:53 |
| kanzure | 'The basic idea of a von Neumann probe is to have a space-probe that is able to navigate the galaxy and use self-replication (see RepRap and bio). The probe would contain hundreds of thousands of digital genomes (sequenced DNA), DNA synthesizers and sequencers, bacteria, embryos, stem cells, copies of the Internet Archive and a significant portion of the WWW in general, | 22:53 |
| newgenome | yeah I know that | 22:53 |
| kanzure | plus the immediate means and tools to copy all of the information and create a material embodiment, kind of like running an unzip utility on top of the thousands of exabytes predicted to be inexistence today. This would probably include many people, societies, even entire civilizations if we can collect enough data and begin to 'debug' civilization. The system might end up using an ion drive and a hydrogen collector, with on-board nucleosynthesis t | 22:53 |
| newgenome | I know about VNMs | 22:53 |
| newgenome | I work with reprap remember? | 22:54 |
| newgenome | anyway | 22:54 |
| newgenome | the best proposal I have seen for total replication is this | 22:54 |
| fenn | wait, how do you define 'total replication'? | 22:55 |
| newgenome | * In 1998, C. Phoenix [1171] informally sketched out a design for a macroscale kinematic replicator a few cubic feet in volume that would use two hydraulic-powered manipulator arms to machine, then assemble, its own components out of a soft plastic feedstock which would then be ultraviolet-cured to yield hard plastic parts, analogous to the stereolithography system offered by Vicale Corp.... | 22:55 |
| newgenome | ...[934]. The acoustically-powered plastic replicator, composed of perhaps ~2000 parts, would be controlled by an onboard 8086-class computer built from cured-plastic fluidic logic elements including 1 KB of RAM, receiving instructions from a 1400-foot long strip of hole-punched control tape. Most details such as specific materials and assembly procedures, basic closure issues, process error... | 22:56 |
| newgenome | ...rates, and accessibility of required machining tolerances were not explicitly addressed. | 22:56 |
| newgenome | http://www.molecularassembler.com/KSRM/3.20.htm | 22:56 |
| kanzure | bah, stop quoting Freitas | 22:56 |
| newgenome | it's not freitas | 22:56 |
| kanzure | KSRM is Freitas | 22:56 |
| newgenome | oh | 22:56 |
| kanzure | Chris Phoenix is the direct of the center for molecular nanotech or something | 22:57 |
| kanzure | ethical stuff | 22:57 |
| kanzure | he's turned somewhat environmentalist recently | 22:57 |
| newgenome | the thing about fluidic logic elements is that you don't need that much accuracy to make them | 22:57 |
| newgenome | this isn't using any nanotech | 22:57 |
| kanzure | Freitas didn't actually have a design for his stuff. | 22:57 |
| newgenome | I know | 22:57 |
| kanzure | or that Phoenix design for example | 22:57 |
| kanzure | :-( | 22:57 |
| kanzure | so | 22:57 |
| kanzure | that's where this project comes in | 22:57 |
| kanzure | http://heybryan.org/exp.html | 22:58 |
| fenn | you dont need much accuracy to make a transistor either | 22:58 |
| kanzure | we have a 'validator' | 22:58 |
| kanzure | yeah, I was thinking about a giant jar with electrodes for a transistor once | 22:58 |
| kanzure | a gel or liquid inside or some such | 22:58 |
| kanzure | anyway, | 22:58 |
| kanzure | we have a 'validator' to check if a design is self-replicating | 22:58 |
| kanzure | sort of. :) | 22:58 |
| fenn | we do? | 22:58 |
| kanzure | well | 22:58 |
| kanzure | no | 22:58 |
| newgenome | but this is elementally cheaper | 22:58 |
| kanzure | but theoretically we do | 22:58 |
| kanzure | if we weren't so lazy | 22:58 |
| * fenn yawns | 22:58 | |
| kanzure | hm? | 22:58 |
| kanzure | look, you validate by checking whether or not all models exist | 22:59 |
| kanzure | so | 22:59 |
| newgenome | as in no machinery is needed to purify silicon and what not | 22:59 |
| kanzure | pseudocode: if the dependencies are nonexistent, it's bullshit | 22:59 |
| kanzure | oh | 22:59 |
| kanzure | I guess this is a good time to bring up http://heybryan.org/fractal.html as well ... since we have so many arguments on the definition of self-replication ;-) | 22:59 |
| newgenome | you could make fluidic logic gates from melted down rocks | 22:59 |
| kanzure | Wittig, the guy on campus, was yelling at me about this | 23:00 |
| kanzure | wait, what ? | 23:00 |
| kanzure | the linkydinky ones? | 23:00 |
| newgenome | you don't have to dedicate machinery to obtaining near pure silicon from the rocks | 23:01 |
| newgenome | no | 23:01 |
| fenn | you can also make logic gates out of rods, sorta like tinkertoys | 23:02 |
| fenn | or relays, or vacuum tubes | 23:02 |
| kanzure | relays? | 23:02 |
| kanzure | yep | 23:02 |
| fenn | and probably a million other things | 23:02 |
| newgenome | fluidic logic gates don't wear out | 23:02 |
| fenn | are you sure about that? | 23:03 |
| newgenome | just as long as stuff doesn't grow in your fluid | 23:03 |
| newgenome | and your fluid is filtered | 23:04 |
| fenn | i've seen some fluidics sequencers for fountains, but they were very simple and rather slow compared to, say, a vacuum tube | 23:04 |
| fenn | so if you can get a trillion switching events out of your vacuum tube, and a trillion switching events out of your sequencer, which is better? | 23:05 |
| fenn | wah | 23:05 |
| newgenome | check this out: http://www.dself.dsl.pipex.com/MUSEUM/COMMS/fluidicgramophone/fluidgram.htm | 23:05 |
| fenn | not a very good diagram | 23:06 |
| newgenome | http://en.wikipedia.org/wiki/Fluidics | 23:07 |
| fenn | there are other amplifiers that have no moving parts, no high purity ingredients, and don't wear out | 23:07 |
| fenn | look up saturable core reactor, aka magnetic amplifier | 23:07 |
| newgenome | then you have to mine iron | 23:07 |
| kanzure | bacteria could do that | 23:08 |
| kanzure | or you just drill and melt and do whatever else it is you do to mine iron | 23:08 |
| newgenome | http://www.blikstein.com/paulo/projects/project_water.html | 23:09 |
| newgenome | good point | 23:09 |
| kanzure | MIT Media Lab? | 23:09 |
| kanzure | we've had people from there show up in here | 23:09 |
| newgenome | yup | 23:09 |
| kanzure | wait | 23:09 |
| kanzure | this is down the hall from David | 23:09 |
| newgenome | but fluidic logic gates do have problems | 23:10 |
| newgenome | especially if your VNM lives in space | 23:10 |
| newgenome | fluids tend to evaporate in a vaccuum | 23:11 |
| fenn | or in general | 23:11 |
| bkero | fluids evaporate in general? | 23:13 |
| fenn | yes | 23:13 |
| bkero | If they've achieved equal temperature with the gasses above them, why would they? | 23:14 |
| newgenome | there is also trouble in obtaining fluid on say an asteroid | 23:14 |
| fenn | bkero: because statistical variation will ensure that a molecule at the surface will reach vaporization temperature | 23:14 |
| fenn | did that make sense? | 23:15 |
| kanzure | newgenome: ice | 23:15 |
| fenn | the tip of the boltzmann distribution will be hot enough to be gaseous | 23:15 |
| newgenome | yeah I just thought that | 23:15 |
| kanzure | :p | 23:15 |
| kanzure | hehe | 23:15 |
| fenn | god wikipedia sucks at math | 23:16 |
| newgenome | so what are your ideas on getting energy for a VNM? | 23:16 |
| newgenome | solar or anything else? | 23:17 |
| fenn | solar power is pretty straightforward | 23:17 |
| fenn | geothermal is also easy | 23:17 |
| fenn | not molten magma stuff, just the small 10-20 C difference between air and underground | 23:18 |
| newgenome | so is the VNM we are designing going to live on earth or some other place | 23:18 |
| fenn | different designs for different environments | 23:18 |
| fenn | hard to experiment with designs for asteroids right now :) | 23:18 |
| newgenome | so let's get something that works on earth | 23:19 |
| kanzure | right | 23:19 |
| kanzure | the idea of 'skdb' is to throw a lot of manufacturing processes into the 'pot' and then find dependency loops between them | 23:19 |
| kanzure | supposedly, these processes that can make each other's machinery and so on will be a 'von neumann probe' | 23:19 |
| newgenome | water and air are readily available | 23:19 |
| kanzure | well, not a VNM, but a self-replicator | 23:20 |
| kanzure | ad hoc design is not going to work really | 23:20 |
| newgenome | macroscale or something a bit smaller? | 23:20 |
| fenn | macroscale is easier to work with | 23:20 |
| kanzure | yep | 23:20 |
| newgenome | microscale replication is very easy | 23:20 |
| fenn | pfff | 23:20 |
| kanzure | hands and some elbow grease | 23:20 |
| kanzure | haha | 23:20 |
| fenn | where did this myth come from? | 23:21 |
| kanzure | that's why Jupiter replicates itself every night! | 23:21 |
| newgenome | John Szostak | 23:21 |
| newgenome | he's trying to recreate some of the first life forms on earth | 23:21 |
| fenn | and what were the first life forms on earth like? | 23:21 |
| kanzure | reverse engineering of the RNA world? | 23:21 |
| newgenome | http://www.youtube.com/watch?v=U6QYDdgP9eg&eurl=http://scienceblogs.com/pharyngula/2008/06/wondering_how_life_got_started.php | 23:21 |
| newgenome | mainly protocells | 23:22 |
| fenn | i refuse to watch a youtube of some guy talking about his blog | 23:22 |
| kanzure | hm, pharnygula has an irc channel | 23:22 |
| kanzure | probably lots of dawkin people :) | 23:22 |
| newgenome | fast forward to about 3:38 | 23:23 |
| newgenome | 4:00 might be better | 23:23 |
| newgenome | http://www.exploringorigins.org/ | 23:23 |
| newgenome | this is good too | 23:23 |
| newgenome | it explains why dna strands are double stranded | 23:24 |
| kanzure | newgenome: Andy's favorite subject is the RNA world | 23:24 |
| newgenome | yeah RNA world is fun stuff | 23:24 |
| kanzure | he's coauthored with Stuart Kauffman and so on | 23:24 |
| kanzure | so he's kind of in the loops .. | 23:24 |
| newgenome | did you watch the video yet | 23:24 |
| kanzure | no, I'm fighting Cell | 23:25 |
| newgenome | lol | 23:25 |
| newgenome | http://www.exploringorigins.org/nucleicacids.html | 23:25 |
| newgenome | the third video on this page is a good example of the type of replication you are talking about | 23:26 |
| newgenome | the point is that on the molecular scale things self organize making replication easier | 23:28 |
| kanzure | replication is a wee bit more than just copying bits of course | 23:28 |
| newgenome | yeah | 23:28 |
| fenn | you have to make the bits too | 23:28 |
| kanzure | bit fairydust? | 23:29 |
| kanzure | pixy/pixeldust? | 23:29 |
| newgenome | on the macroscale it's hard to get self organization going | 23:29 |
| newgenome | you need an assembler | 23:29 |
| fenn | newgenome: feynman said it pretty well "i think god has pretty good quality control on atoms" | 23:30 |
| newgenome | do you have any ideas on making one? | 23:30 |
| fenn | of course this ignores isotopes an so on :0 | 23:30 |
| newgenome | yeah | 23:30 |
| fenn | a macroscale assembler? | 23:30 |
| newgenome | yeah | 23:30 |
| newgenome | an assembler that puts together the machine that makes the bits | 23:31 |
| newgenome | or assemblers | 23:31 |
| fenn | woops, that was drexler, not feynman | 23:31 |
| fenn | newgenome: i think you could do a lot with ceramics and a holy length artifact | 23:32 |
| fenn | like a double ball bar for example | 23:32 |
| fenn | since the replicators don't have to interact with each other much, drift in the calibration standards wouldn't be such a big problem as it is in a constantly churning and intermixing civilization | 23:33 |
| fenn | (like our own) | 23:33 |
| newgenome | I don't know what you mean | 23:33 |
| fenn | as long as all the parts fit together, it doesn't matter if your extruder nozzle is exactly 1mm in diameter | 23:33 |
| newgenome | do you mean like natural selection or having people maintain them | 23:34 |
| kanzure_ | There was a ridiculous lack of laptops. | 23:34 |
| fenn | kanzure_: you should write to nicholas negroponte | 23:34 |
| newgenome | you also have to move that extruder precisely enough to make parts | 23:34 |
| newgenome | which brings me to another point about fluidic logic gates | 23:35 |
| newgenome | they do not have to be made very accurately | 23:35 |
| kanzure | woah | 23:35 |
| newgenome | in order to work | 23:35 |
| kanzure | that was an old message from the other machine | 23:36 |
| kanzure | keyboard on floor | 23:36 |
| kanzure | ignore | 23:36 |
| fenn | yes, the bay area savages | 23:36 |
| fenn | they are so primitive, i hear many of them are forced to ride bicycles to work | 23:36 |
| newgenome | this should help you understand fluidics: http://blog.modernmechanix.com/2007/10/03/fluid-transistor-circuits/ | 23:38 |
| kanzure | can't we just throw this into our (supposed) computational system so that we don't have to remember everything each time we try a new route? | 23:40 |
| kanzure | blah | 23:40 |
| kanzure | that was the idea in the first place .. | 23:40 |
| -!- cow_town [n=cowtown@71.237.209.152] has joined #hplusroadmap | 23:41 | |
| kanzure | Hi cow_town | 23:41 |
| -!- cow_town [n=cowtown@71.237.209.152] has left #hplusroadmap [] | 23:41 | |
| newgenome | so the next big thing on this fluidic system is the replicability of the the hydraulic arms and accuracy of the extruder positioning system | 23:42 |
| kanzure | newgenome: I'm going to hit the sack since I have an 830 calculus class I'd rather not miss, this morning being the first day and the first missed class | 23:46 |
| newgenome | same here | 23:46 |
| kanzure | but there's a bit more to mention | 23:46 |
| kanzure | so maybe that could be for another time? | 23:47 |
| newgenome | cay | 23:47 |
| newgenome | cay | 23:47 |
| newgenome | cya | 23:47 |
| kanzure | heh' | 23:47 |
| -!- newgenome [n=chatzill@resnet-47-165.dorm.utexas.edu] has quit ["ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]"] | 23:47 | |
| kanzure | hm | 23:47 |
| kanzure | he's not an idler? | 23:47 |
| JanK | so this is hplusroadmap. nice to read. | 23:51 |
| kanzure | Hi JanK. | 23:51 |
| * kanzure wonders who JanK is | 23:51 | |
| willPow3r_ | jan k | 23:52 |
| JanK | may be you remember me: metamath.org offlist from exi-chat | 23:52 |
| kanzure | Hm? | 23:52 |
| kanzure | hm | 23:52 |
| kanzure | 23:52 | |
| kanzure | "Jan Klauck" <jklauck@uni-osnabrueck.de> | 23:52 |
| JanK | yes. :) | 23:53 |
| JanK | I was compiling some stuff and tooke the time to have a look at your channel. | 23:54 |
| kanzure | Integrated Assessment Center? | 23:54 |
| kanzure | Red Queen? | 23:54 |
| kanzure | Hm. | 23:54 |
| JanK | not center, just integrated assessment | 23:54 |
| JanK | well, red queen was a (lame) class project | 23:55 |
| kanzure | heh | 23:55 |
| kanzure | red queen's race is a common theme around here | 23:55 |
| JanK | sure, but our implementation sucked | 23:55 |
| JanK | it was about multi agent systems | 23:55 |
| kanzure | in which context? | 23:56 |
| JanK | complex adaptive systems | 23:56 |
| kanzure | agents like, ' "intelligent" agents' | 23:56 |
| kanzure | ha | 23:56 |
| kanzure | *ah | 23:56 |
| JanK | not in this class :( | 23:57 |
| JanK | the agents were anything but intelligent. pure reactive architecture | 23:57 |
| kanzure | more like cellular automata? | 23:58 |
| percent | your mom is a sexual automata | 23:58 |
| percent | I just noticed there aren't any ops here. | 23:58 |
| percent | Why? | 23:58 |
| kanzure | my mom is a stripper | 23:58 |
| kanzure | go away | 23:58 |
| JanK | not like ca | 23:58 |
| kanzure | percent: dunno | 23:58 |
| kanzure | didn't think about it really | 23:58 |
| percent | I just noticed I hadn't been banned yet | 23:59 |
| percent | And I wondered why | 23:59 |
| willPow3r_ | its because you're gay | 23:59 |
| percent | Since I'm so rarely helpful to the conversation | 23:59 |
| percent | there, NOW you guys are acting like hackers | 23:59 |
| percent | the fuck is up with this professional shit | 23:59 |
| kanzure | professional? | 23:59 |
| percent | you're like | 23:59 |
| willPow3r_ | its what happens when you try to accomplish something using intelligence | 23:59 |
| percent | not cursing or talking about goatse | 23:59 |
Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!