--- Day changed Sat Jul 26 2008
00:20 < fenn> p2pfoundation.net is basically an illustrated proof of how fractionated (balkanized) this movement is
00:20 < fenn> it's almost like one of those ideonomy lists
00:22 < fenn> all permutations of {open, free, collective, collaborative, co-} x {action, behavior, identity, intelligence, design}
00:23  * fenn thinks some redirects are in order
00:52 < kanzure> redirects to the actual sites?
00:53 < fenn> no, just glomming concepts back together since they really mean the same thing
01:03 < fenn> seems that the theoretical foundation of wuffie (doctorow's reputation economy) is stronger/more useful than advogato
01:03 < fenn> er, whuffie
01:04 < fenn> advogato tries to make one big happy community, with no room for dissent
01:14 < kanzure> http://en.wikipedia.org/wiki/Randy_Pausch
01:14 < kanzure> yikes
01:17 < kanzure> self-fulfilling prophecy: http://en.wikipedia.org/wiki/Andries_van_Dam 'It's a tradition in the Brown Computer Science Department to add good natured vandalism to Andy's Wikipedia page.[citation needed]'
01:20 < fenn> wtf
01:21 < fenn> "its a tradition in wikipedia to leave harmful material up and delete good articles"
01:22 < kanzure> eh?
01:33 < kanzure> I can't (or won't) understand this: http://www.integralworld.net/wollersheim3.html
01:34 < kanzure> if anybody can hack through whatever it is that the page is trying to say, please interpret for me
01:38 < fenn> lol
01:39 < fenn> i think this has been baked as unitarian universalism
01:39 < kanzure> hm,
01:40 < kanzure> Tony mentions that http://tech.slashdot.org/article.pl?sid=08/07/07/166253 is Nicole Yankelovich of the Brown University Intermedia team that was mentioned in the Randy Pausch "last lecture" fiasco
01:40 < kanzure> hugh.
01:40 < kanzure> *huh.
01:42 < fenn> it would be interesting to take an intersection set of the beliefs of all or most religions
01:43 < fenn> but i think it'd be hard to get a coherent message if you just did the union set
01:44 < fenn> i suppose you could do some kind of "hamming map" showing the distance from one set of beliefs to another
01:45 < fenn> oh yeah, wonderland looks pretty goofy at first but they have a lot of nice code under the hood
01:46 < fenn> (even though its java)
01:48 < kanzure> sounds kinda like http://canonizer.com/
01:48 < kanzure> but canonizer obviously sucks for other reasons
01:49 < fenn> canonizer is too crowd-sourced to provide an unbiased representation
01:50 < fenn> its like, "what people who bother with canonizer believe" or would be if anyone bothered with it
01:50 < fenn> blah not like i care anyway
01:50 < fenn> 90% of people are crap
01:52 < kanzure> it's worse than that
01:52 < fenn> well, whatever your threshold is in general
01:53 < fenn> i'm just deriving a statement from sturgeon's law
01:54 < fenn> An important corollary to SturgeonsLaw:  ... but ninety percent of everybody thinks they are part of the ten percent that's not crap. 
01:55 < fenn> c2 is great.. i wonder how they maintain the high SNR
01:59 < fenn> hmm.. they point out that you can take advantage of people's bias/inability-to-evaluate-their-own-abilities by making them feel good and thus they will trust you and enjoy your presence
01:59 < fenn> er, by saying how capable and leet they are
02:05 < kanzure> heh
02:05 < kanzure> 1337 for sure
02:06 < fenn> shades of optimus keyboard.. http://krotty.livejournal.com/33074.html
02:07 < fenn> i think i'd rather just wear a 2d barcode thingymabob
03:02 < kanzure> oh wait
03:02 < kanzure> fenn: this is the damn hexayurt guy
03:02 < kanzure> Vinay Gupta <hexayurt@gmail.com>
03:02 < kanzure> he's the friend of the Technocalyps fellow
03:06 < fenn> yep
03:07 < fenn> should i watch technocalypse?
03:10 < kanzure> no
03:10 < kanzure> seriously, it's a waste of time
03:11 < fenn> oh well
03:11 < kanzure> you and I could pull something better off
03:11 < kanzure> by playing a video of one of our browsing sessions or something
03:12 < fenn> hmm.. video production's harder than it looks
03:12 < kanzure> no doubt
03:13 < kanzure> I've certainly had my fair share of video production assignments in high school
03:13 < kanzure> the software tends to suck
03:18 < fenn> it's kinda like writing, in that it's hard to know what the viewer understands because you've seen it fifty times in a row
03:18 < kanzure> oh, well, that too
03:19 < kanzure> in this edition of weird searches leading to Bryan's site:
03:19 < kanzure> "salts used to make nickel nanotube by cyclic voltameter"
03:19 < kanzure> second result is me on Google.
03:19 < fenn> all the non-obvious flaws stick out because you're worried about them, but the big overall issues don't stand out at all
03:19 < kanzure> sometimes I wonder if that's just because of having the wrong big idea
03:19 < kanzure> erm, nevermind
03:20 < fenn> that's a very valid statement
03:20 < kanzure> but it could be interpreted to mean "if people don't understand it, it sucks" which is not the point
03:20 < fenn> you can bang your head against a wall with 99% perspiration and not make a damn bit of difference
03:21 < fenn> heh i'm watching the andy pausch video.. the cake is just awesom
03:22 < fenn> +r +e
03:22 < kanzure> ron pausch methinks
03:22 < kanzure> maybe not
03:23 < kanzure> 'randy'
03:23 < kanzure>  okay
03:23 < kanzure> nevermind
03:26 < kanzure> it does seem like an unnecessarily large percentage of those in academia are jerks
03:27 < kanzure> he mentions that others corectly identified him as a jerk
03:36 < kanzure> I do not understand. http://hca.gilead.org.il/tinderbx.html
03:50 < kanzure> heh
03:50 < kanzure> the Byte Scouts
03:50 < fenn> i think the moral of this story is 'dont pay people too much'
03:51 < kanzure> i think this guy was on acid
03:51 < fenn> and possibly 'might makes right'
03:51 < fenn> or something like that
03:58 < kanzure> http://knol.google.com
03:58 < kanzure> 'A knol is an authoritative article about a specific topic.'
03:58 < kanzure> quick, let's go jump on open source
03:58 < kanzure> this was as of Wednesday, hm
04:03 < kanzure> I wonder if Google > Wikipedia or if knol will be > Wikipedia
04:03 < kanzure> if so we might have an opportunity to set the collaborative snowball right
04:13 < kanzure> Heh. http://www.8bitpeoples.com/mp3/get/492/8bp078-01-mr_spastic-net.mp3
04:27 < kanzure> fenn: http://en.wikipedia.org/wiki/Inside_contracting what?
04:28 < kanzure> http://en.wikipedia.org/wiki/American_System_of_Watch_Manufacturing
04:28 < kanzure> "There was probably no greater industrial challenge, no line of manufacturing in the world demanding such a high grade of business and mechanical ability, and such unremitting care and oversight, combined with technical skill and individual dexterity and judgment as is indispensable in systematic watch making. "
04:28 < kanzure> wtf
05:18 < fenn> knol seems like danny hillis' revenge
05:18 < fenn> expert-pedia or whatever
05:18 < kanzure> hm
05:19 < fenn> re the quote, watches are harder to make than the primitive guns of the time
05:19 < kanzure> guess "of the time" is the key phrase there
05:19 < fenn> they werent even cartridge bullets were they?
05:19 < fenn> maybe paper cartridge
05:20 < fenn> i think knol is gross
05:20 < kanzure> "yet another" sort of gross?
05:20 < kanzure> or the implementation? the pages weren't really all that fancy .. just stupid AJAX bullshit again
05:21 < kanzure> and having it in this closed database format
05:21 < kanzure> how original :)
05:21 < fenn> just google taking over the world kind of gross
05:22 < kanzure> google should be smarter than this though
05:22 < fenn> these fucking doctors think their advice is so important, at least a million times more important than anything else on wikipedia, so they have to set up their own little kingdom where they can be big dog
05:22 < kanzure> isn't google made up of the people that know that making the formats to share this data would be more long term sustaintable?
05:22 < kanzure> *sustainable
05:22 < kanzure> *that should know
05:22 < fenn> google is like bell labs in the 70's
05:22 < kanzure> wasn't bell labs mostly the 50s and 60s ?
05:23 < fenn> eh, not really
05:23 < kanzure> transistor?
05:23 < kanzure> google is a powerhouse of postdocs last I heard, but if these guys are so bright, ...
05:24 < fenn> The 1950s saw fewer developments and less activity on the scientific side. ... The 1970s and 1980s saw more and more computer-related inventions at the Bell Laboratories as part of the personal computing revolution
05:25 < fenn> specifically i was thinking of Unix, C, and "worse is better"
05:26 < kanzure> I hardly see Google doing the equivalent of unix
05:27 < fenn> they Are building an operating system.. :\
05:27 < kanzure> not you too :p
05:28 < fenn> anyway, how is knol collaborative?
05:28 < kanzure> first time I swear I've heard of building an operating system by publicity
05:28 < kanzure> oh wait
05:28 < kanzure> Microsoft
05:28 < kanzure> damn.
05:28 < fenn> ms-dos wasn't half bad
05:28 < fenn> i think they stole it from some other company though
05:29 < kanzure> I think I remember some DOS machines and could grok the overal scheme it was working from
05:29 < kanzure> it worked I guess, but I do admit preferance to the nixes
05:29 < kanzure> *preference
05:29 < fenn> early unix was rather warty
05:30 < kanzure> early unix was proprietary I recall :)
05:30 < fenn> but it had principles at least
05:30 < kanzure> I have no experience with the originals
05:30 < fenn> i've only used irix
05:31 < fenn> supposedly it was one of the better ones
05:37 < fenn> ffs knol isnt even organized at all
05:37 < fenn> "Bag o'knols"
05:37 < fenn> 1 2 3 4 5 6 7 Next
05:38 < kanzure> how many per page?
05:39 < fenn> 50
05:39 < kanzure> 350 knols and I know about it?
05:39 < kanzure> so this must be early?
05:39 < kanzure> or possibly stupid
05:39 < fenn> its 95% medical subjects
05:39 < fenn> er.. medical advice
05:41 < fenn> i'd love to see a 'lonk' as uncyclopedia is to wikipedia
05:42 < kanzure> I need a better way to write exp.html
05:42 < kanzure> maybe I should just put up a todo list
05:42 < kanzure> "make git + wiki"
05:43 < kanzure> "make py-yaml validator using gnu units as backend"
05:43 < kanzure> "make tons of shitty python classes to represent manufacturable thinglets"
05:44 < kanzure> actually, git + wiki, I was thinking that it might be more productive to just let each user have their own 'ghosted' git clone on the server side
05:44 < kanzure> I don't know how else to run it really
05:44 < kanzure> maybe git keeps track of user changes when it's on a single host?
05:44 < kanzure> i.e., the differences that chown modifies?
05:44 < fenn> todo list == roadmap no?
05:44 < fenn> whatever floats your boat :)
05:45 < kanzure> yeah, yeah, whatever
05:45 < fenn> why would each user have their own clone? just make branche
05:45 < kanzure> how does this work?
05:45 < fenn> uh, you know what a branch is right?
05:45 < kanzure> thought I did
05:46 < fenn> well, one way to think of it is like you undo a bunch of changes and then start working from there
05:46 < kanzure> sure
05:46  * kanzure did this in some code bases today
05:46 < kanzure> CTRL+Z for the win, and such
05:46 < fenn> another way to think of it is like making a little sandbox to play in that won't mess up everyone else's stuff
05:46 < kanzure> so it's not a full clone?
05:46 < fenn> no
05:46 < kanzure> excellent
05:47 < fenn> but it's more of a svn kind of thing because if you have your own repo you can just work on whatever you're doing with out worrying if it messes up other people's code (since they arent pulling from you until its done)
05:47 < kanzure> http://www-cs-students.stanford.edu/~blynn/gitmagic/ch04.html branch wizardry
05:47 < kanzure> right
05:47 < kanzure> but if it's on a wiki server thing
05:47 < kanzure> oh
05:47 < fenn> then the user doesnt have to download anything
05:48 < kanzure> uh, well
05:48 < kanzure> http is still downloading of course
05:48 < kanzure> my "oh" was the idea of faking a wiki for users
05:48 < kanzure> since they don't git it
05:48 < kanzure> javascript + ajax bullshit => magical 'wiki'
05:48 < kanzure> where they do, in fact, host a mini repo on their end
05:48 < kanzure> either in cookies or on their local system somewhere else
05:49 < kanzure> but that would require some other crappy development
05:49 < fenn> what is a 'mini repo'?
05:49 < kanzure> not full thing
05:49 < kanzure> i.e., biotech.git is 500 MB
05:49 < fenn> how do you track content changes if you dont have the full repo?
05:49 < fenn> (this is the only thing about git that sucks)
05:50 < fenn> or maybe i just havent looked hard enough yet
05:50 < kanzure> that's synchronization, right?
05:50 < kanzure> which kind of sucks with asynchronous distributed development, one would expect
05:50 < fenn> i _thought_ you could check out partial repos and then merge them
05:50 < fenn> but i cant figure out how to do it (havent tried very hard either)
05:51 < fenn> i dont think forcing people to download 500mb is a good idea
05:51 < kanzure> exactly
05:51 < kanzure> 'In contrary to SVN  neither Git nor Mercurial can handle partial repositories, so you always have to download all of it and cannot restrict the download to a ...'
05:51 < fenn> this conflicts with what linus said
05:51 < fenn> supposedly you can track a subset of content
05:51 < kanzure> http://lists.openfabrics.org/pipermail/general/2006-November/028473.html "git cloning partial tree"
05:52 < kanzure> 'The first format is what "git-apply --index-info" reports, and used to reconstruct a partial tree that is used for phony merge base tree when falling back ...'
05:52 < fenn> derr..
05:52 < fenn> but what do you do with this partial checkout once you've modified it?
05:52 < kanzure> supposedly you send it back?
05:53 < kanzure> well
05:53 < kanzure> at least advertise it or something
05:53 < kanzure> I don't know.
05:53 < kanzure> you tell me?
05:53 < fenn> it just doesnt make sense to do that if you can't merge it back in somehow
05:53 < kanzure> which is svn
05:53 < fenn> otherwise you might as well just submit a patch
05:54 < fenn> oh hmm
05:54 < fenn> git-checkout doesnt need a repo to work
05:54 < fenn> so maybe you can't commit once you download the files
05:55 < fenn> s/repo/local repo/
05:55 < kanzure> but what about people wanting to do development work? 500 MB again.
05:56 < fenn> what kind of development work?
05:56 < fenn> there's always ssh
05:56 < kanzure> well, let's say there's only 2 MB of relevant source
05:56 < kanzure> uh, for all of the users?
05:56 < fenn> no
05:56 < fenn> though if someone messed up your repo i suppose you could just undo it
05:57 < kanzure> wait, let's just look at the linux kernel 
05:57 < kanzure> how do they do it?
05:57 < kanzure> how is it that there's an "official version" ?
05:57 < fenn> linus is the benevolent dictator
05:57 < fenn> he says what goes in the kernel
05:57 < kanzure> and then he puts up his checked out repo?
05:57 < fenn> right
05:57 < fenn> actually its more structured than that but that's the general idea
05:57 < kanzure> but that's just svn overall
05:58 < kanzure> in the end, basically.
05:58 < fenn> not really
05:58 < kanzure> uh?
05:58 < fenn> because the code is all out there
05:58 < kanzure> sure, 
05:58 < kanzure> but not to people not in the development circles
05:58 < fenn> its not that hard to patch a kernel
05:58 < fenn> before this there were mailing lists dedicated just to kernel patches
05:58 < fenn> its a step up i believe
05:59 < kanzure> wait, where are the links distributed then if not on mailing lists?
05:59 < kanzure> my point is that the fact that Linus has access to the kernel.org site and can put up an 'official' download is somewhat like subversion in that it's the main release and that's what the majority of people see and so on
05:59 < kanzure> sure the code is still out there
05:59 < fenn> no you dont understand, the patches used to be sent to the mailing list directly
05:59 < kanzure> oh
06:00 < kanzure> not automatically integrated or somesort
06:00 < fenn> ok, so i skipped a bit of important detail it seems
06:00 < fenn> linus doesn't pull from all of the 99 million kernel developers
06:00 < fenn> he only pulls from like 3-4 people
06:00 < kanzure> he probably pulls from a few, no?
06:01 < kanzure> and these few from others
06:01 < fenn> yes, all the way down
06:01 < kanzure> long tail dynamics
06:01 < fenn> but the version history is preserved
06:01 < fenn> as if they were working on one big svn
06:01 < fenn> but without all the svn headaches
06:02 < kanzure> the headaches were the conflicts?
06:02 < fenn> conflicting development, limited branch namespace, commit access, slowness
06:02 < fenn> i guess they had a lot of trouble with three way merge
06:02 < kanzure> it's still generally slow if you're transferring 500 MB between developers :-)
06:02 < fenn> well, dont do that!
06:02 < kanzure> heh
06:03 < fenn> the 500mb repo is a bad idea
06:03 < kanzure> yep
06:03 < fenn> it will never get smaller, only bigger
06:03 < fenn> but that's not what you're trying to do, 99% of that data is irrelevant or redundant
06:04 < kanzure> certainly
06:04 < kanzure> still though
06:04 < kanzure> if the system isn't scalable ..
06:04 < fenn> if you have 500mb of code, something is wrong
06:04 < fenn> look at opencascade, what a fucking mess
06:05 < fenn> sure, it's "open source" and "industrial grade cad" but its impossible to use and change
06:08 < kanzure> a wiki doesn't make sense in this context though
06:08 < fenn> anyway, nobody's going to be doing any sorting and winnowing unless you make it clear what needs to be done by providing a framework
06:08 < kanzure> and people are going to cry without their beloved wiki
06:09 < fenn> a wiki is just html + version control
06:09 < fenn> people are going to cry without their GUI
06:10 < fenn> principle of least surprise
06:12 < fenn> has anyone offered to help sort the biohack stuff?
06:12 < fenn> or is it just territorial bitching 'my way is better' stuff
06:30 < fenn> well, working with partial checkouts is a pain
06:37 < fenn> it seems to work, you just have to be careful with git-add and git-commit to only commit the files you have (otherwise it will delete the ones you havent checked out)
06:38 < fenn> what i did was: git-clone -n ~/git/autogenix.git
06:39 < fenn> git-checkout master test.mdwn
06:39 < fenn> nano test.mdwn
06:39 < fenn> git-add test.mdwn
06:40 < fenn> git-status
06:40 < fenn> git-commit
06:40 < fenn> it should say something like 'one file changed'
06:46 < fenn> you should be able to repeat my results with git-clone -n http://fennetic.net/git/autogenix.git
06:49 < fenn> donuts before bed == bad idea
07:12 < kanzure> sugar?
07:13 < kanzure> or teh fats?
07:13 < kanzure> hm, I wonder if breakfast is open yet
07:15 < kanzure> btw, are you still playing around with ikiwiki mdwn?
07:15 < kanzure> and who's the one doing the territorial bitching, me?
07:16 < kanzure> because no one has even offered to comment on it ;-)
07:16 < kanzure> and by no one I mean absolutely no one
08:33 < kanzure> Those bastards. The flatscreens in the cafeteria had a documentary on insomnia playing.
08:37 < kanzure> anyway, if the git-checkout and git-add method is to work it'll probably need the usage of 'adduser' on the server end of things
08:37 -!- jm [n=jm@p57B9D3F9.dip.t-dialin.net] has joined #hplusroadmap
08:37 < kanzure> I've never implemented something like that, is that wise?
08:37 < kanzure> I mean, adding user accounts on the server?
08:37 < kanzure> Slashdot makes it look like that's what they are doing, but it's probably just fancy mod rewrites (and I'm too lazy to go check the slashcode)
08:37 < kanzure> Sourceforge definitely does it.
08:38 < kanzure> bah, it has to be less computationally intensive than doing the mysql db lookups that mediawiki does
09:21 < kanzure> Tony: "Re your conversations with Gupta, Bauwens et al, my entrepreneurial colleagues have acquired some rights to an electrolysis-based system developed in (from memory) San Antonio which is supposed to do everything from purify water to provide better energy storage than batteries.
09:21 < kanzure> I had a vague thought some weeks ago about having you take a look at it, but I think you were in the middle of your move out of home at the time and it slipped down my priorities."
09:21 < kanzure> hm
09:39 -!- kanzure [i=bryan@66.112.232.74] has quit [Remote closed the connection]
10:23 -!- kanzure [i=bryan@66.112.232.74] has joined #hplusroadmap
10:25 < kanzure> fenn:  'I've got a friend who did exactly this:  made a python wiki engine that uses
10:25 < kanzure> mercurial as revision control: http://dandelion.sheep.art.pl/  also makes it
10:25 < kanzure> possible to sync up wikis, fork them, clone them, many other possibilities.
10:25 < kanzure> I am using this as a personal wiki on my desktop, and also experimenting
10:25 < kanzure> with developing many different front-ends for it.
10:25 < kanzure> '
11:59 -!- kanzure [i=bryan@66.112.232.74] has quit [Remote closed the connection]
14:34 < ybit> http://bazaar-vcs.org/BzrWhy
14:52 < fenn> "Revision objects are simple: r1, r2 etc. In git they are SHA1s"
14:52 < fenn> this is because there's no central repository, so two different r2's can exist at a time
14:57 < fenn> but yeah it's annoying
15:04 < fenn> now i'm wondering if they did it that way because it acts as a checksum
19:27 -!- jm [n=jm@p57B9D3F9.dip.t-dialin.net] has quit [Read error: 110 (Connection timed out)]
23:55 -!- kanzure [n=bryan@adsl-76-229-131-47.dsl.pltn13.sbcglobal.net] has joined #hplusroadmap