--- Day changed Thu Aug 21 2008 00:00 < kanzure> wonder if he was doing what I'm doing now 00:00 < fenn> more or less 00:00 < kanzure> the stealing, I mean 00:00 < kanzure> but it's not really stealing 00:00 < kanzure> since I do have access permissions 00:03 < kanzure> oh, nevermind.. the urls are systematic 00:09 < nsh> also: their mums 00:09 < nsh> is my philosophy 00:17 < kanzure> hm, why isn't the library on AIM? 00:17 < kanzure> it was there earlier today .. 00:20 < kanzure> http://www.lib.utexas.edu/indexes/titles.html?id=372 00:20 < kanzure> bah 00:20 < kanzure> I think these guys are telling lies 00:22 < kanzure> 'Downloading of entire journal issues or complete journal volumes in a systematic fashion is strictly prohibited.' 00:22 < nsh> meh 00:22 < kanzure> 'you may only download them randomly!' 00:22 < kanzure> very well then 00:23 < kanzure> wait, you can't prove it was random 00:23 < kanzure> what the fuck 00:23 < kanzure> we're boned. 00:23 < kanzure> "well, I tried using a pseudo random number generator ... but if you guys have a random number generator, so that I don't systematically access it, that would be awesome." 00:23 < elias`> http://www.virtualp.us/Dilbert-Oct_25_001.jpg 00:23 < kanzure> :) 00:25 < kanzure> interesting, 00:25 < kanzure> searching for "Downloading of entire journal issues or complete journal volumes in a systematic fashion is strictly prohibited." brings up multiple results 00:25 < kanzure> even a string of four characters is rarely found to appear more than once on the web 00:26 < kanzure> http://www.google.com/search?q=%22Downloading+of+entire+journal+issues+or+complete+journal+volumes+in+a+systematic+fashion+is+strictly+prohibited.%22&num=100&hl=en&client=opera&rls=en&hs=gSd&filter=0 00:26 < kanzure> 50 results 00:26 < kanzure> hah 00:34 < kanzure> http://heybryan.org/shots/2008-08-20_irony.png 00:40 < fenn> if you set your user agent to googlebot they'll never know what hit 'em 00:40 < kanzure> oh fuck 00:41 < kanzure> that should have been obvious 00:41 < kanzure> thank you 00:42 < fenn> or perhaps one of the lesser known search engines 00:42 < elias`> they might do something nasty if they notice the IP is not in some expected range. 00:42 < elias`> in that case 00:43 < kanzure> I'm a sekrit google project 00:44 < kanzure> where googlebot is on a university dorm computer 00:44 < kanzure> called backrub 00:45 < kanzure> (2008-08-20 19:49:23) nchaimov: I like the IEEE license, which adds "The use of robots or intelligent agents on this site is a violation of your subscription license agreement" 00:45 < kanzure> http://www3.ntu.edu.sg/lib/collections/db/AAY-2391use.htm 00:46 < kanzure> this is ridiculous 00:48 < kanzure> "I'm not downloading it, I'm caching it" 00:49 < kanzure> "DNS architecture foo" 01:05 < kanzure> fenn, I always thought that webmasters were smart enough to check for the user agent being google /and/ for it to be the ip range that google uses 01:07 < nsh> one day... 01:07 < nsh> there'll be enough bandwidth and cheap storage 01:07 < kanzure> to download the internet? 01:07 < kanzure> google's been doing it, nsh 01:07 < nsh> that somehow, all of google's data, is going to accidentally end up on the web 01:08 < kanzure> hm 01:08 < shobin> what are you doing bryan? 01:08 < nsh> there are sleeper cells of information liberation in all quarters 01:08 < nsh> most don't even know that they are. 01:08 < kanzure> shobin: http://heybryan.org/projects/autoscholar/ 01:09 < kanzure> oh, 01:09 < kanzure> http://heybryan.org/shots/2008-08-17_gdmap.png 01:10 < kanzure> the pink stuff to the left of the yellow outlined square. All of that pink stuff is Nature. 01:10 < shobin> how do you set your user agent to the google bot? 01:10 < kanzure> wget --help | grep agent 01:11 < shobin> very decent image to text conversion 01:12 < shobin> I have most of the 2600 magazines scanned in pdf form, I should run that app on them 01:13 < shobin> lawl nevermind, it's pretty bad 01:13 < kanzure> :) 01:13 < shobin> speaking of which 01:13 < shobin> google has speech to text software now 01:14 < kanzure> I'm looking into some page segmentation algorithms 01:14 < kanzure> they've had that for a while 01:14 < kanzure> call them with your cell phone 01:14 < shobin> it's up now 01:14 < kanzure> and they do searches for you. 01:14 < kanzure> oh 01:14 < shobin> on the youtube candidate channels 01:14 < shobin> yeah I use goog-411 01:15 < shobin> it's good for them, parse through videos with their existing technologies 01:15 < kanzure> you know, 01:15 < kanzure> how many keys are on a cell phone these days, no more than 12 right? 01:15 < kanzure> that should be doable with brain implant tech .. 01:15 < kanzure> esp. re: the man who could control a cursor "with the power of his mind" 01:15 < shobin> i have 19, but no more than 14 should be necessary 01:16 < kanzure> right 01:16 < kanzure> then do a HUD 01:16 < kanzure> and the typical ear mount and speaker mount 01:16 < kanzure> this shouldn't be hard.. 01:16 < shobin> yeah 01:17 < shobin> given everything people have done it should be pretty straightforward 01:17 < shobin> it would be neat to see 01:17 < kanzure> oh, speaking of which 01:17 < kanzure> http://heybryan.org/2008-08-15.html 01:18 < kanzure> I need some electricity gurus to help me figure out how to check feasability requirements on all of this 01:18 < kanzure> I was thinking of just ordering a pcb and handmake the electrodes as described in those documents 01:18 < kanzure> do we have any electricity buff in here? 01:20 < shobin> do you know anything about EEGs bryan? 01:20 < kanzure> Not much. 01:20 < kanzure> I avoid them really 01:20 < kanzure> bkero was thinking of doing some EKG methinks 01:20 < shobin> I've wanted one 01:20 < shobin> too expensive though 01:20 < kanzure> did you see openeeg? 01:20 < shobin> yes 01:20 < kanzure> even the soundcard was too much? 01:21 < shobin> it seemed messy 01:22 < elias`> I'm currently thinking of trying out the emotiv headset whenever they begin selling them 01:23 < shobin> that had a hell of a debut eh? 01:23 < shobin> still, I'm hopeful it'll be good 01:24 < kanzure> bah, that eeg headset? 01:24 < elias`> not sure what they were called, but something that is able to read/predict what you're saying inaudibly (subvocally?) would probably be nice to have too. 01:25 * kanzure once saw a paper about browsers and subvocalization 01:28 < kanzure> http://www.wpi.edu/Pubs/E-project/Available/E-project-010808-102024/unrestricted/IQP_Ceccarini.pdf 01:28 < kanzure> that might have been it 01:28 < kanzure> 'Web Browser Control Using EMG Based Sub Vocal Speech Recognition' 01:28 < kanzure> http://ieeexplore.ieee.org/iel5/9518/30166/01385845.pdf 01:31 < kanzure> I wonder how loud you have to talk. 01:32 < nsh> oh cool 01:32 < nsh> i don't think you have to make any actual noise 01:32 < kanzure> 'Specifically, we use non invasive aggregate surface 01:32 < kanzure> measurements of electromyographic signals or EMGs to 01:32 < kanzure> categorize muscle activation prior to sound generation [3]. 01:32 < kanzure> Such signals arise when reading or speaking to oneself with or 01:32 < kanzure> without actual lip or facial movements. Hence the information 01:32 < kanzure> we are using does not show up using external observation, nor 01:32 < nsh> just "rehearsal speech" 01:32 < kanzure> in current methods used to enhance speech recognition, such 01:32 < kanzure> as machine lip reading. 01:32 < kanzure> ' 01:32 < kanzure> yeah 01:32 < kanzure> awesome 01:32 < nsh> it's a cool technology 01:32 < nsh> i'm waiting for the telepathy app. 01:32 < nsh> there's already some proof of concept 01:33 < kanzure> ' in previous work we showed the adequacy of EMG signals for the control of a virtual joystick and virtual numeric keypad entry [2]. In [1] we demonstrated recognition of a small sub acoustic control vocabulary.' 01:33 < nsh> http://gburt.blogspot.com/2008/03/subvocalization.html 01:35 < kanzure> I wonder how quick speech synthesizers are 01:35 < nsh> that demo looked a bit fishy 01:35 < nsh> took 10 seconds to say "yeah definitely" 01:36 < kanzure> "Little work testing the ability of EMG to perform speech recognition by itself appears to have been done. Parallel work for speech recognition augmentation along the lines of that in our word experiments was performed by Chan [8]. He proposed supplementing voiced speech with EMG in the context of aircraft pilot communication. In their work they studied the feasibility of augmenting auditory speech information with EMG signals recorded from primary fa 01:36 < kanzure> face reading, anyone? 01:36 < kanzure> a game where you have to smile :) 01:37 < kanzure> using separate cheeks for separate controls 01:37 < nsh> microgesture human-computer interface? 01:37 < kanzure> right 01:37 < nsh> what was vinge's last book 01:37 < nsh> rainbows end? 01:37 < kanzure> reading rainbow or something 01:37 < nsh> that sorta thing 01:37 < kanzure> yeah 01:37 < nsh> it's funny 01:38 < kanzure> they say that people with asperger's can't decode nonverbal signals 01:38 < nsh> you ever read the foundation books? 01:38 < nsh> asimove 01:38 < kanzure> yes 01:38 < nsh> the second foundation were "telepathic" 01:38 < kanzure> so if they can't decode them, then let's just measure them excessively 01:38 < nsh> in the sense that they had microgesturing down to a fine art 01:38 < kanzure> same with Zindell and the ascetics 01:38 < nsh> oh? 01:38 < nsh> another sci-fi? 01:39 < kanzure> 'As a master cetic might, I tried to read the truth from the flickers of light reflecting from her bright irises and from the set of her wide mouth. But the only truth that came to me was an old truth: I could no more read her face than I could descry the future.' 01:39 < kanzure> http://heybryan.org/docs/Zindell,%20David%20-%20Neverness%20(v1.0).txt 01:39 < nsh> Neverness? 01:39 < nsh> ah yeah 01:39 < kanzure> http://heybryan.org/quotes.html#Posthuman <-- quotes. 01:40 < nsh> i think i started reading this 01:40 < nsh> from your server 01:40 < nsh> but got distracted and forgot about it 01:41 < kanzure> 'Although I was no cetic, it seemed they were in danger of copying, and perhaps running, each other's programs. Such was the danger of sharing the pit of a lightship-if one could believe the warnings of the cetics and the programmers. So far as I knew, no two pilots had ever faced the same thoughtspace at the same time. When I hinted of this danger, and hinted of my worry, Justine smoothed the folds of her robe, straightened her back stiffly, and told 01:42 < kanzure> 'He must have taken this as a sign of encouragement for he continued, "Once a time on Urradeth there was a cetic who had a great flock of sheep. But the cetic was very busy fashioning metaprograms which he hoped would control his own baser, more mundane programs. Consequently, he had little time to tend his flock. Often they wandered off into the forest or stumbled into snowdrifts, and worse, they ran away because they knew that the cetic wanted their 01:43 * nsh smiles 01:45 < kanzure> "The cetic, after much hard work finally decoded his death program, or, I should say, his fear-of-death program. He purged it from his brain, from his very neurons. And there are many poems written of this-the cetic discovered that it is the fear of death which enslaves us. You might say the dread of the dying self sends us stumbling blindly about our daily tasks as if we are nothing but sleepwalking robots programmed to feed and drink and copulate. F 01:50 < kanzure> 'Her eye twitched then, and I saw what I should have seen long ago: My mother was addicted to toalache-the facial tics were the result of her hiding this shame from her friends, and from herself. I saw other things, too, other programs: The layers of fat girdling her hips, which betrayed her compulsive eating programs and love of chocolate drinks and candies; her arrogant speech patterns, the clipped sentence fragments hinting at her belief that other 01:50 < kanzure> Anyway, I doubt we could get down to that specialized an art ;-) 01:50 < kanzure> with emg 02:09 < bkero> shobin: I'm looking into doing some EKG and EEG things. 02:10 < ybit> anyone know if the openstim project is elite? 02:11 < ybit> i don't want to waste my time sending them an email so that i can register on the site 02:18 < kanzure> don't bother 02:18 < kanzure> I know everybody active on that project 02:18 < kanzure> it's Superkuh and Ed Boyden. 02:18 < kanzure> good luck. 02:18 < kanzure> http://edboyden.org/ 02:18 < kanzure> http://superkuh.ath.cx/ 02:26 < ybit> i already bothered 02:26 < ybit> it says it is open to all, we'll see 02:31 < ybit> i know you all have seen this before, i have atleast, but it's still cool: http://www.auvsi.org/competitions/water.cfm 02:41 < kanzure> http://books.google.com/books?id=pUhIQlp6EAwC&pg=PA129&dq=%22David+Zindell%22&ei=PtWsSIbgFqXayAS_zp2ABw&client=opera&sig=ACfU3U0v_FgDzr6-KjFIJCgPrAMPQ8dWHw <-- hah 02:42 < marainein> where's the hah? 02:42 < kanzure> The name of the book is "Future on Ice" 02:42 < kanzure> his short story was in the context of a city on ice 02:42 < kanzure> So, naturally, he would be skeptical. 05:30 < kanzure> Oh crap. ScienceDirect has an HTML validation error. 05:30 < kanzure> name="rememberid"id="rememberMe" 05:33 < ybit> are you extracting all the articles from the site? 05:35 < ybit> i fully support you, just curious how much hdd space that will consume 05:38 < kanzure> Nature was 40 GB. 05:38 < kanzure> and that was probably only half of Nature really 05:39 < ybit> heh, i didn't realize you had already grabbed Nature 05:43 < kanzure> Nature grabbed me. 05:44 < ybit> good one 05:50 < ybit> joking with you of course 05:50 < ybit> what modules did you use? 05:53 < kanzure> WWW::Mechanize ? 05:54 < kanzure> what are you asking? 05:56 < ybit> you answered, i was just curious 06:01 < kanzure> I don't know what you mean by modules though. 06:03 < ybit> python has modules, i was curious if you were using urllib, and you are 06:03 < ybit> urllib2* 06:04 < ybit> urllib would be considered a module, not sure what else to call them atm 06:50 < kanzure> huh, this is peculiar 06:50 < kanzure> the number earlier was 1998 06:50 < kanzure> now it's 1997 journals 06:50 < kanzure> hrm 09:25 < bkero> lolperl 09:28 < kanzure> ? 09:30 * bkero shrugs 09:35 < bkero> Sleep deprivation 09:37 < kanzure> Sleep is for the weak. 09:37 < kanzure> And women. 09:38 < bkero> http://forums.somethingawful.com/showthread.php?threadid=2934761 09:39 < bkero> My brain stem needs 15 minutes of rest per night. 09:39 < bkero> The rest is just to fight muscle fatigue. 09:39 < kanzure> Explain to me somethingawful. 09:39 < kanzure> I think I missed that duckwagon ... even though I was around at the right time (2001?) 09:49 < bkero> It's a bunch of internet 1.0 people who are cycinal of the rest of the internet. 09:49 < kanzure> neat 09:49 < kanzure> sounds like surfraw 09:52 < bkero> Sort of 09:52 < bkero> Less unixy though 20:01 < nsh> what's kickin'? 20:01 < nsh> *yawns* 20:24 < kanzure> http://www.livescribe.com/smartpen/techspecs.html 20:24 < kanzure> from Kurt, the NJ contact. 20:25 < kanzure> how is this 'searchable' ? 20:42 < nsh> mmm 20:42 < nsh> watching that Dawkins Ventor discussion 20:43 < nsh> Dawkins is saying the chromosome is pure information 20:44 < nsh> "it's more than just saying you can pick up a chromosome and put it somewhere else: it's pure information, you could put it in a book, send it over the internet. you could put it on a magnetic disc and in a thousand years and then with the technology they'll have then, it will be possible to reconstruct whatever living organism that was here now" 20:44 < nsh> WRONG 20:45 < nsh> i hate it when the leading scientists in a field of academic endeavour are so painfully wrong 20:46 < kanzure> you refer to the lack of phenotypic preservation? 20:46 < kanzure> and the fact that you don't have the cell membranes and so on ? 20:46 < kanzure> I mean, surely the repair mechanisms would be enough to work off of 20:47 < kanzure> 'surely' <-- nevermind 20:47 < nsh> the repair mechanisms of *what*? 20:47 < nsh> letters in a book 20:47 < nsh> patterns of magnetism on a cdrom? 20:47 < nsh> you still need an organism to do anything with dna 20:47 < nsh> in trashing the opening quotation 20:48 < nsh> dawkins has neglected to admit that it is still an accurate set of statements 20:48 < nsh> despite being misinformed 20:48 < kanzure> theoretically you don't need an entire organism because of evolution 20:48 < kanzure> but I bet Dawkins isn't thinking about this 20:49 < nsh> not sure i understand what you mean, kanz 20:50 < nsh> my point was the dna is meaningless until it is interpreted 20:50 < nsh> it is the coupling of the digital data with the complexity of the interpreting environment that produces the final product 20:50 < kanzure> my point is that if you are in the future and you have cow dna and you don't have a cow, you could theoretically repeat the evolutionary steps to get back to a cow 20:51 < nsh> well, i would suppose it'd depend on what cow-analogues you did have 20:51 < kanzure> no, assume you have nothing analogous to a cow 20:51 < kanzure> I'm talking about redoing the RNA world from scratch ;-) 20:51 < nsh> oh, no. 20:51 < kanzure> and the Urey-Grey/Miller thingy experiments etc. 20:51 < kanzure> yeah 20:51 < kanzure> it'd suck 20:51 < nsh> you couldn't make a cow. 20:51 < kanzure> nature did it once before 20:51 < kanzure> Dawkins isn't thinking about this of course 20:51 < nsh> there are billions and billions of things that were 'incidental' in making a cow 20:51 < kanzure> true 20:52 < nsh> the same outcome would not occur twice 20:52 < nsh> maybe for very simple organisms 20:52 < nsh> but it's like the first fractals 20:52 < kanzure> sure 20:52 < nsh> mandelbrot noticed that the results of certain nonlinear calculations were different on different machines 20:52 < nsh> sometimes different on the same machine at different times 20:52 < nsh> he was like: WTF 20:52 < nsh> then he realised sensitive dependence on initial conditions 20:53 < nsh> small differences get blown up because the system feeds back into itself information 20:53 < kanzure> sure 21:04 < fenn> i thought you were going to go on about the Kit gene, but then you took a stupid turn 21:04 < fenn> of course you need a tape reader 21:04 < fenn> of course you need a printer 21:05 < fenn> pure information doesn't do anything, so how does the fact that DNA doesn't do (much) anything make it not pure information? 21:42 < kanzure> heybryan.org/sciencedirect_alerts.html <-- masochism 21:45 < nsh> Kit gene, fenn? 21:46 < kanzure> 50 bucks to whoever can figure out the md5 encoding scheme on sciencedirect. 21:48 < nsh> for their pdf names? 21:49 < kanzure> Look at all of their URLs. They always have an md5 variable in them. This variable is required or else the link fails. 21:49 < nsh> http://www.sciencedirect.com/science?_ob=MImg&_imagekey=B6TBX-4HHWWG7-1-3&_cdi=5154&_user=10&_orig=search&_coverDate=04%2F30%2F2006&_sk=999689996&view=c&wchp=dGLzVlz-zSkzk&md5=937c41165fa1c3e34e1ed1ce46ff586c&ie=/sdarticle.pdf 21:49 < nsh> hmm 21:50 < nsh> pretty nontrivial task to guess 21:50 < nsh> impossible if they're using a seed 21:50 < nsh> *salt 21:50 < kanzure> right 21:50 < kanzure> They're generating one for each link though 21:51 < kanzure> and it seems that these md5 sums work over the years 21:51 * nsh nods 21:51 < kanzure> i.e., Google search results from 2005 still work 21:51 < kanzure> so I doubt they are storing all of the md5s they hand out 21:51 < kanzure> instead they're regenerating them 21:51 < nsh> they probably take some concatenation of the title and author and date 21:51 < nsh> and maybe some other fixed secret string 21:51 < kanzure> maybe 21:51 < kanzure> but they use these md5 strings for non-article related pages as well 21:51 < nsh> hrm 21:52 < kanzure> I could just search for the md5 string they give for the links or something, but this makes for a ridiculous amount of downloading of similar pages that really shouldn't matter all that much 21:52 < nsh> that just seems... unnecessarily obfuscational 21:52 < kanzure> yes 21:52 * nsh hates sciencedirect 21:52 < nsh> someone should hax them bad 21:53 * nsh wishes he still had friends who had friends who had 0days 21:54 < kanzure> I'm amazed that this bullshit isn't out on torrents yet 21:55 < kanzure> this md5 system is really sucking 21:55 < nsh> yeah 22:01 < willPow3r> who uses torrents any more? 22:01 < willPow3r> usenet is where it's at 22:05 < nsh> free reliable usenet server? 22:05 < nsh> (and comprehensive) 22:07 < bkero> Usenet is so 1980s. 22:07 < kanzure> I've never used a newsreader. 22:08 * kanzure should give up his geekbadge 22:08 < bkero> You're fired. 22:09 < bkero> You've screwed up too many times, cost the department too much money. 22:09 < bkero> I want your gun and your badge. 22:09 < bkero> You're off the case. 22:13 < kanzure> hm 22:13 < kanzure> according to #md5crack, sciencedirect's md5 hash sums are not valid md5 hash sums 22:19 < nsh> why do you want to be able to create them again? 22:19 < nsh> to avoid downloading linking pages and extracting them? 22:20 < nsh> the latter seems like the path of least resistance 22:20 < nsh> where would you get a url from except a linking page anyway? 22:21 < kanzure> the urls follow a pattern 22:21 < kanzure> if I don't need the md5s, then I don't have to download every single page 22:21 < kanzure> I could just skip to the pages that list the pdfs 22:21 < nsh> what i mean is 22:22 < nsh> where are you getting the information for the final urls (past the pages you want to skip) ? 22:22 < nsh> what's telling you where you want to end up? 22:23 < nsh> wouldn't the page listing the pdfs be dynamically generated from a POST request? 22:23 < kanzure> requiring an MD5 ... 22:23 < nsh> mm 22:23 < kanzure> that I don't have .. 22:23 < nsh> rightright 22:23 < kanzure> (in the POST) 22:23 * nsh nods 22:24 < nsh> but i don't understand that's in between... 22:24 < nsh> *what 22:24 < kanzure> eh 22:24 < kanzure> the md5 is differnet for each volume 22:24 < kanzure> and each edition 22:24 < nsh> give example pls 22:24 < kanzure> and each journal 22:24 < nsh> ok 22:24 < kanzure> ok, give me a sec to get an example 22:24 < kanzure> I'm fighting Gogeta 22:24 < nsh> that's still a whole lot less requests than you'd end up doing for the pdfs themselves 22:25 < nsh> assuming bandwidth isn't an issue 22:25 < nsh> it all seems a bit unnecessaary to avoid crawling 22:29 < kanzure> it's hard to explain, but basically it's annoying that I have to extract the md5 sum from each page 22:29 < kanzure> as opposed to just bruteforcing and trudging through each of the links on my own 22:29 < kanzure> it's just annoying, that's all 22:29 < kanzure> and I'm hungry. I should probably go eat. 22:30 < kanzure> oh, wait 22:31 < nsh> mm 22:42 < kanzure> Hah! "Previous vol/iss " 22:42 < kanzure> what a wonderful link 22:42 < kanzure> problems solved. 22:43 < nsh> ah, awesome 22:44 < nsh> ugh 22:44 < nsh> ventor is an idiot 22:44 < nsh> yes, we're going to grafiti all over genomes leaving little "hi mum" and "this organism brought to you by ibm and nike trainers" 22:44 < nsh> messages 22:45 < nsh> NO, DOES NOT WORK THAT WAY 22:47 < nsh> "First, our studies provide convincing evidence that the genome is pervasively transcribed, such that the majority of its bases can be found in primary transcripts, including non-protein-coding transcripts, and those that extensively overlap one another. " 22:47 < nsh> junk dna does *not* exist 22:48 < Phreedom> nsh: someone's junk is somebody else's treasure? ;) 22:48 < bkero> Then what are Exons? :P 22:49 < nsh> Introns 22:49 < nsh> they're stuff we don't know wot dat it dus yet 22:50 < bkero> Or vestigial? 22:50 < nsh> vestigial doesn't really work on the gene scale 22:50 < nsh> it's more like "backup" 22:50 < nsh> you could say that all the skills i learnt at school that haven't used this week are vestigial 22:50 < bkero> Then what about retroviruses? 22:51 < nsh> you mean that they serve no purpose for the organism? 22:51 < bkero> sure 22:51 < nsh> everything in DNA does *something* 22:51 < nsh> and not very much of it is for the same end 22:52 < nsh> in fact, i'd venture that the stuff that is cooperative is so only incidentally, and begrundingly 22:52 < nsh> *udgingly 22:52 < nsh> it's like a society 22:52 < bkero> But if a retrovirus sneaks itself into our genome, sometimes it won't do anything until it's activated by something else. 22:53 < nsh> who goes to work to keep the economy going? no-one, you go to work to get a wage, etc. 22:53 < nsh> oh right 22:53 < nsh> sure there's dormancy 22:53 < nsh> but it's not final 22:53 < nsh> so i guess you could use some steganography 22:53 < bkero> It could be if there's nothing to activate it anymore. 22:53 < nsh> nah, there's loads of mechanisms whereby partial information could be reincorporated 22:54 < nsh> 'random' recombination, transcription errors, etc. 22:54 < nsh> there's no black hole for information in biological systems 22:54 < nsh> is what i'm trying to get across 22:55 < nsh> it's very hard to lose information on such a basic level 22:55 < bkero> You're saying that everything in a genome does something. 22:55 * nsh nods 22:55 < nsh> it's like physics, there's no microscopic variable that doesn't input back into the system 22:56 < bkero> What about if some horizontal transfer takes place and cuts off part of the genome? 22:56 < nsh> gene level extinction? 22:56 < nsh> sure, that must happen, have happened a lot 22:56 < nsh> depending on the degree of redundancy 22:56 < nsh> i mean, things must be selected against 22:56 < nsh> for things to be selected for 22:57 < nsh> but then you still can't permanently mark anything genomically 22:57 < bkero> But what if the transfer only cut off a part of the genome, and did so in more than 1 host? 22:57 < nsh> not sure i follow, sorry 22:58 < bkero> Essentially a 10: GOTO 30; 20: DO_SHIT; 30: DO_MORE_SHIT; 22:58 < nsh> ok, but that's a static piece of code 22:58 < nsh> if there are other bits of code that are fucking around with that code 22:59 < nsh> then 20 will get/may get unskipped at some point 22:59 < nsh> perhaps is asymptopically likely to 22:59 < nsh> *totically 23:00 < bkero> Heh, I've got a meeting to go to. I get what you're saying. 23:00 * nsh nods 23:00 < nsh> enjoy meeting! (fat chance ;-) 23:00 < bkero> 1:1 meeting with boss 23:00 * bkero should take some biology/microbio/genomics classes 23:01 * nsh should take some * classes 23:02 < nsh> meh 23:02 < nsh> they're both not doing well at this argument (dawkins and venter) 23:02 < nsh> but i think venter is right 23:03 < nsh> virally introduced genes could easily (i would say must) have been appropriated into (for example) mammalian biochemical pathways 23:03 < nsh> making it much more difficult to give tree (acyclic directed graph) taxonomic structure to gene evolution 23:04 < nsh> i would say that the evolutionary path of genes in multiply connected and to a certain degree re-entrant 23:04 < nsh> *is 23:04 < nsh> (certain degree because genes {we reasonably safely} assume can't travel back in time and be incorporated into organisms no longer extant) 23:04 < nsh> s/{we /we { 23:04 < nsh> / 23:06 * bkero hasn't taken any science classes in 5 years. 23:07 * bkero needs to SCIENCE more. 23:09 < nsh> yay SCIENCE!! 23:12 * bkero is a comptuer science major and the only biology information he has was gleaned off of random things. 23:12 * bkero is haxing himself some neuroscience though. 23:14 < nsh> computer science and neuroscience has been 'on the cusp' of convergence for ever now, so i guess it's a good combination of things to know about 23:14 < nsh> personally, i'm dubious 23:14 < bkero> Fucking with brains is dangerous. 23:14 < nsh> but that's because i'm an olympic standard cynic 23:14 < bkero> And irreversible. 23:14 * nsh nods 23:14 < nsh> quite 23:15 < nsh> and i don't think our model of computatation is sophisticated enough to even begin understanding or modelling neurology computationally 23:15 < bkero> I'm a cynic when it comes to things reaching me. 23:15 < nsh> i don't think we even have the math yet 23:15 < bkero> Or things materializing. 23:15 * nsh smiles 23:16 < bkero> I also argue points for sides I'm completely against/don't think is true simply for the sake of hardening the other sides arguments. ;) 23:16 < bkero> See: earlier about the exons. 23:17 < bkero> Is more research done to find data behind a reason, or a reason behind data? 23:17 < fenn> nsh see (first google result) 23:17 < fenn> http://www.cosmosmagazine.com/news/288/mutant-mice-challenge-heredity-laws 23:19 < fenn> what's venter arguing? that horizontal gene transfer is real? of course it's real, we do it all the time 23:40 < nsh> dawkins said that transgenics runs the risk of making molecular taxonomical classification of genes intractable for future scientists 23:40 < nsh> venter said that we were quite unlikely to shift more genes around than already occurs naturally 23:40 < nsh> and dawkins seemed to think that horizonal gene transfer was actually quite rare 23:41 < nsh> and that the evolutionary history of most genes is by descent 23:42 < nsh> dawkins writes about the argument here, it seems: http://richarddawkins.net/article,2287,DLD08---Life-a-gene-centric-view,Richard-Dawkins-Craig-Venter 23:42 < nsh> (haven't read yet) 23:43 < nsh> oh no 23:43 < nsh> just random internewbs discussing it 23:49 < nsh> [[[ 23:49 < nsh> But now, Ouzounis and his team of researchers have discovered that microbes can do this on a horizontal level, passing on the genes that can lead to antibiotic resistance, for example. By using a method called GeneTrace, a technique developed by Victor Kunin, the team observed more than 600,000 vertical transfers, coupled with 90,000 gene loss events and approximately 40,000 horizontal gene transfers. 23:49 < nsh> This does not mean that the ‘tree’ becomes conceptually redundant, as it forms the framework or skeleton for Ouzounis and his team’s research. It just makes it far more complicated than was previously thought. “We used these trees as the scaffold of the net, on which we looked for the evidence of horizontally transferred genes," explains Kunin, previously a PhD student in Christos Ouzounis's group. 23:49 < nsh> ]]] - http://www.scienceagogo.com/news/gene_transfer_wrong.shtml 23:49 < nsh> good analogy 23:51 < kanzure> screw computational modelling of brains 23:51 < kanzure> I don't need mind uploading and neither do you 23:52 < Phreedom> you have nothing to upload? ;P 23:53 < Phreedom> modelling is very useful for many tasks... not sure about mind uploading though 23:53 < kanzure> yeah, nothing left to upload .. i've been out of my mind since the beginning 23:53 < kanzure> get with the program 23:53 < kanzure> something like that. 23:54 * kanzure goes to hunt for food 23:58 * bkero also hunts for food. 23:58 * bkero stabs kanzure and puts him over a fire.