IRC log started Wed May 17 00:00:01 2000 [msg(TUNES)] permlog 2000.0517 -:- water [water@tnt-10-61.tscnet.net] has joined #tunes -:- SignOff eihrul: #TUNES ([x]chat) -:- SignOff I440r: #TUNES (brb) -:- mark4 [mark4@purplecoder.com] has joined #tunes hi mark -:- mark4 is now known as I440r oh hi dood thanxd for the tellin me about tcn... he cool :P i was really pissed off earlier shhh dun tellhim i sed that... np :) go tohis head :P heh 12:40am neway nite ppl :) k * water/#tunes is impressed by the detail and quality of flightgear 12:50am abi: flightgear? fare: bugger all, i dunno 01:30am -:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #Tunes -:- SignOff Kyle_L: #TUNES (Leaving) -:- ult [ult@user-38lcn61.dialup.mindspring.com] has joined #Tunes humhumdeedum -:- SignOff ult: #TUNES (Leaving) 05:20am -:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #Tunes -:- NetSplit: varley.openprojects.net split from tolkien.openprojects.net [06:28am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net] -:- Netjoined: varley.openprojects.net tolkien.openprojects.net -:- abi [nef@bespin.dhs.org] has joined #tunes -:- thomas [thomas@193.217.63.152] has joined #tunes -:- coreyr [coreyr@net255ip95.parklink.com] has joined #tunes -:- Fare [rideaufr@lantier.enst.fr] has joined #tunes -:- AcidReign [wade@ruby.cqu.EDU.AU] has joined #tunes -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- smoke [smoke@16dyn28.delft.casema.net] has joined #tunes -:- water [water@tnt-10-61.tscnet.net] has joined #tunes -:- I440r [mark4@purplecoder.com] has joined #tunes -:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #tunes -:- NetSplit: forward.openprojects.net split from irc.linux.com [07:16am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [forward.openprojects.net] -:- Netjoined: forward.openprojects.net irc.linux.com -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- SignOff water: #TUNES (Ping timeout for water[tnt-10-61.tscnet.net]) -:- water [water@tnt-9-246.tscnet.net] has joined #tunes Fare: flightgear is the gnu flight simulator 09:20am -:- hcf [nef@207-172-225-56.s56.tnt1.pld.me.dialup.rcn.com] has joined #tunes re bri is down i see or r we split dunno, the topic is desync'd here once again not for me 09:40am -:- hcf has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System: http://www.tunes.org/ || Slate Language: http://www.tunes.org/~water/slate-home.html -:- zhlubbish [nlevitt@bacon.ummu.umich.edu] has joined #tunes hi zhlubbish hello there did you want to talk about something, zlub? no, just came to see if anybody was having interesting conversations hm not currently finger zhlubbish heh hops oops rather you're right not currently 09:50am ah wappinger's falls i guess i'll have to look elsewhere adios, everybody -:- SignOff zhlubbish: #TUNES (Leaving) 10:00am -:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #Tunes re kyle h i ! anything to discuss? I don't think so. yeah that's what i thought :P 10:20am -:- NetSplit: lackey.openprojects.net split from irc.linux.com [10:26am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [lackey.openprojects.net] -:- Netjoined: lackey.openprojects.net irc.linux.com -:- I440r [mark4@purplecoder.com] has joined #tunes re 10:50am -:- SignOff water: #TUNES (The Tao went that-a-way!) -:- dalvarez [dalvarez@tristan.sc.cs.tu-bs.de] has joined #tunes -:- plonk [plonk@cpt-dial-196-30-184-98.mweb.co.za] has joined #tunes -:- NetSplit: forward.openprojects.net split from irc.linux.com [12:32pm] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [forward.openprojects.net] -:- Netjoined: forward.openprojects.net irc.linux.com -:- I440r [mark4@purplecoder.com] has joined #tunes -:- SignOff dalvarez: #TUNES (Leaving) -:- SignOff plonk: #TUNES ([x]chat) * Fare/#Tunes is back yay aya! 01:10pm -:- eihrul [lee@usr5-ppp187.lvdi.net] has joined #tunes -:- SignOff hcf: #TUNES (Ping timeout for hcf[207-172-225-56.s56.tnt1.pld.me.dialup.rcn.com]) -:- hcf [nef@207-172-225-57.s57.tnt1.pld.me.dialup.rcn.com] has joined #tunes -:- vvater [water@tnt-10-149.tscnet.net] has joined #tunes -:- vvater is now known as water Fare: you there? hm abi: beep Fare i beeped Fare abi: beep Fare i beeped Fare hey not necessary (i already did it ;) he likes multiple beeps *likes*? does it turn him on or something? makes it easier to determine where theyre from oic * water/#tunes looks over yesterday's conversations again 02:30pm -:- tcn [Tom@cci-209150250052.clarityconnect.net] has joined #tunes re tcn hey how ya been? not bad, i'm on vacation for 2 weeks i also imposed a public deadline on myself for slate specs well, not exactly public, but oh well i'm still debating with myself about the merit of deadlines well, this is a simple deadline and i have a lot of time and resources right now to finish it true as long as you got the willpower :) hey.. ya? I mean heh one of our customers set a deadline for Monday because he wants this thing done before he leaves for vacation lol so wouldn't ya know, we hit a major snag today i hate those kinds of deadlines good for you ;) deep-seated problem, needs lots of work.. hm so have you and l440r really hit it off? poor fellas :) ahh, similar projects.. yeah i noticed just happened to be on the same side of the argument ok did you see our discussion about software-based protection? 03:00pm i skimmed it think it's workable? i'd have to look at the conversation again sorry i've been thinking only about HLL stuff for the last week hmm afaik, though, sw-based is supposed to be faster and safer than hw-based it really requires higher-level code doesn't it? * Fare/#Tunes is back well it needs code that isn't as low-level as c :) hey fare as in, no arbitrary memory-access.. Fare: i'm going to have a lot of info on slate to publish to the tunes list in a week or so tcn: sure but there are varying degrees of that (yes! squeak can now share code directly across the net, object-by-object, changeset by changeset!) cool yeah 2.8 will have this object-oriented mail system (as well as other new apps) water: good Fare: i'm working out the whole source of the confusion yesterday over namespace access, but the solution will likely be a bit unorthodox, though much more tunes-friendly * water/#tunes has been toying a bit with Maude lately, testing ideas 03:10pm water: what will the spec likely entail? the basic object mdoel, syntax, evaluation specification, primitives, and the set of standard meta-objects percentage-wise, how much have you completed of the spec, either on paper or in your head? well the problem is that i'm trying out a bunch of different ideas at the same time but to give an estimate: mmm say 60% how that "unorthodox"? ideas that have to reconciled with eachother? everything on the website is 90% final no, different ways of doing the same thing Fare: well, it's not immediately obvious unless you think in terms of tunes ideas, which most people don't, even on Tunes mlist unorthodox solution, how? well my ideas on it aren't final yet so you dont want to say something you might renig on? or confuse people who don't need to be confused :) i'll basically say that higher-order equational rewrite is really cool, and leave it at that 03:20pm if you want to know more, look at maude abi: maude? it has been said that maude is a reflective rewriting logic language at http://maude.csl.sri.com/ or http://www.csl.sri.com/~duran/ * tcn/#tunes is away (working on retro3) water: why not have static lexical scoping? tcn: good tcn: publish it! heh fare: I did well anything dynamic in slate can be made static by not letting things touch it hm tcn: great? s/?/!/ * Fare/#Tunes goes retro.tunes.org * water/#tunes does too * Fare/#Tunes sees that its.html has been replaced by its.txt btw, Alan Bawden has a new address. abi: Alan Bawden? Alan Bawden is an ex-MIT LISP hacker at http://www.bawden.org/alan/ thx heh which re-directs to the old site :) no, the site has moved http://achilles.bu.edu/bawden/ ? the lispm will be at bellingham in a around a month ok it costs a bit more than I initially thought how long will it be there? maybe I should have bought the one from england; a bit more expensive, but better and nearer water: until I get my PhD and come fetch it or until no one at Bellingham wants it anymore oh good, then it'll be there for quite a while :) or forever, if you fall in love with it and don't want to part how much for it? er... how much did you pay? current price, shipping&handling&everything included is around $1500 how old is it? it must be a nice workstation 03:30pm how old is it? coreyr: I took the MacIvory model II board. what is linear naming? yes, Fare, why don't you explain linear naming? :) I think the Ivory I is 1987 water: linear naming? I think Alan Bawden already explains it well in his PhD thesis Fare: to you. heh :P i meant for the channel as i for the people who don't read phd theses like other people read the comics s/i/in/ -:- SignOff smoke: #TUNES (znurk) well, a linear name is just a name with a contract that only one entity will manipulate it the name will be used once only and can be recycled after its first and unique use. so what does that mean for the user/programmer of the system? where is the contract at? coreyr: in languages such as Clean (with uniqueness typing) or Haskell (with monads), it's statically checked by the type checker in many compilers, it is inferred by abstract interpretation i dont comprehend the benefits (which is essentially the same as type reconstruction; just that such types need not be human-manipulable) yes, fare, explain the benefits, please coreyr: in a distributed system, maintaining non-linear names is VERY COSTLY even in a concurrent system, coherently accessing a non-linear resource means expensive locking how does it relate to tunes? Fare: and when ur done explain this add it to the tunes glos s/ain/aining/ it basically means that naming is always done by contract, and you explicitly and constructively share the names as necessary, instead of making access to names totally open, so that you have to be paranoid about how the name gets used even in a centralized system, non-linear resource access means that complex protocols might have to be designed as for resource management accross modules 03:40pm Fare: so how about a glossary entry? Fare: and while ur at it, write entries for all the terms in the glossary's todo section even in a non-modular one-developer centralized system, linear constraints can bring compiler optimizations that won't otherwise be possible water: great idea * Fare/#Tunes copy-pastes from the channel well, actually there is already an entry about linear logic, but I'll enhance it right Fare: while ur in an update mode, update the hll specs well, just minor fleshing out of the text would be good enough there the formal lambda-notation for the ideas can wait since it's not trivial water: can you edit the HLL specs to flag the points you seem mistreated? and insert questions in it? pretty much all of them heh Fare: and when ur done w/ the glossary and hll specs, feel free to move onto the todos for other sections well cvs access is unavailable to me :( :( yeah tell me about it, it sucks uh? how that, unavailable to you? well, only my windows installation can connect to the net, and i tried wincvs and it failed miserably and combing through cvs manuals does *not* sound like fun dont you have a shell on bespin? did you try the wincvs-compatible ssh? i didn't know there was one yeah, putty can let you edit files on bespin from a losedoze host corey: no thanks, vi can take a hike putty rocks. hm Google tells me about http://pyrite.cfas.washington.edu/pipermail/orca-devel/2000-February/000336.html btw, on putty use the terminal option " font has X-window encoding" 03:50pm ok this is going to take me a bit to figure out -:- SignOff tcn: #TUNES (ircII EPIC4-2000 -- Accept no limitations) of course, LOADLIN is an enabler god damn it! 04:00pm just great i don't suppose anyone has a local copy of the slate docs? :) u rm'd 'em? no, one of the pages got replaced with a working scratch copy and i didn't notice it until after the backup :( 1min hmm heh it's the 'detail' page if i had a decent version manager, i wouldn't have to mess with this water: how that "local copy" ? water: wincvs ? water: bespin/tmp/slate-detail.html heh you never listen, do you? hcf: thx hum, one symbolics video card was named the "FrameThrower" :) ok fixed 04:10pm * Fare/#Tunes bounces a message from Peter Van Roy to review@ -:- nuria [nuria@44-MALA-X11.libre.retevision.es] has joined #tunes what's a tune? hi hi, so... hmm anyone care to field that one besides me? :) nuria: do you program? a little, why? (yes this is relevant to your question) tunes is about a computing system based on a reflective programming language the language is high-level, but it can dive into different parts of its implementation and interface in order to adapt itself to system-level tasks and what advantages does it take? or you can modify the language to fit the particular kind of programming or using you happen to be doing "take"? abi: tunes well, tunes is a free reflective computing system at http://www.tunes.org or for programming languages what the internet is to networks or like flypaper for hallucinogenic programmers or discombobulated or obnubilated or AOTA hm xcuse my english * water/#tunes fixes that stuff at the bottom of abi's factoid * Fare/#Tunes is back nuria: in another context, a tune is also a little melodic fragment of music the advantage is that you don't have to rely on a core group of guru programmers to adapt your system to how you like it ahhh have you ever used lisp? is it declarative programming? 04:20pm it can be, but not necessarily i have studied smthing, but never used it reflection allows you to change the semantics of the language and even what the language is for the most part well, lisp has good reflection abilities have you looked at the tunes review of languages? where? abi: review? review is http://www.tunes.org/Review/ those pages link to all sorts of programming languages and oses hm anyway hm 04:30pm where's that new glossary entry, Fare? im planing a project about interacting humans and computers, have you got any sugestion in language terms? HCI stuff? you mean programming languages, of course, i suppose yes well there are VPL's like ToonTalk and ProGraph, or user interfaces like Morphic what aspect of HCI are you most interested in? languge understanding natural language understanding?? what's the most advanced reading you've done on the subject? :( great hmmm im just a 3th course computer science student... what does "3th" mean? 3rd i mean what kind of parser work have you done? do you get into knowledge-representation issues at all? :( i was hoping to find some kind of help here.... well if i were you, i'd look into ATN's, parsers, and KR theory ok, let's see do a search on google ATN? heh google? abi: paip? i think paip is "Paradigms of Artificial Intelligence Programming, Case Studies in Common Lisp" by Peter Norvig at http://www.norvig.com/paip.html you don't know what google is? sorry www.google.com ok i believe sicp also has stuff on ATN's, but is more programming and reflection related 04:40pm abi: sicp? sicp is probably Structure and Interpretation of Computer Programs at http://mitpress.mit.edu/sicp/ there's code there for ATN's thnx, but what is ATN? ATN's are very basic, though. for real natlang parsing you should look into some more advanced stuff water: new glossary entry on linear logic? yes! but what should i start with? it depends on what you already know water: what do you think I'm doing that prevents me from looking at the irky conversation here? you probably don't know enough about parsing yet, so i suggest looking at parser information Fare: ok Fare: i guess i'm too used to doing all sorts of other non-Tunes-related things :P s/to doing/to you doing/ nuria: ATNs are a specific way for building parsers that work in limited situations for natural languages there was some url that BillT posted to the mlist about natural language parsing, but i don't recall when or what it was does sil.org cover this poop? hm a little, the "computing" link on their homepage takes you to interesting things http://www.sil.org/computing/hermitcrab/ 04:50pm i find it all very interesting and i thank u all i'm sure you do np hope to see u soon -:- nuria [nuria@44-MALA-X11.libre.retevision.es] has left #tunes [] bleh i take it most cs undergrads are like him or no? 05:00pm -:- SignOff hcf: #TUNES (Leaving) hm really quiet now hi ppl hey * water/#tunes listens to some b-sides by Moby i heard you were discussing sw-based memory protection the other day, l440r well tcn was.... oh i dont think it was workin out too good... huh? imnot sure its as practicable as all that... what wasn't working out? software based memory protection dude sw-based memory protection has already been done many times yes. i know 05:30pm so what do you mean? it slows you down ALOT not necessarily we were trying to come up with a way thtat didnt slowe u down very much there's this one os coding page i read that talked about how you could actually attain *greater* speeds than hw-based mem-protection unfortunately i don't remember where it was at that depends on the processor. pmode on a x86 is not faster than real mode i do remember though that it required a specific kind of strategy for implementing it well, x86 just sucks all around and unless you know what you are doing you can actually generate SLOWER code yeah i know because any time you do a 16 it operation in pmode it will take an override. * water/#tunes recalls learning pmode intricacies back in '93 pmode is a bad solution to a stupid design flaw i avoid pmode. like i said, the x86 just sucks all around its not needed in application software... it sure does hehe gimme motorola any day :) heh well i just code in HLLs lately, so i've lost my edge with low-level stuff brb, rebooting to new kernel ( i hope :) -:- SignOff I440r: #TUNES ([x]chat) -:- I440r [mark4@purplecoder.com] has joined #tunes dammit hm? kernel-2.3.99-pre6 is a fscking PAIN!!! heh * water/#tunes sticks with old reliable 2.2 :) x86 has nice features, tho: byte-wise addressing, transparent i-cache coherency, etc. it gets as far as decompressing kernel then HANGS have to hit reset yeah but they don't make up for the fact that optimizing for x86 is a rela pain s/rela/real/ I440r: hardware memory protection also slows down a LOT yes i know however, in a kernel like linux that is acceptable. 05:40pm _with proper static analyses_, software protection is better and faster otoh we are a HLL croud, so we don't like unrestricted pointers in code although without such analyses, dumb hardware protection wins. i.e. if you want arbitrary unsafe C programs, you can't beat hardware protection. hardware willbe doing that job for ever more. coders arent interested in writing good code any more. they want RIGHT NOW not right and in C if you want an OS in C, for C, try Linux or BSD. * water/#tunes likes "right now" in smalltalk :) Tunes is not for you. i run linux exclusivly erm almost... or lisp, but common lisp seems to enjoy vexing me i installed nt so i could do work stuff from home water: I hope the LispM will change your mind i hope it will, but there's nothing like the LispM without the hardware currently well written code doesnt write outside its own memory any way. water: a $5000 OpenGenera/Tru64 license? eek that much for a decent lisp environment? I think that MCL, Allegro CL, LispWorks, are around the same order of magnitude for their full versions, too yeah but they honestly don't offer anything like what Squeak does a LispM does yes i know and the non-LispM try