IRC log started Tue Jul 13 00:00:00 1999
[msg(TUNES)] permlog 1999.0713
<hcf> can tunes be properly reshaped?
<water> the project?
<abi> well, the project is too large for one person :)
<water> abi says it all
<hcf> abi: u said it
<water> sigh...
<water> people just don't analyze things for causality patterns like i try to do
<hcf> can they not be taught?
<hcf> and what r causality patterns?
<water> causality = "why things are so" or "what causes such-and-such"
<water> "a" causes "b"
<water> or "a" contributes to "b" being the case
<hcf> thotso
<water> that's why i point at Fare
<water> i must see this thing through
<water> the universe lets me have no peace
<water> :(
<hcf> how do u propose we fix things?
<water> i don't have definite answers for such things
<water> not at this point
<hcf> where do we start?
<water> well, let's think about the possibilities
<water> 1) i branch from tunes
<water> n/m screw numbering systems
<Fare> n/m ?
<water> we could try to convince Fare to do something
<hcf> a hash instead?
<hcf> abi: n/m is nevermind
<Fare> what numbering systems?
<water> i was going to itemize options
<water> "1) i branch from tunes"
12:10am
<water> that's unnecessary structure
<hcf> so, theres: fix fare or replace fare?
<water> not just that
<water> but those could categorize other solutions
<hcf> or leave fare in place as a figure head
<water> it's just that replacing fare for me might mean replacing the project and social club he's made
<water> and not sharing anything with them, even going far enough to forbid gpl
<Fare> I'm ready to be replaced, but not by water
<Fare> we have different goals enough
<water> fare, you're not ready to be replaced by anyone
<Fare> I'm ready to be replaced by someone who gets code written
<water> you've got this entire project running into the ground quite well
<Fare> not by someone even more theoretic than I
<water> exactly
<water> yes, but you don't analyze anything
<Fare> ?
<water> especially not your ideas of tunes
<Fare> I have ideas of tunes and ideas not of tunes
<Fare> I keep them separate
<water> you don't analyze them at all
<water> you don't refine them
<water> you don't question your answers
<water> hell, you don't even address what godel's incompleteness and undecidability ideas have for tunes or reflection
<water> and look what your several years of work have you
<water> you're not a leader, you're an obstacle
<hcf> uv established that
<water> you don't push people to explore, you tell them what's out there
<water> ok
<water> but he had to hear that
<Fare> water: you say rubbish
<hcf> oh god, here we again
<hcf> s/we/we go/
<Fare> I'm quite conscious of issues of incompleteness and undecidability
<water> i'm about to give up for tonight anyway
<water> not for reflection. you don't address it anywhere on the site
<Fare> I'm not a leader, but I'm all there is to Tunes right now;
<Fare> I've never hidden the fact that I came up to lead by lack of another better leader
<hcf> a depsnser of 'feel free's
<water> you don't address anything that i've looked into
<Fare> then look into it and let the tunes project alone!
<water> you've even said 'wtf is that?' about some of the ideas i've researched
<Fare> water: yes, and I'll continue to say it, whenever you speak some alien language
12:20am
<water> it just points out your ignorance
<Fare> water: you extract words from various context without precision
<water> yes, that's my style
<Fare> it's ok to be ignorant, if you're open-minded
<Fare> water: your style is confusing
<water> i've learned way too much to be bothered with being precise all the time
<water> i've researched ten times the scale of topics that apply to tunes
<Fare> then be precise when you explain things only
<Fare> water: good. So found your better larger project
<Fare> and leave us quiet
<water> no
<Fare> what was alternative #2 ?
<Fare> open your own mailing-list, somewhere.
<water> no
<Fare> I'm sure Tril is willing to host it
<hcf> Fare: shut up
<water> i didn't join this project for no good reason
<water> and i'm not a fan of internet based r&d anyway
<water> not for collaboration anyway
<water> well, good night
<water> this has been quite a discussion for me
<hcf> the desire to reach the goals should outweigh everything else
<water> i still haven't said everything i'd like to say to Fare, but i've already said more than he's willing to listen to by far.
* Fare/#Tunes shuts up and goes rewriting his implement1 paper with understandable examples
-:- water [water@ppp-tnt-148.tscnet.net] has left #tunes []
<hcf> Fare: wait a sec
<Fare> hcf: what?
<hcf> u mentioned abi doing a note system
<hcf> r u aware of memoserv?
<Fare> memoserv?
<hcf> /msg memoserv help
<Fare> thanks!
<hcf> pretty damn easy
<Fare> now going to work
12:30am
* hcf/#tunes is away: (afk)
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us734.javanet.com])
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- abo [d96-abo@brown01.nada.kth.se] has joined #tunes
-:- Linuxlov1 [user2832@p3E9E1CE1.dip.t-dialin.net] has joined #Tunes
-:- Linuxlov1 [user2832@p3E9E1CE1.dip.t-dialin.net] has left #Tunes []
-:- La [user5570@210.121.233.231] has joined #Tunes
-:- SignOff La: #TUNES (Leaving)
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- SignOff abo: #TUNES (Ping timeout for abo[brown01.nada.kth.se])
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  Quick rehubbing will take place in a moment, you will see some splits and joins.
-:- NetSplit: forward.openprojects.net split from sterling.openprojects.net [10:26am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [forward.openprojects.net]
-:- Netjoined: forward.openprojects.net sterling.openprojects.net
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #Tunes
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- Netjoined: asimov.openprojects.net irc.linux.com
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #Tunes
-:- NetSplit: irc.linux.com split from asimov.openprojects.net [10:28am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [irc.linux.com]
!sterling.openprojects.net!! Remote CONNECT irc.linux.com 8005 from dilbert
-:- Netjoined: irc.linux.com asimov.openprojects.net
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #Tunes
-:- NetSplit: tolkien.openprojects.net split from varley.openprojects.net [10:29am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [tolkien.openprojects.net]
!sterling.openprojects.net!! Remote CONNECT tolkien.openprojects.net 8005 from dilbert
-:- Netjoined: tolkien.openprojects.net varley.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Assuming everything remains stable, we shouldn't need to rehub again for a while.  Thanks for your patience!
-:- s_r [sr@phila-dialup268.nni.com] has joined #tunes
* Fare/#Tunes has found how to make implement1 more "sexy" to the CS community -- advertising concurrency aspects
<Tril> .cx is not the north pole. never believe _QZ
<Tril> it's a small island country near indonesia
<s_r> .nu?
<Tril> that's Niue and it's also not the north pole
!brin.openprojects.net!! Remote CONNECT huxley.openprojects.net 8005 from dilbert
<Tril> see www.cx and www.nu
<s_r> what's antarctica?
11:10am
<Tril> AQ
<Fare> what's sugercandyland?
<Fare> s/er/ar/
<s_r> are there any networks on .aq?
<Tril> no idae
<Fare> aq is not a country; there cannot be a .aq domain
<Fare> it's split between countries with naval power
<Tril> there's plenty domains that ain't countries, Fare
<s_r> .ae
<Tril> Fare: I think you and water need to learn to work together better
<Fare> sure
<Fare> I think he has some kind of inferiority complex of some kind for being self-taught
<s_r> fare aren't you self-taught?
<Fare> only half so.
<Fare> I have an academic cover.
11:20am
<Fare> I'm even on the road to a PhD
<Fare> (not a glorious road, tho)
<s_r> have you found any professors who were your intellectual superior?
<Fare> and even this "self-teaching" was done in large part thanks to the learning atmosphere of ENS
<Tril> why can't water ignore the pragmatic coders
<Tril> He doesnt have a problem ignoring the lurkers on the mailing list
<Fare> s_r: I have found many co-students, professors, and researchers, who were indeed, and by far, my intellectual superiors.
<Fare> and I insist: by far.
<Fare> Going to the ENS was a most enlightening experience for that.
<Fare> now, the intellect is not a linear, total order
<s_r> Tril: why do you say that the coders are pragmatic?
<Tril> s_R I'm reading yesterday's log, water said it
<Fare> I still like to believe that there are a few points where I have something original to teach.
<s_r> Fare have you ever played "chess" or "go"?
<Fare> s_r: I'm a desperately bad player at either game.
<s_r> do you enjoy "go"?
<Fare> I can't focus enough.
<Fare> s_r: in as much as I understand it, yes. But I'm too weak to really enjoy it properly.
<Fare> s_r: its subtleties are mostly out of my reach.
<Tril> MR_Wrong are you here?
<s_r> i've heard people say that go is more strategic a game than chess
<Fare> s_r: most people say that, too
>>> s_r [sr@phila-dialup268.nni.com] requested PING 931901445 from #tunes
* Tril/#TUNES dislikes 2-player competitive 0-sum games
<Fare> s_r: seen http://www.ai.univ-paris8.fr/~cazenave N
<Fare> s/N/?/
<Tril> the mindset of chess or go or any of those is that when you win, the other loses. It's a bad mindset to get into.
<Fare> Tril: take any constant sum game, remove the constant...
<Tril> and what is good for one, is bad for the other
<Fare> Tril: or add "nature" as the loser.
<Tril> those games can't be changed like that.
<Fare> Tril: go is more subtle than chess.
<Tril> new games must be developed.
<Fare> Tril: programming tunes is such a game :)
* Fare/#Tunes feels the call of 61453
<Tril> Mr_Wrong has some administrative ideas. we should arrange a meeting with Beholder, water and Mr_Wrong and Fare
<Fare> well, going to my Bro's
<Fare> where in the world is Mr_Wrong?
<Tril> not that kind of meeting!
<Fare> why don't they get funded to participate to ICFP? ;->
11:30am
<Fare> we could meet in Paris...
<Tril> MR_Wrong is at pacbell.net in California
<Tril> San Fran
<Fare> better ific than kard!
<Tril> ?
<s_r> fare: when tunes is finished, i will program a "go" game for it
<Tril> if you knew what tunes was you could program it now
<Tril> water says tunes won't accept new or different ideas? 
<Tril> water, get in here now!!
<Tril> see you later Fare
-:- smkl [sami@MCDII.rdyn.saunalahti.fi] has joined #tunes
<Tril> hi, smkl
<smkl> hello Tril
-:- Tril has changed the topic on channel #Tunes to: what a bunch of ignorant fellows we are
<Tril> water disqualifies himself to be leader by his own definition
<Tril> hehe...
<Tril> good job, water
11:40am
* Tril/#TUNES talks back in time
<Fare> ?
<Tril> Fare: I agree with water, you are too stubborn
<Tril> However, i Think he is wrong in assuming you believe tunes to be scheme
<Fare> I am stubborn, sure
<Fare> you need be stubborn to keep that project alive without code
<Fare> and without theory
<Fare> and without funding
<Fare> and without anything
<Fare> If I were more code-wise, maybe things would be better, tho
<Tril> Stop encouraging him to split the project
<Tril> they're too close to work as a split
<Tril> you MUST cooperate with him
<Fare> I must try.
<Fare> it takes two people to cooperate :)
<Fare> really gotta go
* Fare/#Tunes is away 61453
<Tril> ok, 
<Tril> bye
<Tril> abi: be water
<abi> you know i'm not going to accept that
<Tril> abi: be tril
<Tril> abi: babi
<Tril> abi:e abi
<Tril> ack
11:50am
<Tril> abi cam
-:- hcf [nef@me-portland-us744.javanet.com] has joined #tunes
<Tril> hey, hcf
<hcf> hoy Tril
12:10pm
<Tril> when does water usually show up?
<hcf> earlier and later than now
<hcf> Tril: u get my memo?
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<Tril> memo? from memoserv? is that automatic? I havent seen any
<Tril> or do I have to check in to get my memo?
<hcf> it should tell u have a memo waiting when u log in
<Tril> hcf: I guess I wasn't identified to nickserv
<Tril> but then how come I got ops?  I think the services are all !#@*$'d up.
<hcf> try /msg memoserv list
<Tril> I thought that since water tried to make that #modtunes the other day
<Tril> ok i got it
<hcf> whats seems screwed up about services?
<hcf> the #modtunes services seem to work fine
<Tril> then why was everybody having problems seing people in the channel ?
<hcf> probly just a network glitch
<hcf> is it persisting?
<Tril> I should NOT have ops in #TUNES if I'm not identified to NickServ.
<Tril> are there dox for these loggers?
<hcf> memoserv should of notified u when i sent the memo (~12 hrs ago)
<Tril> maybe, but I'm online 24hrs and I don't read my scrollback
<Tril> I only read the IRCLOG when someone says I need to
12:20pm
<Tril> i.e. email is better to send me something
<Tril> i got a notice of a new memo
<Tril> that contains absoultely nothing :)
<hcf> '...' isnt nothing
* Tril/#tunes hits hcf over the head with a solar mass of AOL disks
<hcf> so u check those urls out?
<Tril> i am 
<Tril> test
<Tril> test2
12:30pm
* Tril/#TUNES is away: (afk) [BX-MsgLog Off]
-:- hcf is now known as hcf_
-:- SignOff hcf_: #TUNES (Leaving)
-:- hcf_ [nef@me-portland-us744.javanet.com] has joined #tunes
<s_r> hey hcf
12:50pm
<hcf_> hoy s_r
-:- hcf_ is now known as hcf
01:00pm
-:- s_r [sr@phila-dialup268.nni.com] has left #tunes []
-:- eStormy [stormy@rain.futuresouth.com] has joined #Tunes
<eStormy> great topic
01:50pm
-:- binEng [Anders@dialup81-1-59.swipnet.se] has joined #tunes
<hcf> hoy bin
<binEng> hi hcf
* binEng/#tunes wonders over the topic... and Mr_Wrong
02:10pm
-:- hcf has changed the topic on channel #Tunes to: Free Reflective Computing System <http://www.tunes.org> || When Egos Collide <http://slashdot.org/features/99/07/12/1639202.shtml>
-:- hcf has changed the topic on channel #Tunes to: Free Reflective Computing System <http://www.tunes.org> || When Egos Collide <http://slashdot.org/features/99/07/12/1639202.shtml> see also: http://portslave.mnic.net/other.html
-:- SignOff smkl: #TUNES (Read error to smkl[MCDII.rdyn.saunalahti.fi]: EOF from client)
-:- smkl [sami@MCDII.rdyn.saunalahti.fi] has joined #tunes
-:- Nate [argh@tnt-1-215-56k.portsmouth.zoomnet.net] has joined #Tunes
-:- krz [kzaback@gnat6.owo.com] has joined #tunes
<smkl> hello Nate and krz
<krz> hello
<abi> hoy, krz
-:- SignOff binEng: #TUNES (Read error to binEng[dialup81-1-59.swipnet.se]: Connection reset by peer)
-:- binEng [Anders@dialup81-1-59.swipnet.se] has joined #tunes
-:- krz [kzaback@gnat6.owo.com] has left #tunes []
03:20pm
-:- s_r [sr@phila-dialup496.nni.com] has joined #tunes
<_QZ> greetings
<s_r> QZ how are you going to implement dynarec in BriX?
<_QZ> eh?
<s_r> you said brix would have dynamic recompilation
<_QZ> btw it is 'brix' or 'BRiX', and never any other combination of capitalization
<_QZ> ya when u compile source code it becomes linked to the code binary
<_QZ> when u modify that source it will recompile the binary
<_QZ> if u want to make changes to a binary but also want the old one then u need to duplicate the source object to remove the link to the binary
<_QZ> the original source object will maintain its link to the original binary and the new source object will create a new binary and link to it
<_QZ> u can also modify the kernel source and the kernel will change without rebooting, as long as u didnt modify any data structures
03:30pm
* hcf/#tunes is bored, somebody talk
<binEng> k
<binEng> Suggestions for features in RPGs/MUDs?
* binEng/#tunes is working on pinning down akos
<hcf> as if it were a dead butterfly?
<binEng> :/ nah
<binEng> fleshing out?
03:50pm
-:- SignOff Nate: #TUNES (Leaving)
<s_r> hcf are you designing an OS
<s_r> ?
<hcf> s_r: kinda, but not sharing any details
<s_r> sort of like ms?
<_QZ> haha
<_QZ> hcf: u arent making an os
-:- s_r [sr@phila-dialup496.nni.com] has left #tunes []
04:10pm
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us744.javanet.com])
-:- hcf [nef@me-portland-us107.javanet.com] has joined #tunes
-:- s_r [sr@phila-dialup460.nni.com] has joined #tunes
-:- SignOff smkl: #TUNES (Ping timeout for smkl[MCDII.rdyn.saunalahti.fi])
-:- SignOff s_r: #TUNES (Read error to s_r[phila-dialup460.nni.com]: Connection reset by peer)
-:- s_r [s_r@phila-dialup165.nni.com] has joined #tunes
* hcf/#tunes is away: (bbiaf)
-:- smkl [sami@MCDII.rdyn.saunalahti.fi] has joined #tunes
* hcf/#tunes is back
-:- s_r [s_r@phila-dialup165.nni.com] has left #tunes []
<binEng> has now changed his primary email address to bineng@tpp.dhs.org
<binEng> crap :)
<binEng> /msg did it
<hcf> uhm ok
<binEng> now I get how this work
<binEng> ...which also means I'm done here for today. Good day and good night.
-:- SignOff binEng: #TUNES ( <k!14>)
05:50pm
-:- SignOff smkl: #TUNES (sleeping ...)
<_QZ> does #tpp have any meaningful conversations?
06:30pm
<hcf> _QZ: some times
06:40pm
-:- SignOff hcf: #TUNES (Leaving)
-:- s_r [sr@phila-dialup026.nni.com] has joined #tunes
<Mr_Wrong> Hi everyone
<s_r> what brings you here Mr_Wrong?
<Mr_Wrong> I came to stir up controversy
<s_r> here to introduce a new revolutionary OS?
<Mr_Wrong> Not really
<s_r> WrongOS
<Mr_Wrong> That already exists, it's called Windows
<s_r> heh
<s_r> Windows 2k is a fat pig
-:- hcf [nef@me-portland-us111.javanet.com] has joined #tunes
<Mr_Wrong> Fare, did you finally kick water off the project?
<hcf> Mr_Wrong: u want that to happen?
<Mr_Wrong> He sounds like I did... well... in high school
<Mr_Wrong> No, I want him to learn to communicate with people better
<Mr_Wrong> "You're wrong... go read my 50 page paper to understand why"
<hcf> heh, yeah
<Mr_Wrong> That's not going to win him any converts
08:10pm
<Mr_Wrong> You know, Tunes should probably try to adopt someone else's kernel rather than writing their own
<Mr_Wrong> It doesn't seem to make any sense to write yet another kernel
<s_r> like Mach?
<Mr_Wrong> Preferably not Mach, but Mach might be a good start
<s_r> but most kernels are inadequate
<Mr_Wrong> Preferably a kernel without filesystems and bloat like that in it
<s_r> a standard microkernel?
<s_r> like BLT?
<Mr_Wrong> All the kernel needs to do is securely multiplex system resources
<s_r> blt?
<Mr_Wrong> Or an exokernel
<s_r> openblt?
<abi> openblt is just yet another uk
<Mr_Wrong> Too bad there are only two exokernels and they're both still experimental
<s_r> what are the names of the two exokernels?
<Mr_Wrong> I suppose it would be easy enough to build a kernel on top of Flux
<hcf> Mr_Wrong: besides code related issues, what do u think tunes should do?
<Mr_Wrong> I think it should document in more detail what Tunes is intended to be, and with less fluff
<s_r> you're code oriented
<Mr_Wrong> The existing documentation meanders somewhat
<Mr_Wrong> I think if we get some code it will attract some more coders
<hcf> Mr_Wrong: whatever about the ego clashes?
<s_r> there is Retro
<hcf> s/whatever/what about/
<Mr_Wrong> hcf:  Hard decisions must be made. If project members can't agree well enough to move ahead, the project needs to split
<Mr_Wrong> And most importantly, someone eventually needs to decide how things should be, and say "This is how things are gonna be.  Either put up or shut up
<Mr_Wrong> "
<Mr_Wrong> It doesn't matter whether that's done democratically or autocratically,  but it needs to happen, there needs to be a final decision on the current outstanding issues
<Mr_Wrong> The nice thing about democracy is that it guarantees that the majority of coders will be satisfied with the decisions that are made
<Mr_Wrong> Since the project isn't really anything without coders  :)
<Mr_Wrong> Of course, if you're going to be democratic you have to decide who your 'citizens' are.  It doesn't make any sense to have people vote who aren't valuable to the project
<Mr_Wrong> Am I babbling?
<hcf> no, just echoing pretty much what i think
<Mr_Wrong> Plus, I got the impression Fare doesn't really want to do the management tasks of leading the project.  I think he wants to design and code
<hcf> and write papers
<Mr_Wrong> That's a good thing too
<Mr_Wrong> So maybe someone needs to help him with the management stuff
08:20pm
<Mr_Wrong> Can someone write up pending questions that are awaiting answers for the project to move forward?
<s_r> Mr_Wrong: when do you think we'll see actual code for tunes?
<Mr_Wrong> Some people think there's actual code right now
<Mr_Wrong> Given that the project has been around a couple of years, I don't know that we'll ever see code for it
<s_r> Mr_Wrong: do you count Retro?
<Mr_Wrong> s_r:  I haven't seen Retro yet
-:- s_r [sr@phila-dialup026.nni.com] has left #tunes []
<Mr_Wrong> Have you guys looked at Spin?
<Mr_Wrong> That's a dynamic extensible OS
<Mr_Wrong> Perhaps relevant to this project
<Mr_Wrong> I don't know where the word dynamic came from there.
<Mr_Wrong> It just popped out of my fingers... It's just an extensible os
<hcf> spin, the uni of wash. project?
<Mr_Wrong> Yes
<hcf> "SPIN OS from University of Washington, which allows applications to migrate code in kernel space, implementing new functionality or replacing default policies." --tunes os review
<hcf> thats all thats said about it
<Mr_Wrong> That's the basic idea
<hcf> (in the review)
<Mr_Wrong> Though I think the exokernel ideais better
<Mr_Wrong> Keep everything in user space and use libraries rather than servers
<Mr_Wrong> There's also VSTa, which is quite cool and actually boots  :)
<Mr_Wrong> I think it would be quite easy to build Tunes on top of that
08:30pm
<Mr_Wrong> I like their design methodology
<Mr_Wrong> STa Design Methodology 
<Mr_Wrong>     If you can put it in a library, put it in a library.
<Mr_Wrong>     If you can put it in a server, put it in a server.
<Mr_Wrong>     Only when you have tried and failed both of the above, put it in the kernel. 
-:- Kaufmann [Kaufmann@200.224.105.168] has joined #Tunes
<Mr_Wrong> Hi Kaufmann
<Kaufmann> My Internet connection is unbelievably slow today.
<hcf> hoy Kaufmann
<Kaufmann> You can actually see the individual bits as they flow through the cable.
<hcf> 2secs, not bad
<hcf> Kaufmann: whats ur hp url?
<Kaufmann> on top of that - isn't it like illegal for an ISP not to have 24x7 staff?
<Kaufmann> hcf: have none
<Kaufmann> I'm waiting until there's a more intelligent way to have a 
<Kaufmann> "net presence".
<Kaufmann> Actually, that fits in nicely with my global distributed AI information system idea.
08:40pm
<hcf> Kaufmann: how will ur 'archive of lore on code obfuscation' be available?
<Kaufmann> hcf: I'm working on it... right now, only problem is I don't have a place to put it
<Mr_Wrong> How the hell does water even get up in the morning?
<hcf> Kaufmann: tril probly wouldnt mind hosting it
<Mr_Wrong> His head is so full of crap I don't know how he works out which foot to put on the floor first
<Mr_Wrong> It sounds like he took a cursory look at a bunch of ideas/writers he thought were cool and now thinks he's smarter than everyone else
<Kaufmann> I'm thinking of coding it up so that it'll be "annotateable" (bad word) by readers, like tpp.dhs.org
<Mr_Wrong> I doubt he's even read godel, he's probablu just read _Godel Escher Bach_
<Mr_Wrong> Or read/heard someone mention him
<Mr_Wrong> Sorry, I was just scrolling back to look at what happened after I left last night
<Kaufmann> Gödel is kewl
<Kaufmann> Oh BEAUTIFUL, now Slashdot forgot my freaking preferences. No wonder it's taking hours to load
<Kaufmann> At least in Slashdot I know "Edit preferences" is in users.pl?op=preferences
<Kaufmann> This is something that I want to fix in INOM
<Kaufmann> By the way, INOM is my project for a high-level model for universal Internet objects
<Kaufmann> You use an OO interface for individual servers/sites. So, for instance, in Perl you do something like:
<Kaufmann> use INOM; $this_site = Site->new("my.site.com");
<Mr_Wrong> Shouldn't accessing an object be independent of where that object resides?
<Mr_Wrong> I would think you'd just use some sort of name service to get a handle and access the object with the handle
<Mr_Wrong> ala CORBA
<Kaufmann> The class Site would provide plenty of predefined methods, so you can do something like: $this_site->search("edit preferences");
<Mr_Wrong> CORBA already provides all that sort of stuff
<Kaufmann> Mr_Wrong: yes, universal name services would also be included in the INOM package (using Perl as an example).
<Mr_Wrong> Is it a new model or just a library of functions for accessing existing services?
08:50pm
<Kaufmann> Mr_Wrong: CORBA is specialized toward applications and interoperability. The purpose of INOM is for programs to have access to a consistent high-level interface to the Internet.
<Mr_Wrong> Ok, so it's a library
<Kaufmann> I see the resemblance to CORBA, though.
<Mr_Wrong> It would be cool if object handles in a distributed object system could also be web pages and the like
<Mr_Wrong> So a URL would just be another kind of handle
<Mr_Wrong> It seems like the web is already on the way in that direction, but the lack of a standardized location service keeps it from getting all the way there
<Kaufmann> But the WWW/HTTP model wasn't designed with this sort of thing in mind... it's more free-form
<Mr_Wrong> Right, but there's no reason the use of http couldn't be made transparent to the user
<Kaufmann> Then again, I might be going in the direction of Xanadu... but who knows
<Mr_Wrong> The problem I see is that URLs are location-specific
<Mr_Wrong> URNs were what I was reading about where you have a unique name for any resource and you look up in a registry to find out where that resource is located, including all mirrors and the like
<Mr_Wrong> Kind of like a DNS for resources rather than just hosts
<Mr_Wrong> Now *that* would be awesome, particularly if it were decentralized
<Kaufmann> Anyway, I think that INOM and this kind of thing are much more rational systems for information transfer on the Internet than the Web
<Mr_Wrong> <--- anarchist, misses the days of UUCP addressing
<Mr_Wrong> I don't think you're being ambitious enough :)
<Kaufmann> You want ambitious? Lemme tell you about MECCA.
<Mr_Wrong> Cool, tell me
>>> hcf [nef@me-portland-us111.javanet.com] requested PING 931922533 from TUNES
<Kaufmann> MECCA (MEmetic Culture/Catalog/Archive) was derived a few weeks ago from NODE (Node-Oriented Database Engines), which I had designed some time before as a general framework for, appropriately enough, node-oriented database engines - essentially smart storage systems inspired by symbolic AI, who would try to mimic the way the human mind learns, through representations of nodes (pieces of data - photos, text, etc.) and "keywords" (who perform most symbolic 
09:00pm
<Mr_Wrong> You're talking about semantic networks
<Kaufmann> NODE is still alive, but MECCA is much cooler as a conceptual whole
<Kaufmann> Mr_Wrong: kinda
<Mr_Wrong> So what does it do?
<hcf> Kaufmann: that description got cut off at 'perform most symbolic'
<Kaufmann> It's really in embryonic phase, and not really fleshed out, but the idea is that it'd be an AI, run as software on a distributed network of volunteer computers (like SETI@home or distributed.net) that would serve as an archive and culture (think "bacteria culture") of memes.
<Kaufmann> hcf: "(who perform most symbolic work)"
<Mr_Wrong> What would be its function?  I mean, what would I do with it?
<Kaufmann> It'd have several input and output interfaces... parts of it would be constantly traveling the Web, lurking at IRC channels, etc. to "learn new things" - acquire and modify memes. And it'd likewise have Web and IRC interfaces (among other media), so that you could talk to it like you talk to anybody - a really resourceful anybody, but nonetheless
<Mr_Wrong> How would you keep its SNR high?
<Mr_Wrong> i.e. how do you filter out the crap?  There is a *lot* of crap out there in the world.
09:10pm
<Kaufmann> It doesn't have one specific purpose - it's just meant to be a memetic culture (somewhat like the old Frisco Community Memory). But, for instance, there could be an avatar of MECCA in this channel... if you wanted to know, eg, who Mr_Wrong is, you'd say "MECCA, who is Mr_Wrong?" It'd look for the name in its storage system, find data about you (that it'd acquire through the Web, or through reading your posts on IRC, etc.), resolve conflicts through contex
<Kaufmann> (cont'd) and return an answer in English (or whatever language you happen to be talking in)
<Kaufmann> Wrong, in regard to this question - how do people do that?
<Mr_Wrong> They build up filters through experience
<eStormy> Kaufmann: i had something similar in mind for my os
<Mr_Wrong> And from what their mom tells them
<Mr_Wrong> Wow, everybody has an OS.
<eStormy> Kaufmann: though not as well thought out  (c:
<Kaufmann> Exactly. So we'd give the system a mechanism for associating "positive" and "negative" impulses with memes or sets thereof, and let it learn from there
<Kaufmann> LOL Mr_Wrong
<Mr_Wrong> Personally, I think OSes need to get to the point where nobody needs to care what OS they're running
<Kaufmann> I think "my" OS is Tunes
<eStormy> yes, you can all be compatible with mine.  (c;
<Mr_Wrong> Actually, I expect the OS to become fairly integrated with the hardware
<Mr_Wrong> Microsoft is thinking about doing something like that with PCs to try to kill Linux
* Kaufmann/#Tunes finally understands that the string at the end of eStormy's posts is actually a genetically altered smiley.
<Mr_Wrong> But a lot of PDAs already come with the OS in ROM
<eStormy> i have a big nose
<eStormy> at least virtually
<Mr_Wrong> They have surgery for that, you know
<eStormy> my nose irl is normal
<Kaufmann> Mr_Wrong: but an OS that is not modifiable through traditional means (coding) nor reflective has its usefulness very limited
<Mr_Wrong> Kaufmann:  Well, I was thinking more along the lines of the kernel itself
<Mr_Wrong> And when computers are little disposable things and you can have ten of them on your person, I don't think the extensibility of the OS is going to matter all that much
<Kaufmann> Mr_Wrong, but isn't the idea to be kernelless?
<Mr_Wrong> All software will be free, but it won't matter because you'll have to buy the appliances to run it on
<Kaufmann> You're probably right
<Mr_Wrong> Kaufmann:  Well, I was thinking move the kernel down to the hardware itself and don't put any abstractions in it
<Kaufmann> Then again, a bit later we'll all be uploading ourselves
<Mr_Wrong> It may get to the point where general purpose hardware will be owned by hackers/developers only and the general consumer will have just appliances
<Mr_Wrong> And then it will be a question of closed/open specs
<Mr_Wrong> rather than closed/open source
<Mr_Wrong> That's just a random thought, anyway
<Kaufmann> I have an interesting (I think) article about how PCs in general are all wrong for regular people
<Mr_Wrong> What do you think regular people should have?
<Mr_Wrong> I think people in general are far too aware that they're using a computer
<Kaufmann> I think that if the industry hadn't become so obsessed with PCs, it would have evolved enough in the direction of appliances like you're describing that general-purpose computers and the such would become hacker tools
09:20pm
<Mr_Wrong> They have to put too much thought into using the computer and not enough into what they're actually trying to accomplish
<Mr_Wrong> Exactly.  Maybe that's where we're headed
<Mr_Wrong> Because of what I just said
<Kaufmann> That's what I think too. And it's proven scientifically too.
<Kaufmann> Computers often result in decreased productivity
<Mr_Wrong> And that';s why Linux will probably never get where Windows is now
<Mr_Wrong> But then again, Windows won't stay where Windows is
<Mr_Wrong> So we need to be going after *that* market, not the PC market
<Kaufmann> That's what my article says... that the UI should be ubiquitous
<Mr_Wrong> We need to flood the market with OSS appliances... people don't need to know they're using OSS, just us developers.
<Kaufmann> Mr_Wrong: agreed! In fact, that's my biggest problem with Linux - the whole "Linux for the masses" movement
<Mr_Wrong> We *need* to beat MS to the punch
<Kaufmann> Now let me find the bloody article...
<Mr_Wrong> Well, it's nice to try to get popularity, but by the time we really start encroaching on Windows, Microsoft will already have its appliances out
<Mr_Wrong> In fact, I'll bet they're working on it right now
<Kaufmann> Popularity is good, but the whole notion behind the state of "personal computers" today is bad
<Mr_Wrong> Microsoft is gonna kick our asses in the appliance market because they have Intel in their pocket.
<Mr_Wrong> We need to keep courting Motorola or someone like that
<Kaufmann> Besides, the UNIX model kinda becomes problematic in the appliance computer model.
<Mr_Wrong> Kaufmann:  Microsoft likes that  :)
<Mr_Wrong> There is no reason to stick with UNIX
<Mr_Wrong> I suppose the other thing we could go after is the server market, but those are all gonna be appliances too
<Kaufmann> Mr_Wrong: yep, true... hardware will become very important again... and so will the people who control the hardware industry
<Mr_Wrong> We need waaay more projects like the uCsimm
<Kaufmann> There is no reason to stick with UNIX... then again, Windows is a much worse solution
<Mr_Wrong> We need to make our own PDAs too
<Kaufmann> We need more OSS alternative projects, not just Linux
<Kaufmann> Hmm... PDAs are kewl
<Mr_Wrong> Linux is just a means to an end
<Mr_Wrong> And unfortunately it's winning now because it's becoming less relevant
<Mr_Wrong> Specifically because ot the appliance thing
<Kaufmann> In fact, I think that a Tunes-like system would be very good in an appliance computer
<Mr_Wrong> I was thinking that today myself
<Mr_Wrong> OTOH, we always have the PalmPilot
<Mr_Wrong> I guess we don't really need to design our own appliances
<Mr_Wrong> But we *do* need to keep close to the manufacturers of such appliances
<Mr_Wrong> Holy shit, I made my machine swap
<Mr_Wrong> I have 256 megs of RAM
<Mr_Wrong> X is using LOTS of ram
<Mr_Wrong> Probably because I have pixmaps coming out my ass
<Kaufmann> Tunes would be really cool in a WC... with a THUMBS (isn't that the name?) glove, a HUD and a 3D UI
<Mr_Wrong> X is using 184 megs  :)
<Mr_Wrong> Is there a way I can look at resource usage within X?
<Kaufmann> I dunno
<Kaufmann> Mr_Wrong: you could ask MECCA something like that...
<Mr_Wrong> Ok, where do I go to ask?  :)
<Mr_Wrong> Show me the code, baby
<Kaufmann> It'd provide you with a technical description and some links on the Web
-:- SignOff eStormy: #TUNES (eStormy has no reason)
<Kaufmann> GOD, is my connection slow
09:30pm
<Mr_Wrong> I suppose I shouldn't mention I have DSL to my house.
* Kaufmann/#Tunes starts to cry
<_QZ> Mr_Wrong: what X server?
<Mr_Wrong> XFree86 3.3.3.1
<Mr_Wrong> It's probably a problem with E or something
<_QZ> same here and im only using 48meg
<Mr_Wrong> I was using a pixmap theme in Gnome along with transparent terminals
<Mr_Wrong> And a pixmap theme in E
<_QZ> E is only using 4meg
<Kaufmann> ARGH.... Slashdot is downloading at 44 bytes/second.
<Kaufmann> Heh... tomorrow I get MacOS X Server
<Kaufmann> I'm going to call up my ISP tomorrow and see how much a private line would cost
<Kaufmann> Does anyone else like the cartoon "Dr. Katz"?
<Mr_Wrong> Damn, I'm gonna restart X, brb
-:- SignOff Mr_Wrong: #TUNES (leaving)
09:40pm
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #tunes
<Kaufmann> wb
<Kaufmann> Mr_Wrong: ever watched "Dr. Katz"?
<Mr_Wrong> Yeah, once
<Mr_Wrong> It was this client that was using all the memory
<Mr_Wrong> I have a feeling it stores the entire scrollback as a pixmap or something
<Mr_Wrong> And I had a hell of a lot of scrollback
<Kaufmann> I really like this cartoon
<Kaufmann> CC stuff tend to be surprisingly good
<Mr_Wrong> Yeah, I've noticed that
09:50pm
<_QZ> Mr_Wrong: what client
<Mr_Wrong> irssi
<Mr_Wrong> DSL is nice when you're looking at tons of screenshots
<Mr_Wrong> Hmmm, it kind of sucks that you need to store pixmaps for GTK.  The gradient generator is one of the nice things about BlackBox
<_QZ> Mr_Wrong: so what kinda os are u looking for?
<Mr_Wrong> One that doesn't limit me with meaningless abstractions
10:00pm
<_QZ> do u want it to be small and fast?
<Mr_Wrong> I want it to be able to be small and fast.
<_QZ> so u dont want tunes then
<Kaufmann> QZ: Tunes should be able to be small and fast too
<_QZ> Kaufmann: tunes requires its language to run
<_QZ> along with its source code
<Kaufmann> Tunes users should be able to use the abstractions that they feel most comfortable with
<_QZ> Kaufmann: do u know how tunes works?
<Kaufmann> QZ: which doesn't necessarily translate into a problem... not if you KISS
<Kaufmann> QZ: it doesn't! It's still in design phase ;)
<_QZ> if u are running an mp3 player and editing text then tunes will compile the source code for both programs into a single program
<Kaufmann> So?
<_QZ> part of program A, part of program B, part of program A, ...
<Mr_Wrong> That sounds kind of dumb, since there is no reason for the mp3 player's code to be combined with the text editor's code
<_QZ> that means it can only be as small as the source code and the language
<Kaufmann> Also, I'd like to point out that most programs would be very small if it weren't for the UI
<_QZ> Mr_Wrong: that is how it handles multitasking
<_QZ> Mr_Wrong: it doesnt preempt code it merges code into a single stream
<Mr_Wrong> Most modern processors already provide hardware for multitaking.
<Mr_Wrong> Multitasking, too.
<_QZ> Mr_Wrong: fare thinks hardware multitasking is way to slow
<Mr_Wrong> Well, I suppose you could look upon it as a hack for not having dynamic compilation
<_QZ> if done right the tunes method could be faster
<Mr_Wrong> Yeah, but it means you won't be able to run any native code on the system at all
<Mr_Wrong> You'll have to write everything in the tunes language
10:10pm
<_QZ> right
<Kaufmann> No, with emulation
<Mr_Wrong> Which means any deficiencies in the tunes language will be deficiencies in the OS
<_QZ> the language does extensive type/bounds checking to make sure the code wont crash the system since it uses no memroy protection at all
<Mr_Wrong> The world can't even agree on a common language now]
<_QZ> when u first run a program the language will quickly compile it with no optimization
<Mr_Wrong> Hmmm, that does sound like an interesting protection model
<Mr_Wrong> But definitely a bitch to debug
<_QZ> as the program runs the language will recompile it with better optimization
<Mr_Wrong> Yeah, Self does the same thing
<_QZ> the longer it runs the better it optimizes
<Mr_Wrong> Of course, you'll run into a problem on 32 bit architectures, because you won't be able to have more than 4GB of virtual memory because everything will use one address space
<Kaufmann> Why does one need 4GB of virtual memory?
<Mr_Wrong> Hmm, as I recall Bill Gates said something similar about 640K
<Mr_Wrong> We have machines with 8GB of *physical* memory
<Kaufmann> Mr_Wrong: I'm sorry. What I really meant was: Why does one need 4GB of virtual memory on a 32-bit machine?
<Mr_Wrong> I guess it won't be a serious limitation anyway, since people will move more toward 64 bit
<Kaufmann> xactly
<Mr_Wrong> On the other hand, I'm a hacker.  I want something I can play with on microcontrollers as well
<Mr_Wrong> Tunes will be replacing hardware with a very complex compiler
<Kaufmann> Most of the techniques used in systems programming are hacks to go around other hacks... the problem with not having a good design in the first place
<Kaufmann> LOL Mr_Wrong
<Mr_Wrong> And it will have to recompile code to thread it into the stream every time I start a new application
<_QZ> Mr_Wrong: brix can be used on any processor
<Mr_Wrong> What was the URL for brix again?
<_QZ> brix?
<abi> i think brix is an OS at http://www.qzx.com/brix or not a vacuum or not suckiful.
<Mr_Wrong> thanx
<Mr_Wrong> Woo, lots of documentation
<Mr_Wrong> Too bad I can't read any of it
<_QZ> pick the pages that u wanna read the most?
<Kaufmann> I'll probably be moving away from these issues... I'd rather be studying something more fun, like AI
<_QZ> s/?//
<Mr_Wrong> I just want a kernel I can standardize on so I don't have to worry about kernel development any more
<_QZ> just a normal uk?
<Mr_Wrong> And not something that forces me to use servers to run my OS, because, as Fare has pointed out, context switches are expensive
<Mr_Wrong> Actually, I want an exokernel
<Mr_Wrong> Then when we finish the exokernel we can start doing the real stuff
10:20pm
<Mr_Wrong> VSTa is close enough to an exokernel for me
<Mr_Wrong> Maybe that's what I should be working on
<Kaufmann> I want assemblers, nano-component neural-network AI, uploading for the masses, and eternity in cyber-Nirvana
<Mr_Wrong> I want something I can actually get in a reasonable amount of time
<Kaufmann> Not me. I'm a dreamer.
<Kaufmann> MECCA will probably not even exist until the middle of the next decade
<Kaufmann> Oh yes - I also want a rational revolution
<Mr_Wrong> VSTa's kernel is 40k
<Mr_Wrong> and supports everything.
<Mr_Wrong> And I think the solution to the problem of hardware multitasking's being too slow is to fix hardware multitasking.
<Mr_Wrong> Single address space OSes are one step in the right direction
<Mr_Wrong> Though they don't solve the problem of memory protection... you still have to keep track of which pages are protected in a given context
<Mr_Wrong> BRB, my indicators of activity on channels has broken
-:- SignOff Mr_Wrong: #TUNES (leaving)
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #tunes
10:30pm
<_QZ> Mr_Wrong: some of the info pages are now working
<Mr_Wrong> Cool
<_QZ> ignore the functions on the kernel page
<_QZ> the real functions are at http://www.qzx.com/brix/api.cgi
10:40pm
<Mr_Wrong> Why do you have register names in your API definition?
<_QZ> cuz it passes arguments that way
<Mr_Wrong> But the registers on other architectures will be different
<_QZ> the language takes care of that
10:50pm
-:- SignOff hcf: #TUNES (Leaving)
<Kaufmann> Well, I'll be going now. Buh-bye.
-:- SignOff Kaufmann: #TUNES ("Nationalism is an infantile disease. It is the measles of mankind." - Albert Einstein. For more information regarding the good and evil of nationalism, contact Fare at OpenProjectsNet #Tunes. (Heh.))
11:20pm
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
[msg(TUNES)] newlog 1999.0714
IRC log ended Wed Jul 14 00:00:01 1999