2008-04-26.log

--- Day changed Sat Apr 26 2008
fennhttp://everything2.com/e2node/Santa%2520Machine00:22
kanzurenote the ref to Damien Broderick00:23
kanzureDamien and I have been talking recently.00:23
fennyes, have you read that book>00:24
kanzureactually, no00:24
kanzurebut I probably should :)00:24
kanzureI also see a Hamilton ref and Stephenson ref00:24
kanzureI've read Hamilton's Mindstar story, it started off with some good hard scifi, though not quite hard enough for my tastes (it could have been harder, more edge and cut sort of thing)00:24
kanzureshould I bother writing something on that everything2 node?00:25
kanzureor just get back to reading up on git00:25
fenne2 has an even more strict and irrational editorial policy than wikipedia00:25
fennthey seem to be proud of their literary consistence above all else00:26
fennconsistency00:26
fennso you can't just add a line here or there00:26
kanzureit looks like you can add a few paragraphs that are 'self-referential' or of the same theme/style and be good00:27
kanzurebut I haven't recursed through e2 explicitly before.00:27
fenndont try, it will take a couple lifetimes00:29
kanzurehttp://www.ugrad.physics.mcgill.ca/wiki/index.php/Other_Courses weird amount of content00:35
fennhyperphysics is good for just about anything physics00:38
fennif you want to know formulas and their derivation00:38
kanzureyep00:38
kanzure'You may notice that the output here looks very much like the first portion of the output from "git pull". This is no coincidence. The new changes have been "fetched" into the current repository and are stored into "origin/master" and have not been into the current "master" branch. Remember that "master" is our current branch. So now, "origin/master" is the state of the master branch that exists in the "origin" repository, (the on00:45
kanzureugh?00:46
kanzure@ http://cworth.org/hgbook-git/tour/00:46
kanzureI think that's just saying "this is the same repo we originally cloned, not the one with the updates"00:48
kanzurehm, perhaps not00:48
fennduh.. pardon me but why is mercurial documentation in a git repo?00:49
kanzureas an example, perhaps00:49
fenndid it change names?00:49
fennok, so mercurial is definitely not git :)00:50
fenni had to look that up00:50
fenni guess he just used the mercurial tutorial as a template00:52
kanzureeven after reading the tutorial some things are still hazy00:56
kanzurefor example, suppose I git-clone a repo00:56
kanzurethen I do some changes, and then spawn my own local branch,00:56
kanzureand then want to commit my spawned branch in the main repo that I originally got the first from.00:56
kanzurehow ? 00:56
fenngit push origin branchname ?01:11
fennassuming you want the branch to show up as a branch in the main repo01:11
fennbut usually it seems people use pull more than push01:14
fenni.e. there's no "main repo"01:14
fenni dont understand why they use the extended ref-spec here: "git push origin devel-something:refs/heads/devel-something"01:16
kanzurewhat is the use of a 'extended ref-spec' or what is it supposed to be?01:22
fennif you want to push your changes to a different branch01:23
fennworks when i do it my way; seems to work their way but i dont see it in git branch -r 01:33
kanzurehttp://heybryan.org/mediawiki/index.php/2008-04-25#News <-- me did a news dump01:46
kanzurere: supermemo http://heybryan.org/shots/March14th02007.PNG - screenshot of my use of it01:57
kanzurehttp://supermemo.com/01:57
kanzurehttp://supermemo.com/articles/genius.htm - gets most things right, but is missing a few key points that I've learned since I last read it, 2006?01:58
kanzureit's in the news under the heading "Never forget anything again" ehhh01:59
kanzurewoo02:01
kanzurehttp://en.wikipedia.org/wiki/Taxonomy_of_manufacturing_processes02:01
fenni was pretty surprised when i saw these graphs (which i had arrived at independently): http://www.supermemo.com/articles/sleepchart.htm02:02
kanzurewish I would have read that before trying out polyphasic sleep in 200502:03
kanzureit was during spring break, so I lost some good, useful time02:03
fenneh? it doesnt say anything about polyphasic sleep02:04
kanzureoh02:04
kanzurewrong article, hold on02:04
kanzurehttp://www.supermemo.com/articles/polyphasic.htm02:04
kanzurehm, from that page - 'One minute of insight can be worth a century of shoveling!'02:06
fenni'd like to see some scientific studies with more than one person's experience02:06
kanzurehttp://heybryan.org/thinking.html re: insight02:06
kanzuresure, he has references in there02:06
fenn"Anything against the free running rhythm is bound to result in mental torture and cognitive deficit"  <- that's for damn sure02:08
kanzuresounds like school02:08
fennmy thoughts exactly02:08
fennyou'd think there would be all sorts of weird sleep-pattern-inducing drugs out there02:11
fennbeyond just stimulants and sleeping pills02:11
kanzureI don't know if we know enough about sleep to understand that02:14
kanzurehowever, I can definitely tell when I am "out of it"02:14
kanzurehm, speaking of which, now might be some good time for me to fetch some sleep02:15
fennhah02:15
kanzure3D real-time spatiotemporal dendritic voltages could be used to recreate dendritic objects a.k.a. perceptions in other ways - 53:50 in the video; hm, I guess he didn't really talk about what I was thinking about02:33
kanzurebut it's definitely unspoken re: aggregation, spill-over, and top-down rule-based management or organization of the aggregating layers of the brain towards strategies and approaches to living/thinking/etc., 02:33
kanzure1999 group, IIRC, was able to hack into a cat visual cortex to retrieve visual data that was going through the eye (a television, actually)02:34
kanzurebut I think they were doing spikes02:35
kanzurewhich doesn't make too much sense02:35
kanzuremaybe they were doing patch-clamp, but that's destructive, no?02:35
kanzureoh02:38
kanzuredistance between synapses - spillover from one input is almost impossible to avoid, to another one. Within 200 nm there's another synapse that is related to a different neuron. What the brain is trying to do is control the voltage within a certain voxel compartment - not so much a particular neuron. This would be an attempt to control pain/energy/action. Electromagnetic dendritic objects. We do not see the world. What we do is we 02:38
kanzureso the brain is overall trying to control the voltage within a minicolumn02:39
kanzurethere we go02:39
kanzureso what would voltage controlling theoretics say about that02:39
kanzurein electronics02:40
kanzurewould this imply that the top-down management of the brain procs are there to do voltage minimization/maximization to get the systems to 'equilibriate' ? would that make sense, from an electronics point of view?02:42
-!- marainein [n=marainei@220-253-10-191.VIC.netspace.net.au] has quit ["Ex-Chat"]02:44
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has quit ["Leaving."]02:46
-!- marainein [n=marainei@220-253-10-191.VIC.netspace.net.au] has joined #hplusroadmap04:10
-!- Splicer2 [n=p@h17n1c1o261.bredband.skanova.com] has joined #hplusroadmap05:51
-!- Splicer [n=p@h100n2c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]06:07
-!- Splicer [n=p@h44n3c1o261.bredband.skanova.com] has joined #hplusroadmap07:04
-!- Splicer2 [n=p@h17n1c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]07:19
-!- Splicer2 [n=p@h44n3c1o261.bredband.skanova.com] has joined #hplusroadmap07:19
-!- marainein [n=marainei@220-253-10-191.VIC.netspace.net.au] has quit ["Ex-Chat"]07:20
-!- Splicer [n=p@h44n3c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]07:26
-!- Splicer [n=p@h70n1c1o261.bredband.skanova.com] has joined #hplusroadmap07:49
-!- Splicer2 [n=p@h44n3c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]07:59
-!- Splicer2 [n=p@h70n1c1o261.bredband.skanova.com] has joined #hplusroadmap08:00
-!- Splicer [n=p@h70n1c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]08:12
-!- Splicer [n=p@h90n3c1o261.bredband.skanova.com] has joined #hplusroadmap08:15
-!- Splicer2 [n=p@h70n1c1o261.bredband.skanova.com] has quit [Read error: 104 (Connection reset by peer)]08:17
-!- Splicer2 [n=p@h86n3c1o261.bredband.skanova.com] has joined #hplusroadmap11:07
-!- Splicer [n=p@h90n3c1o261.bredband.skanova.com] has quit [Read error: 110 (Connection timed out)]11:19
-!- Splicer2 [n=p@h86n3c1o261.bredband.skanova.com] has quit []11:42
-!- kanzure [n=bryan@cpe-70-113-54-112.austin.res.rr.com] has joined #hplusroadmap14:10
-!- krebs [n=krebs@208.106.116.196] has joined #hplusroadmap15:46
-!- krebs_ [n=krebs@208.106.116.196] has quit [Read error: 104 (Connection reset by peer)]15:46
-!- marainein [n=marainei@220-253-155-121.VIC.netspace.net.au] has joined #hplusroadmap15:50
kanzurehttp://heybryan.org/shots/img_8028.jpg - my other shelf, mostly scifi and programming, but this scifi is old stuff - it doesn't include my 'new' (since 2005?) stacks16:41
maraineinbooks, books16:45
marainein"* for dummies"...:P16:46
kanzurehehe16:51
maraineinthey're not bad...I have an ebook of bioinformatics for dummies16:53
kanzureawesome :-)16:54
kanzureMost of my texts are in16:54
kanzurehttp://heybryan.org/shots/img_8032.jpg16:54
kanzureand then a lot that are digital ...16:54
maraineinI think that's my first year biology book17:03
kanzureWhat's that firefox plugin that lets you analyze flash? To see if it's loading anything from the site?17:07
kanzureflashtracer?17:08
maraineinI didn't know there was one17:08
kanzurehttp://www.refactoring.com/17:49
kanzureI think this is everything for agx-get.py - http://heybryan.org/mediawiki/index.php/2008-04-2618:20
kanzureexcept that I wasn't writing in python, oops18:20
kanzurehttp://en.wikipedia.org/wiki/Captain_Copyright <- This is the best thing I ever read. Too bad it was Canada and not the US.18:24
kanzureHeh, Paul is good -18:29
kanzureOr (paraphrased: :-)18:29
kanzurehttp://www.imdb.com/title/tt0076759/quotes18:29
kanzure"""18:29
kanzureGold Leader: Pardon me for asking, sir, but what good are semantic wikis and18:29
kanzuredesktops going to be against Virgle?18:29
kanzureGeneral Dodonna: Well, the Empire doesn't consider a small cgi script on a18:29
kanzureshared server or desktop to be any threat, or they'd have a tighter defense.18:29
kanzure,,,18:29
kanzureCommander #1: We've analyzed their attack on Knol, sir, and there is a18:29
kanzuredanger. Should I have your Golden Parachute standing by?18:29
kanzureGovernor Schmidt: Evacuate? In our moment of triumph? I think you18:29
kanzureoverestimate their chances.18:29
kanzure"""18:29
* fenn paddles ferociously in the torrent. trying to keep up18:34
fenn"the configuration18:35
fennscript"18:35
fenndoes what exactly?18:35
kanzureit's bad to call it a configuration script18:41
kanzurecall it a 'runtime script'18:41
kanzureagx-get lets you use parameters,18:41
kanzurethe default behavior is to download the metadata file18:42
kanzureanything else would mean downloading a portion of the skdb git repo18:42
kanzure(for that particular project)18:42
kanzuredifferent 'runtime scripts' for different 'objects' or whatever will let you specify different things18:42
kanzurelike comfiness ;-)18:42
kanzurebut I recall that being a bad example18:42
-!- marainein [n=marainei@220-253-155-121.VIC.netspace.net.au] has quit ["Ex-Chat"]19:10
fennjust had an idea for a "rapid prototyping" process: get a package of barbecue skewers, feed them into your cnc chop-saw which cuts them to length based on a simple 3d mesh representing the object you want20:12
fennthen you drill holes into 'gumdrops' some roughly spherical shape that's cheap and commonly available20:13
fennand the computer gives you instructions to put them together to reconstruct the 3d mesh20:13
kanzurehm? barbeque skewers are what?20:13
kanzurewhile I might live in Texas, I try to avoid barbeques20:14
fenna readily available cheap package of pointy bamboo rods20:14
fennabout 14 inches long 1/8" wide20:14
fennthe gumdrops kinda mess the process up, maybe using straws and bolting/riveting them together would work better (like an an electrical conduit dome)20:15
fennyou could make straws from a continuous roll of tape, somehow20:16
fennit would be like those lawn ornament wire deer people put up with christmas lights20:17
kanzure#git is confusing me ... if you clone a git repo, do some updates, commit, then push, but within the time that it took you to do all of that, somebody *else* cloned a git repo, you can't push your changes to the main repo since it's "checked out"20:18
kanzurebut wtf is this useful for if you can't put your code changes back into the main repo20:18
fennchecked out? what does that mean? (sounds like some ancient lock-based rcs)20:19
kanzurethat's what #git is saying20:19
kanzure"checked out"20:19
kanzureI certainly didn't use that term with 'em.20:19
fenni think that phrase only applies to using a specific branch20:20
fennor messing with cvs20:20
fennaha - you mentioned 'checked out' first20:23
kanzuredid i?20:24
fennhttp://colabti.org/irclogger/irclogger_log/git?date=2008-04-2720:24
kanzure00:1520:24
kanzurethat's not me20:25
fennoh oops. do you understand the difference between a branch and a cloned repo?20:25
kanzurenot yet20:26
kanzureplease help :-(20:26
fennok, a branch is like, say you undo a bunch of changes and then start editing an old copy20:26
fennnow you're moving along a different timeline20:26
kanzurebut isn't that a repo20:27
kanzureyou just cloned the "old version"20:27
fennthat's a branch20:27
fennboth timelines are recorded in a repo20:27
fenntimeline == branch20:27
kanzureokay, so if you have version 4 at the moment, and I clone from version 2, does my version 2 repo have version 4 in it somewhere ?20:27
kanzurehm20:28
fennum.. forget about cloning 'version 2'20:28
kanzurebut you just said20:28
kanzure'undoing a bunch of changes'w20:28
kanzureso if your timeline has A, B, C, D (current)20:28
kanzureand you clone B20:28
kanzureeh20:28
fennyeah, so your working directory is showing an old version20:28
fennyou can't clone B20:28
kanzureoh?20:28
fenn(not without making thing harder to understand)20:28
fenni can hit you with a ruler if you like :)20:29
kanzureokay, so you can *branch* from an old version, but you can't clone an old version20:29
fennclone means get the whole repo, which includes all the branches20:29
fennthe closest thing to a 'main repo' is some branch that you intend to be distributed widely20:31
kanzurebut then why were they talking about 'checking out' ?20:32
kanzurehrme.fjkdlfjaldkfjal20:32
fennthe neat thing is that all the changes on development branches that get included in the distribution branch will show up there as if they were just edited on the distribution branch20:32
fenn'check out' means, 'show me this branch (at this version) in the working directory'20:32
fennits a bad analogy i think20:33
kanzurebut they said locks20:33
kanzurewait20:33
kanzureaha20:33
kanzurethey didn't20:33
kanzurethat's the difference20:33
kanzureI thought they were saying something was locked20:33
kanzurebut in this model you don't "lock a branch down"20:33
fennits not locks exactly20:34
kanzure(2008-04-26 19:29:30) jdq: I would suggest only to push to bare repostories and update those with working copies only by pulling20:34
kanzurethat strategy sounds like a good idea20:34
fennsee, you dont normally push to someone else's repository, they will pull from you if they want to, and then if there's problems with the merge they get fixed when the person is pulling20:34
kanzureright20:35
fennotherwise you're throwing monkeywrenches on their workbench20:35
fennyou could rebase your changes on the current version of the repo you're trying to push to (if you dont want to obligate the recipient to fix merge errors)20:36
kanzurebut the recipient might have other changes of his own of course20:36
kanzureoh20:36
kanzurenevermind20:36
kanzurehe could then pull those changes (or you can push to him)20:36
kanzurebut that's no guarantee that his modifications don't conflict with yours20:36
kanzureso rebasing doesn't sound possible, unless the developer hasn't done anything to the files20:36
kanzurehm20:36
kanzurebut that's not much of an issue20:36
kanzureI can live with that20:36
fennyou'd still test your changes after rebasing of course20:37
fennin a large percentage of changes it will work20:37
fennit's not magic20:37
fenni.e. it wont re-base semantic changes20:38
kanzurehttp://neuralensemble.org/20:41
kanzure"Increasingly, the real limit on what computational scientists can accomplish is how quickly and reliably they can translate their ideas into working code."20:41
kanzureGregory V. Wilson20:41
fenni wonder when that was written20:43
kanzurehttp://www.americanscientist.org/template/AssetDetail/assetid/48548?&print=yes20:43
kanzure200620:43
kanzure'I therefore started asking scientists how they wrote their programs. The answers were sobering. Whereas a few knew more than most of the commercial software developers I'd worked with, the overwhelming majority were still using ancient text editors like Vi and Notepad, sharing files with colleagues by emailing them around and testing by, well, actually, not testing their programs systematically at all.'20:44
kanzure'I finally asked a friend who was pursuing a doctorate in particle physics why he insisted on doing everything the hard way. Why not use an integrated development environment with a symbolic debugger? Why not write unit tests? Why not use a version-control system? His answer was, "What's a version-control system?"'20:44
kanzurehahah20:44
fennhm i think i've read the 'computational literacy' project that guy started 20:45
fennits like, programming for scientists20:45
kanzure'20:45
kanzureMy friend was intelligent and intimately familiar with the problems of writing large programs—he had inherited more than 100,000 lines of computer code and had already added 20,000 more. Discovering that he didn't even know what version control meant was like finding a chemist who didn't realize she needed to clean her test tubes between experiments. It wasn't a happy conversation for him either. Halfway through my explanation, 20:45
kanzureah, good20:45
kanzurebecause that's retarded20:45
kanzuresomebody writing 100k lines of code and not knowing about rcs?20:45
kanzurehttp://en.wikipedia.org/wiki/Gregory_V._Wilson20:47
kanzureOram, Andy & Wilson, Gregory V. (2007), Beautiful Code: Leading Programmers Explain How They Think, O'Reilly Media, Inc., ISBN 978-0596510046 20:48
fennoh look, a reference to some dead tree20:48
fennhow quaint20:48
fennthe whole 'comutational scientist' thing just screams 'anachronism'20:49
kanzureI think the problem is this:20:49
fennah right, software carpentry20:50
kanzurerapid advancements on the edge, versus the tendency to 'preserve integrity' of the already established Scientific Institutions and The Way20:50
fennMy Way or the High Way20:50
kanzureunfortunately, 'My Way' is usually also some highway like road infrastructure20:51
kanzurewith predetermined destination20:51
fennsoftware carpentry is actually pretty good20:52
kanzurehm?20:52
kanzureexplain what you're on about ?20:52
fennhttp://swc.scipy.org/20:52
fennthe result of g. wilson's "shocking revelation"20:53
fennkinda makes you wonder just what they're teaching in schools20:53
kanzurehow to bend over.20:54
fennLOL "Knowing how to read code is as useful as knowing how to read a proof"20:55
kanzurewhy is that funny20:55
fennsorry, it's only funny if you dont believe in proofs20:55
fennsee, the difference between code and proofs is that code actually does something20:56
kanzureeh, 20:56
kanzuresort of20:56
fennoh come on, math theorems that cite proofs are just doing a glorified include statement20:56
kanzureit's s/proof/ with ... koan, theorem, poem, something that is itself not solvable, but somehow allows you to work over it so that you can figure out some new context20:56
kanzureor whatever20:57
kanzuresure20:57
kanzurethe real difference, I think20:57
kanzureis that the 'advanced' mathematics theorems are more like code that hasn't been written yet20:57
kanzureit's formalized pseudocode or something20:57
kanzuresome of the ancient recursion examples with series for example20:57
kanzurethey didn't have code to write20:57
kanzureso instead they had that formalized system20:57
kanzureand locked it in via education, 20:57
fennyes20:57
kanzurebut now that we can do it in code, I suggest we stop doing it20:57
kanzureteaching it, I mean20:58
fennand the syntax is horrible, unfortunately20:58
kanzuretell them about it, sure,but show the equivalency20:58
kanzureand show how the machine is able to do this20:58
kanzureit's quite formalized.20:58
kanzureand *grounded*20:58
kanzurere: symbol grounding20:58
kanzurethis is a recent insight that I've been toying with20:58
kanzurethe symbol grounding problem in ai isn't ai-specific20:58
fennmath is grounded?20:58
kanzureit's completely relevant to all humans - we're a walking, talking example of symbol grounding20:58
kanzuremath is grounded in *us*20:59
kanzurein mathematicians20:59
* fenn looks around for a number20:59
kanzureor people becoming mathematicians20:59
fenni dont see any numbers20:59
kanzuremath has a social context20:59
kanzureI don't mean to say that math is ideological20:59
kanzurebut that math is a formalization and is grounded in society20:59
kanzurethis formalization isn't Of The Divine Void20:59
fennit's very hard to convince some people of that21:00
kanzuredo you agree with it, though?21:00
fennyes of course21:00
kanzureblah21:01
kanzurethis is just the basic strand of thought re: the need for skdb and so on21:01
fenni think the whole symbol grounding thing has to do with perception and procedural learning21:02
fennif you can take a basket of apples and count them, there's a number. if your culture just says 'many' then there is no number21:02
kanzureright, because counting and numbering are mental processes, physically grounded in your brain circuits that you develop to solve problems21:03
kanzureusually problems that teachers present to you in, in the case of school; or on your own when you sit down to focus on a problem21:03
fenncounting is an application of generalization21:04
fennyou dont learn generalization in school21:04
fenngeneralization is taking all the lumpy green and red things in front of you and realizing they are all the same abstract class21:05
kanzureOOPism is the dark side, it leads to stuff like string theory21:06
kanzureand then my Holy Ai Church that I was developing ... erm.21:06
fennis an apple not an apple?21:06
kanzureno, truly it is not21:06
fennthen you refute number theory too?21:06
kanzurean apple is *physically there*21:06
kanzureyou can go touch it, eat it, and I can't possibly elaborate all of the possible interactions with the apple21:07
fennyou can do that with any other apple21:07
fenn(unless it's an apple that looks like jesus)21:07
kanzureso if the apple has jesus mapped on it21:07
kanzureit is therefore not an apple now?21:07
kanzurewhat the hell type of reasoning is that? :-)21:08
fennit's a special case apple21:08
fennyou can do extra things with it, like sell it on ebay21:08
kanzureI am more a programmer; more interested in processes, not 'object orientation'. I can't link you to how I jumped out of OOPism in my development, so I can't cite the full philosophical arguments, but I'll think about it so that I can describe the inherent problems in OOPism to you.21:10
fennyeah it's limiting but it's easy21:11
fennc++ made a lot of stupid mistakes that make things hard21:11
kanzureno, I don't even mean C++21:11
fennwell, a lot of what goes in the definition of "oop" is from languages like C++ and java21:11
kanzureAh, well, that's not what I am talking about21:12
fennok21:12
fennmaybe it's better to define what you mean then21:12
kanzureOOPism, to me, is the philosophy of objects out there in the world, that all have these 'classes' inherent to them, they are of Some Type, and they are all intricately related to, aha. I just remembered a diagram I made back in 2004 to describe my Holy Ai system.21:12
kanzurehttp://heybryan.org/diagrams/arrow_of_time_system3.JPG21:13
kanzurehm21:13
fenndid you watch the geoff hinton video?21:13
kanzurehttp://heybryan.org/projects/aot_system/arrow_of_time_system3.JPG21:13
kanzureno, not yet21:13
fennheh so your ai will be done in a couple billion years by that timeline :)21:13
kanzure'Objects' just aren't true. Don't see it as an apple, but rather the seed of a tree, as the Salthe-ecological gravitational thermo complex that is aggregating materials into that redness/greenness, an attempt in mutation history towards evolvability in whatever environment that tree was competing in etc.21:15
kanzureerm, that uses way too many loosely used terms;21:15
fennwell the reason i brought up geoff hinton is his self-programmed algorithms cluster raw data into regions automatically21:15
kanzurematter/energy signals, processes, not objects21:15
fennthey arent equivalent, but they are in discrete clumps21:16
kanzurethink of the galaxies as a matter/energy processing system21:16
kanzurean 'ecology'21:16
fennok, so an apple is a process21:16
kanzureright, so we have our biologists go out there and attempt to classify them or whatever21:16
fenni'm sure you'll get uppity about my use of the term "is"21:16
kanzuregood luck with that whole 'tree of life' thing21:16
kanzurebut I really like those trees of course, they are interesting21:17
fenni think it's just english language bias (sapir-whorf)21:18
kanzurethe point isn't to find a 'true tree of life' - that sort of information is very, very hard to find from evolutionary history, without self-documenting organisms of course21:18
kanzuresapir-whorf was even cited in my intro to speech class, the state mandated class in high school21:18
fennthe point isnt to recover the version history?21:19
fennoh, well.. taxonomy is different from phylogeny21:19
fenn(i think taxonomy is bullshit)21:19
kanzure'version history' would be great -- too bad the biosphere didn't spawn with 'git' installed by default21:19
kanzurebut anyway, we're getting off track here21:20
kanzurewhile you might be a process that is defined by who your ancestors are, your environment also influences you to some extent, no? Are we to track all contexts that have ever existed?21:20
kanzuresometimes I find myself imagining going on this very big quest21:20
kanzureto try to recover all of the data21:20
kanzurethat could help me figure out the 'original state'21:20
kanzurethe seed data that spawned the universe21:20
kanzurethe idea is that the universe is stochastic, thus why we have entropy and so on21:21
kanzure(entropy is just "energy not used" and unnecessary noise etc.) 21:21
kanzureso if I could go sample enough data and figure out the original state, then maybe we could simulate specific aspects or something21:21
kanzurebut I don't really place much on this idea any more21:21
fenndoesnt stochastic mean non-predictable?21:22
kanzureno?21:22
kanzurestochastic != random21:22
fennsomewhere between random and deterministic21:22
kanzuredeterministic rules with random seed21:22
fennrandom seed is a computer thing21:22
fennbecause they didnt have hardware RN generators21:22
fennits a way to make a deterministic system appear random21:23
fenns/random/stochastic/21:23
fenni think its just another way of saying 'probabilistic'21:24
kanzurehm, I have an idea21:24
kanzureNG: Would it be accurate to say that a good deal of the thematic burden of the Neverness novels, especially The Wild and War in Heaven, is the necessity that we prefer our actual, physical environment over virtual realities, no matter how beguiling?21:25
kanzureDZ: I would say that is exactly true. And more, I would say that the so-called virtual realities are misnamed: they should be called something like21:25
kanzure "simulated experiences." Because they aren't real, and can never be so, any more than a map can be the territory. And more, for the same reason that a map is necessarily less detailed 21:25
kanzurethan the territory that it describes, a virtual reality can only ever be a pale shadow of the real thing. 21:25
kanzureSuch constructs might prove amusing, or even useful and illuminating, but how could they ever take the place of the essential reality that they represent?21:25
fenneh, i guess21:26
kanzureoh, btw,21:26
kanzurethe guys in #git recommend *against* partial downloading of git repos21:26
fennyeah i know21:27
kanzureso then. the whole idea of agx-get only getting parts of an skdb file is a bad idea, it seems21:27
kanzureunless skdb files != their own git repos.21:27
fennthat might work, but it neglects relationships between package21:27
fennpackages*21:27
kanzurehow so?21:27
fennsay package A v1.0 only works with package B version 1.221:28
kanzureso? the metadata for A would say that21:28
fennright21:28
kanzurewe're talking about whether or not the dot skdb file is secretly a dot git21:28
kanzureand now we're just saying it's secretly a dir or secretly a dot tar21:28
kanzureI guess this is okay21:28
kanzuredoesn't change much.21:28
kanzuremakes it easier actually21:28
fenndo users really need to get the whole devel history?21:28
kanzurenope, they just want the current skdb package 21:29
kanzureand if they want to make modifications or something, go right ahead21:29
kanzurebut you might have to merge with an old branch or something21:29
fenni think it makes sense to have the whole SKDB in one big git repo21:29
fennthen you can track where someone copied code from your module and modified it21:29
kanzurebut you can't have gits within git21:29
fennthen you update your module, you can say 'hey better fix those other modules too'21:30
kanzureand you can't easily partially git a part of that 'global git SKDB'21:30
kanzureso everybody everywhere has to download the full thing21:30
kanzurenah, I think it's okay to make this an aggregation layer21:30
kanzurewe're fine, nothing is different21:30
kanzureI'm just stalling21:31
kanzureso that I don't have to ALT+TAB over to the window with the python code21:31
fennyou can download single files from a git repo easy enough21:31
fennheh21:31
fennprocrastination is my favorite pastime21:31
kanzureI think I'm quoted on wikiquotes re: procrastination21:31
kanzureyou can procrastinate later or something21:31
fennso you associate OOP with neurotypical "windowpane" mentality21:39
kanzurewell, yes, but I didn't think of it like that before21:39
kanzurebut I'm confident that yes lots of 'neurotypical' dopamine addicts like OOPism21:40
fennmakes sense21:41
kanzurebut the thing is that even I loved OOPism21:41
kanzurefor years ... so. 21:41
fennoh, that was the original point of reciprocality.org21:41
kanzuremappers v. packers21:41
kanzureor something21:41
fennhe was a program-management consultant21:41
kanzure"This project began as a bit of practical industrial psychology, and ended up unfolding into an understanding of how most people in most human societies have a consistently distorted view of everything. Not everyone is caught in the confusion, and as the picture emerged, 21:43
kanzurean alternative model of relationships between observable phenomena that seems to be experienced by creative programmers in software engineering, star diagnosticians in medicine, great physicists and mathematicians, so-called ADHD children, people who "Know Quality" in industry, 21:43
kanzurepoets, painters, sculptors and mystics became describable - but only in its own terms. The alternative picture is wholly rational, but not reductionistic. Best of all, it is scientifically grounded and experimentally testable. If the experiments fail we can junk it. If they work, we've learned something important. " - Alan G Carter21:43
fennsee, the problem is that not 'blackboxing' stuff by putting it into categories, means you have to _think_ which is actually _hard_21:44
kanzureblackboxing doesn't require categories per-se,21:45
kanzureblackboxing is about blackboxing functionality21:45
kanzurenot about blackboxing objects (layering etc.)21:45
kanzurehttp://web.archive.org/web/20060427111402/www.reciprocality.org/Reciprocality/add.html21:45
fenni wonder whats up with the website.. why dont they just redirect to one of the many mirrors21:46
kanzure' This was at variance with usual thinking in the field, which holds that most people are simply useless by nature. ' haha :-) And for good reason too :-)21:46
kanzure' It seemed to me more likely that everyone was born a mapper, but somehow most people got "flipped" into the weaker packer mindset through social pressure. It seemed that packers were distressed by any reference - even implicit - to the mapper worldview. Yet this was a denied, neurotic kind of distress rather than an explicit disagreement about ideology.'21:47
fennits funny that he falls into the trap of categorizing people as one thing or another (maybe ignoring some aspects of reality on purpose)21:49
fennpresumably he's thinking about a few specific people he knows21:50
kanzureright21:51
kanzurethe categorization though is linked to neuroscience in terms of the dopamine inhibitors and cybernetic feedback theories21:51
kanzurebut21:51
fennthis sorta reminds me of how its better for you to eat fresh vegetables, but it's actually _easier_ to get/store/trade dried beans and preserved food21:51
kanzurehow so?21:52
fennand so agriculturalists have taken over the world21:52
kanzurehm21:52
fennbecause they grow food that can be preserved21:52
fennobviously some preservation methods are better than others21:53
kanzureI'm trying to think of a way to address your issue with him classifying the people21:53
kanzureI think that in this case, he can classify people in that OOPist sense21:53
kanzureespecially since he is proposing a method of converting people from one type to another21:53
kanzurehe's talking with actually dealing with them21:53
kanzurehm21:53
kanzuresomething's wrong here21:53
fennit's not just 'convert from one type to another' - thats whiteboard mentality21:54
fennits about getting a person to see what's in front of them, to kick the self-boredom habit, and to figure out what they're doing21:54
kanzureright21:54
kanzurebut "its funny that he falls into the trap of categorizing people as one thing or another"21:54
fennbut its easier to say 'we had a packer, made him a mapper' because there's less information being transmitted21:54
kanzureah, I guess we can attribute that to the problems with writing21:55
kanzurethat 'trap' is because he has to fall into that trap in order to write, methinks21:55
kanzureproblems with writing -> self-expression stuff. "read what I meant, not what I said"21:55
fennbut see, the same data compression is going on in your head all the time. it's how you can remember 90 bazillion things in the first place21:55
kanzurere: perl monk motto "do what I mean, not what I say"21:55
kanzuresure, but it's a compression *process*21:56
kanzureit's feedback, systems, cybernetics.21:56
kanzurewe're both meta like that, IMHO21:56
fenni can run a simulation of your mental state and try to reconstruct what you meant, but i can never truly get what you actually meant (without a brain-pod connection)21:56
fenninterestingly enough, this problem is why i got into genetic engineering21:57
kanzurewait, brain-pod connection?21:57
kanzurepod?21:57
kanzurewhat?21:57
fennbypass the difficulties of language by using a direct neural connection from one brain to another21:58
kanzurefunny thing is that the brains will still have to learn to communicate or something21:58
fenni'm not sure it would actually work, now that i know more about how brains work21:58
fennyou'd have to have nanites crawlin all over your neurons, able to transmit impulses (and connections diagrams?) from anywhere inside the brain21:59
kanzuremaybe - possibly fMRI and rTMS could do it21:59
kanzureI've thought about those possibilities before, yeah21:59
kanzurealso DNIs21:59
fennnah those are too coarse21:59
kanzurehttp://heybryan.org/docs/neuro/21:59
kanzuresure21:59
fennget down to the sub-axon level maybe21:59
kanzureI'm not so interested in a DNI per se with another brain even22:00
fennwhy not?22:00
kanzurebut rather, there's a *ton* of information in the brain22:00
kanzurehehe, hold on a sec22:00
kanzureMarkram in the vid was saying that he generates 1 TB of debug data per sec22:00
fennbig whoop22:00
kanzuredon't you think we could do some interesting things when aggregating that?22:00
kanzureif we could tap into the information within our brain?22:00
kanzurenot information like perceptions22:00
fennsure, you could store your dreams, say what you actually meant22:00
kanzure(but maybe perceptions, if you can do that trick, heh)22:00
kanzureno, I mean like all of the 'entropy'22:01
fennsomeone would have to do some "strange days" mind zap to really understand though22:01
fennand they wouldnt have connections for all your concepts (without replicating your brain)22:01
fennlike, you might see something and start laughing (in the reconstructed simulation) but not know why it was funny because you didnt have the memory and associations22:01
kanzureperhaps we could find a way to insert new microcolumns, or have them interfaced to your more meta cognitive processes22:01
kanzurebut this is far out there22:02
fennyou cant physically insert new microcolumns because they're so tightly interwoven22:02
kanzuremostly because I am not sure to what extent microcolumns are organized within the head, i.e., they do have some long-range connections, but do they need them? can they manage with just up-propagating connections and so on? aren't they supposed to be layered?22:02
kanzureetc.22:02
kanzurekeep the new ones somewhere else22:02
kanzurebrain in a jar 22:02
fennbut you could 'logically' insert new connections with nanites sending packets22:03
kanzurenanites are a blackswan22:03
kanzureavoid22:03
fennwhatever22:03
fennuse TMS22:03
kanzurenanites are more of a blackswan than macroscale self-replicating machines ;-)22:03
kanzureyeah, TMS would be interesting to work with22:03
kanzureEd does laser-guiding neuro growth too22:03
kanzureso in combination the two could be very interesting22:03
kanzurehttp://edboyden.org/22:03
fennuse electrodes that you've poked into the brain tissue, disrupting a certain acceptable number of connections22:03
kanzurethe idea is that you can use lasers to help specifically target axon/dendritic growth22:03
fenni say nanites because i'd rather not have needles ripping up my brain22:04
kanzurehave you seen my 3D mesh idea for brain engineering22:04
fennno22:04
fennbtw nanites dont have to be self-repliating22:04
fennthink of it as a really small rfid tag22:05
kanzuregiant MEA, except 3D; have brain grow *into* this area, thus when it's completely done you have this huge infrastructure for stimulation with electricity or neurochemicals (if you have some sort of fancy MEMS mechanical distribution network built into it)22:05
fenni have no idea what you just said22:06
fennwhat's an MEA?22:06
kanzuremicroelectrode array22:06
kanzurehttp://heybryan.org/mediawiki/index.php/Brain_implants22:06
fennok, how do you get this into the brain without shredding all the connections?22:06
kanzurepossibly without the last character22:06
kanzurefenn: "have brain grow into this area"22:06
kanzuretissue engineering - grow a new brain22:06
kanzurebrain in a jar22:06
fennbrains dont generally grow except in infants22:06
fenns/infant/embryo/22:06
kanzurecorrect22:07
kanzurehave you considered that we might be screwed? that our brains are ultimately intractable? we can try to prolong their lives, even interface with other systems, even try to scan in the physical connections, but the vascularization issues + nanotech blackswan stuff is a negatory 22:07
kanzureI do not mean to say I am not optimistic and that I am not working towards solutions22:08
fennum, yah i've considered it22:08
kanzureI mean that maybe our second generations can do something better22:08
kanzureso that maybe our 'copies' that we physically build, engineer from what we know, so that they might have a chance in a digital age or something blah blah blah22:08
fennmy solution was to make an 'improved human' though i didnt think of it as human at the time22:08
kanzurebut really, beyond surface brain implants22:09
kanzureand trying to do nanotech, eh22:09
fennit was a sort of living space station22:09
fennbut biological22:09
kanzureZindell's moonbrain22:09
fennbut more than just a brain22:09
kanzuremade out of 'neurologics'22:09
fennthinking only gets you so far22:09
fennyou need craploads of sensors, manipulators, communication..22:10
fennwhat good is an eyeball in earth orbit?22:10
fennor in the asteroid belt22:10
kanzureuhh22:10
fenncan hardly see anything without a telescope22:10
kanzurehm22:10
kanzuretake a look at this22:10
kanzure  http://www.mel.nist.gov/psl/22:10
kanzure"For many years, the Manufacturing Systems Integration Division (MSID) has22:10
kanzurebeen involved in the definition of a neutral representation of product data,22:10
kanzuremost recently realized through the STEP standard. With that effort well22:10
kanzureunderway, another candidate area for a division focus is the representation22:10
kanzureof manufacturing process. Like product data, process data is also used22:10
kanzurethroughout the life cycle of a product, from early indications of22:10
kanzuremanufacturing process flagged during design, through process planning,22:10
kanzurevalidation, production scheduling and control. In addition, the notion of22:11
kanzureprocess also underlies the entire manufacturing cycle, coordinating the22:11
kanzureworkflow within engineering and shop floor manufacturing. The Process22:11
fennthen *bam* you just got pulverized by a meteor going mach 20022:11
kanzureSpecification Language (PSL) defines a neutral representation for22:11
kanzuremanufacturing processes that supports automated reasoning. Process data is22:11
kanzureused throughout the life cycle of a product, from early indications of22:11
kanzuremanufacturing process flagged during design, through process planning,22:11
kanzurevalidation, production scheduling and control. In addition, the notion of22:11
kanzureprocess also underlies the entire manufacturing cycle, coordinating the22:11
kanzureworkflow within engineering and shop floor manufacturing."22:11
kanzureAnd:22:11
kanzure  http://www.mel.nist.gov/psl/ontology.html22:11
kanzure  http://www.mel.nist.gov/psl/ontology.html22:11
kanzure"The purpose of PSL-Core is to axiomatise a set of intuitive semantic22:11
kanzureprimitives that is adequate for describing the fundamental concepts of22:11
kanzuremanufacturing processes. Consequently, this characterization of basic22:11
kanzureprocesses makes few assumptions about their nature beyond what is needed for22:11
kanzuredescribing those processes, and the Core is therefore rather weak in terms22:11
kanzureof logical expressiveness. In particular, PSL-Core is not strong enough to22:11
kanzureprovide definitions of the many auxiliary notions that become necessary to22:11
kanzuredescribe all intuitions about manufacturing processes."22:11
fennyeah yeah it sounds great until you have to actually deal with it22:11
fenni'm glad they're working on it, but based on my experience with STEP i dont have high hopes22:12
kanzureif they're working on it, then we can eventually approach them22:12
kanzureAlso, even more serious:22:12
kanzure  http://www.mfgquote.com/22:12
kanzure"MFG.com is an online marketplace that saves you time and money by instantly22:12
kanzurematching a buyer's drawings and specifications to qualified suppliers. No22:12
kanzuremore searching or worrying."22:12
kanzurehuh22:12
kanzureinteresting22:12
fennyeah very cool22:12
fenni wonder how they do quality control22:13
-!- marainein [n=marainei@220-253-155-121.VIC.netspace.net.au] has joined #hplusroadmap22:13
kanzureI wonder how they do it at all ...22:14
* kanzure goes to find dinner22:14
fennpeople look at it, bid on the job22:14
fennits like ebay (or rentacoder.com really)22:15
kanzureoh, it sounded like it could match specs to supplies22:48
kanzureutomatically22:48
kanzurebecause they say 'instantly'22:48
kanzureI guess they mean something else ;-)22:48
fenninstantly is relative.. usually this process takes weeks or months22:53
kanzureI was hoping they may have had our agx-make GA jig-puzzle piece22:58
kanzureguess not :-) - they are more social in nature, although a social equivalent wouldn't be a bad idea actually22:58
kanzurehmm22:58
kanzure(momentary) 'experience' is a signal22:59
kanzureto the brain.22:59
kanzurea very massive, parallel signal of course22:59
kanzurebut nevertheless, to the brain it is still a signal; it seems to be a matter of letting certain experiences manage how the brain is to cope with 'experience' but what's the selection process for this?22:59
fennbecause experience is intertwingled with your current brain state23:08
kanzurehm, scenemusic is playing "Cyborg Jeff - HTML Quest"23:10
fenni'm re-thinking my possession by multiple spirits in high school as a breakdown in communication between different areas of my brain23:12
fenni thought my soul had left my body and then started attracting various visiting spirits23:13
fennthere was a dog, a fox-woman, a snake-lizard-futuresoldier, and an arab mathematician23:14
kanzurebreakdown?23:14
kanzurehow serious was this?23:15
fennit wasn't very apparent at all, actually23:15
kanzurebrb23:15
fennactually it must have been in middle school, because the arab taught me spanish23:19
fennthere should be a name-origin wiki23:20

Generated by irclog2html.py 2.15.0.dev0 by Marius Gedminas - find it at mg.pov.lt!