IRC log started Mon Dec 27 00:00:01 1999 <_ruiner_> a 14 year old kid The problem I found with brix is that it's almost entirely scripted, no compiled software. Would make performance hurt _ruiner_: Hehe, our hardware's being developed by a 50-year old kid <_ruiner_> even if he doesn't pull it off, he's somebody to keep a close eye on....lots of ambition and the smarts to back it up * Downix/#tunes nods eihrul is someone else I'd keep an eye on [notice(TUNES)] permlog 1999.1227 -TUNES(tunes@bespin.dhs.org)- permlog 1999.1227 <_ruiner_> I don't know much about him Not much older than lar1, has already began, abandoned, restarted an OS project * _ruiner_/#tunes has oodles of ambition, and supposedly the smarts, just not the motivation nor the guts to make it happen <_ruiner_> hmmmm..... _ruiner_: My problem is a overly cautious nature. * Downix/#tunes has taken 27-months to get as far as he has from initial concept to the point where the hardware's spec is mostly done sound system and I/O still not done <_ruiner_> thats not a fault as far as I see it <_ruiner_> wait....are you making this to be almost like a console? almost I'm designing it to be flexible, change the interface ASIC and you can turn it from a PDA into an Internet Appliance to a workstation <_ruiner_> errrrrrr...... * _ruiner_/#tunes doesn't like that idea Well, the ASIC and a few availible add-ons <_ruiner_> stifles innovation it's not the same chips in any case, only "sister" chips Like, for a PDA you'd use the "Loki" chipset it has a narrower bus design and lacks the expention capabilities 12:10am while for a workstation you'd want "Sigfried", which has a higher-capacity memory bus, multiple accelerator CPU expention, etc. inside they're the same cores, just certain parts made availible to the outside world Saves development costs Ok, brb back What do you mean by stifles innovation? <_ruiner_> from other hardware manufacturers 12:20am How so? We're not a "major" company by any means, not Intel or anything of the like <_ruiner_> I mean....if the sound card is built into the motherboard, how is say creative labs going to release a new sound blaster for your computer? _ruiner_: Simple, plug it in. <_ruiner_> right....so you need to make it as easy for other parties to make stuff for you <_ruiner_> I mean, ahhhhh...... <_ruiner_> going that route eh? _ruiner_: We have already worked on that. We'll still support PCI, PCI-X and the like <_ruiner_> don't let isa die! _ruiner_: Plus have a few new options that most standard manufacturers don't have access to yet ISA has a place with us <_ruiner_> I'm just a fuddy duddy....I still wish for the days of dos to come again _ruiner_: Say Creative Labs wants a real "in" on our system. Instead of making a generic PCI card, they instead access our "sound expantion pins" which allows a third-party DSP to hook up with our internal sound system We've taken a "we've got it, but we're not perfect" approach with everything Everything can be augmented or even by-passed if need be except the memory control ok, brb 12:30am <_ruiner_> gotta go....later -:- SignOff _ruiner_: #TUNES (destroy what destroys you) anyone alive? yep ok hcf, what did you think of my explination of Eddas? i wasnt paying much attention scroll up then. hehe I think that's the first time I ever explained it's expantion system 12:50am i'm not into hw Ok hcf: What are you into? os/pl pl? programming language? damnit, abi should of had that anyway, yes Ok Well, would such a system be interesting to work with? i guess ok, what do you like to work with? 01:00am i dont kno ok 01:10am -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us121.javanet.com]) -:- SignOff Downix: #TUNES (BitchX FTP Site -- ftp://sunsite.unc.edu/pub/Linux/system/network/chat/irc) -:- SignOff colorg: #TUNES (Ping timeout for colorg[humbubba.smart.net]) -:- SignOff eihrul: #TUNES ([x]chat) -:- AlonzoTG [Alonzo@216-164-129-62.s316.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes -:- SignOff AlonzoTG: #TUNES (Have Nice Day :)) -:- SignOff Fufie: #TUNES (Leaving) !irq:*! drow cracked crack.linuxppc.org! !irq:*! congrats! -:- Fare [fare@quatramaran.ens.fr] has joined #tunes !NickServ:*! netgod used GETPASS on bm -:- lambda [cwz@cogeco-88-83.cgocable.net] has joined #tunes okay I've got a scheme compiler working to elf to elf? linking format? great yeah it's "Scheme" but it's actually my friends' it doesn't even support pairs yet 09:00am chmod +x it and run it it was compiled with it good can I peek at the compiler source? not yet it's basically all my friend's on his cvs and he doesn't want it out yet it's really minimal it has no (cons) no run-time environment it's sort of like assembly with lisp syntax well a bit higher-level than that I might write a little arithmetic evaluator with stop/copy gc in it today 09:20am or perhaps an rpn calculator ok 09:30am -:- AlTeRnAtIvE [AlTeRnAtIv@ppp89.greco.com.br] has joined #tunes hi all -:- AlonzoTG [Alonzo@216-164-129-62.s316.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes Any programmer here ? I kan code to some extent... bbiaf 10:20am -:- SignOff AlTeRnAtIvE: #TUNES (Leaving) if you want an OS you *might* want coders 10:30am yup -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes gakuk hello Fare 10:40am ole sami 10:50am -:- hcf [nef@me-portland-us341.javanet.com] has joined #tunes * AlonzoTG/#tunes hurls a laser guided flaming woodchuck at hcf !is:*! someone tell me why gline is broken .. -:- Kaufmann [Kaufmann@200.224.105.179] has joined #tunes hey gang 11:20am I'm seriously considering getting a Nino 300 next year R$120 a month, for 12 months But I'll only do it if I know that I'll be able to extirpate WinCE from it -:- Kaufmann is now known as KaufmannBRB -:- KaufmannBRB is now known as Kaufmann so what is tunes? exactly exactly what? it's the greatest vapourware of all times Yeah abi, tunes? tunes is a free reflective computing system at http://www.tunes.org or for programming languages what the internet is to networks or like flypaper for hallucinogenic programmers or discombobulated or obnubilated abi, tunes is also none of the above i'm sorry, but that's too long, Kaufmann it's a project to integrate all software into a one reflective platform argh a platform designed to unify software abi, tunes is also AOTA okay, Kaufmann. The Perl motto is "There's more than one way to do it." Tunes is also very cool Tunes adds "but all these ways are ultimately the same." *applause* ugh that's disgusting so is perl though 11:30am Eh? what's disgusting? look at perl it has the syntax from hell * AlonzoTG/#tunes works lambda over with a 2x4 Perl is good very expressive when I first used perl I thought it was a joke perl is expedient -:- ginge [user5856@212.114.33.149] has joined #tunes -:- ginge [user5856@212.114.33.149] has left #tunes [] sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f}); the above is a Perl quine -:- hcf has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System || Towards an Open System || perl lovehate open A, 'a'; $a = 0; while (){$a = $_ || 0; $a++;} close A; open A, '>a'; print A $a; close A; print $a; exit; it's also impossible to compile the above is a web page access counter (courtesy of yours truly) lambda, perlcc map { $_->[0] } sort { $a->[1] cmp $b->[1] } map { [$_, $f->($_)] } @_; the above is the Schwartzian Transform for a given function $f also courtesy of yours truly @hash{@arr1,@arr2}=(2)x@{[@arr1,@arr2]};map{$hash{$_}--}(@arr1,@arr2);@foo=grep{!$hash{$_}}keys %hash; (define (fact n) (if (= n 1) 1 (* n (fact (- n 1))))) om the above code looks for the intersection of two arrays * AlonzoTG/#tunes tries to figure out the mysteries of mov ax,1 nope, it's now movl $1, %eax * AlonzoTG/#tunes smashes Fare with an Iron Butterfly lambda, a better way is (define (fact 1) 1) (define (fact n) (* n (fact (- n 1)))), by way of pattern matching. fact = fact = foldl (*) 1 . enumFromTo 1 ack Kaufmann: well I was using Scheme a better way is oh well... I'm trying to write my OS... fact = foldl (*) 1 . enumFromTo 1, of course I bet all of you have boot-sectors... =P I need one. =/ smkl, no fair... that's not recursive :) no you don't need it to be recursive it's better: it's primitive recursive lambda, /of course/ you don't need it to be recursive. That's not my point. try this (define (f n) (let loop ((a 2) (b 1) (c 1) (count n)) (if (zero? count) a (f (+ a (* b 2) (* c 3)) a b (- count 1))))) calculates f(n) = n if n < 3 and -:- Downix [down@d-gnaps-175.ici.net] has joined #tunes hey-ya all f(n) = f(n - 1) + 2*f(n - 2) + 3*f(n - 3) otherwise Kaufmann: I've been doing a bit of research into multimedia kiosks for you. Of course it doesn't have to be recursive. sub fact { my $a = 1; map { $a *= $_ } (1..$_[0]); return $a } Downix, cool 11:40am Anything I should know? Kaufmann: How many units will they need, how soon will they need the bid in by, and when would they be installed? (define (fact n) (let loop ((a 1) (accum 1)) (if (= a n) accum (loop (+ a 1) (* accum a))))) Downix, heck, it's the government... nobody has any idea yet! Kaufmann: Well, I figured out how to turn the old blueprints I have into a working machine, as well as how I could drop my own chipset in to replace the blueprints at any point. here is a problem I ambition for tunes: write a program that understands the above specification, a dumps an optimized version, an explained human-readable version, and a proof that all these versions are equivalent. okay, okay, okay... let's cut out the coding competition... obviously lambda's language of choice is Scheme Fare, I've actually been considering the problem of proving the equivalence of two functions (true? (last-statement 'Kaufmann)) ==> #t ;) Kaufmann: reduce them both to lambda calculus Kaufmann: assuming both languages are lambda calculus complete How is Scheme anyways? I talked to a compiler developer yesterday and he said that HLL's do not deliver the same performance as C or straight ASM actually in real systems fftw? i think fftw is at http://www.fftw.org/ -- proves that monads are faster than ASM! a good lisp compiler can beat a c compiler lambda: He was pointing out that there are few good HLL compilers out there. I'm working on one as we speak Oh? well mainly one of my friends is but I'm helping for lisp-like languages lambda, the question is: how do you prove that the iterative and recursive versions are equivalent? It may seem obvious to us - both can be easily expanded to versions that multiply a list from 1 to the number - but a computing system will need either strong AI or a whole lotta heuristics. Kaufmann: reduce them both to lambda calculus lambda: I am concerned over a language to include with my system, that's why I'm asking smkl, FFTW is great! lambda: in what meta-formalism? lambda, not a complete solution you could then do it much like you prove two mathematical functions to be equivalent !is:*! any services admins awake ? alonzo church, before any programming language, devised the idea of a true calculus of functions a function is a rule for computing a vlue !^lilo:*! I'm awake yes, I know, I know dammit well then there's the answer use standard AI techniques automatic theorem proving etc. [msg(tunes)] permlog 1999.1227 the techniques are well-known what I'm asking for is the suitable metaformalism of course 11:50am metaformalism? I've got Norvig here with me Kaufmann: As soon as you get the information, I could post a solution. the problem is Downix, thanks for the help, man this technique won't necessarily work for a lot of languages Downix, believe me, I'm trying to find someone who knows all this... but Rio de Janeiro is one confused city Kaufmann: Anytime. (I could use the publicity anyways) lambda, but in Tunes they can always be transformed into an appropriate representation of the algorithm brb -:- SignOff Downix: #TUNES (Read error: 3.1415926535 (Excessive pi)) even if the language is lambda calculkus complete you can't necessarily convert all code in the language into lambda calculus you can always transform the function from one language to another you'd need a language that is complete for every other language types aren't the same in all languages etc. perl is based on strings only a question of representation. Everything done in Tunes will be extensible and self-modifiable by nature, so this won't be a concern well if I plot f(x) = sqrt(x) on a graph if I do it with real numbers I don't actually have a complete plot but if I do it with complex numbers I do nu? -:- eihrul [lee@usr5-ppp63.lvdi.net] has joined #tunes btw perlcc isn't *really* a compiler :P lambda, but it yields compiled Perl code ;) And That's What Really Matters no it simply bypasses the parsing stage of the perl interpereter :P perl can be compiled to perl machine perl is slow perl will always be slow :P anyway, I'm leaving. All I have to say is: Tunes is good. Buh-bye. it's impossible for perl to not be slow -:- SignOff Kaufmann: #TUNES (kathyanne gets kick'd / as leaves fall on still waters / she is here no more. (KILL ALL THE EVIL PERL-HATERS!!!)) you can always change the specs of perl 12:00pm then it isn't perl if you'd be larry wall it would be look at tcl dynamically-scoped has (or had) no actual system of arithmetic it parsed numbers from strings each time lisp was dynamically scoped too well yeah then guy l steele came along and made it lexically scoped Scheme has always been lexically scoped I believe (since steele and sussman made it) dynamic scoping was a Bad Thing (tm) the ACM gave him an award for getting rid of it WOOHOO!!!! I HAVE BINARIES!!! 8 bytes of the first sphere pre-alpha!!! =) have you read the paper about implicit parameters? can't say that I have smkl: worth reading? i think it's quite interesting smkl: url? !scoop:*! lilo! it was at haskell mailing list well i'm not on it 12:10pm so if you can direct me to somewhere where i can read it :) and it's archives aren't updated :( and perhaps to where i can subscribe to the haskell mailing list didn't know there was one -:- Downix [down@d-gnaps-269.ici.net] has joined #tunes there smkl: ? eihrul: haskell.org this is pretty sick what is haskell? i heard haskell was lazy purely functional programming language at http://www.haskell.org/ i can't find any references to the paper, want DCC? hehe ok what is scheme? scheme is a dialect of Lisp that stresses conceptual elegance and simplicity or ((nearly) (perfect)) damn url ripoff abi, url? lambda: wish i knew what is Downix? you are doing it right now lol what is sex? somebody said sex was the LISP Symbolic EXpression (latter renamed SEXP or S-EXP) lol what is love? love is expensive what is gay sex? lol! abi, gay sex is PERL abi, what is gay sex? gay sex is PERL hahahaha abi, forget abi lambda: I forgot abi abi, abi is a complete moron! abi? lambda? what is abi? i am probably a complete moron! rofl abi, Larry Wall? lambda: i don't know abi, Larry Wall is an evil-Lisp-hater who must be exterminated at all costs lambda: bugger all, i dunno abi, Daleks? downix: bugger all, i dunno abi, Daleks are an evil species of salt n pepper shakers hehe Daleks? Daleks are an evil species of salt n pepper shakers abi is a stupid perl hack :P lambda: url ripoff? its the same software hcf: yeah I know, I just know oz I was in #macdev back in the day oic smkl: you can dcc me paper? 12:20pm eihrul: yeah do so :) -:- windsk8r [mibin@62.11.104.25] has joined #tunes windsk8r: go away. we don't trade mp3s hehe ;) main(){fork();main();} -:- windsk8r is now known as mibin 12:30pm smkl: interesting -:- SignOff mibin: #TUNES (Ping timeout for mibin[62.11.104.25]) 12:50pm -:- SignOff lambda: #TUNES (Ping timeout for lambda[cogeco-88-83.cgocable.net]) abi: comonads? bugger all, i dunno, eihrul eihrul: http://www.cse.ogi.edu/~dick/dick.html's papers thanks may be some other useful things in alta's 42hits for comonads 01:10pm -:- mibin [mibin@62.11.104.112] has joined #tunes -:- Fufie [stig@tunnel-44-14.vpn.uib.no] has joined #tunes hi 01:20pm lo hi hey -:- SignOff mibin: #TUNES (Ping timeout for mibin[62.11.104.112]) grr why are theseing ups tracking numbers not working? no idea 01:30pm code-generation is tricky.. :-/ what are you writing? 01:40pm doing work on the weaver/code-generator for my hypermedia-system 01:50pm -:- SignOff Downix: #TUNES ([BX] Chester Cheeta uses BitchX. Ayeuhayueuhayueuh!) Fufie: aop type of thing? 02:00pm yep still haven't gotten the stinky lisp books (next day air), at 3 days and counting it's been lispmas.. cannot expect things to move too fast gah! but i paid for next day air! -:- mibin [mibin@62.11.103.198] has joined #tunes somewhere, over the raindow s/d/b/ 02:10pm Fare: can you give me a simple definition of a comonad and its usefullness? :) abi, comonad? mibin: bugger all, i dunno comonad? ahem, what's it? heh Fare: i was reading a paper on dynamic scoping smkl sent me Fare: it mentioned how comonads can be used to model the effects of dynamic scoping in the conclusion much as monads can be used to model computation though didn't give much of a reason why... is the paper freely available? 02:20pm -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us341.javanet.com]) . . 03:50pm whatcha doing, alonzo ? I'm responding to a message on my new OS mailinglist. =) I had a bunch of people I was keeping in a netscape draft... but the header was getting up to 12-15 members so I coppied them all into an E-groups mailinglist... (It was time.) ahh, that draft folder the list is sphereos. is it worth joining? i am already in many mailing lists, but if this one has an high signal/noise ratio ... ahh, it is probably a private mailing list no, its completely open... If you don't join you are free to go in and read the archive. om ahh cool 04:00pm -:- mibin_ [mibin@62.11.103.198] has joined #tunes -:- SignOff mibin_: #TUNES (main(){fork();main();}) om -:- SignOff AlonzoTG: #TUNES (Have Nice Day :)) 05:40pm -:- SignOff mibin: #TUNES (Ping timeout for mibin[62.11.103.198]) * Fare/#tunes is back * eihrul/#tunes is still trying to grok comonads. eihrul: where from? eihrul: http://www.cse.ogi.edu/~dick/dick.html's papers 06:00pm abi: comonads are at http://www.cse.ogi.edu/~dick/dick.html 06:10pm -:- Kaufmann [Kaufmann@dial709.infolink.com.br] has joined #tunes FOO! Hey guys, I finally broke the +2 default score treshold! Congratulate me -:- air [brand@p0wer.qzx.com] has joined #tunes 06:50pm my free talk & learn alphabet cebter arrived today and is more fun than i thought it would be :) LOL Congratulate me, air, for I am now one of the Slashdot Elite and what does that mean? * eihrul/#tunes futilely tries to regrok monads. air, for one thing, it means I now have a +2 default score slashdot elite? isn't that kind of an oxymoron? LOL Fare eihrul, got the literature with you?