IRC log started Sat Jan 15 00:00:02 2000
[msg(TUNES)] permlog 2000.0115
<water> oh yeah, <> aren't the right symbols
<water> doh!
<water> fixed
<water> darn, i'm getting a bit tired
<hcf> whats slated ;) for after the bnf?
<water> heh
<water> actually the syntax and semantics will probably get finished up simultaneously
<water> there are issues that affect both
<hcf> dont u need whitespace in list?
<water> although most of the message-passing semantics will be a functional version of self's, the adding of beta-style patterns affects both syntax and semantics
<water> oh yeah
12:10am
-:- SignOff air: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
<water> i'm checking out the semantics specs on some related languages, icuc
12:40am
-:- SignOff water: #TUNES (Leaving)
-:- SignOff hcf: #TUNES (Leaving)
-:- AlonzoTG [Alonzo@216-164-129-165.s419.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
-:- up2nite [up2nite@dnai-216-15-42-98.cust.dnai.com] has joined #tunes
-:- SignOff up2nite: #TUNES ([x]chat)
-:- Downix [down@d-gnaps-188.ici.net] has joined #tunes
-:- ircusr [ircusr@212.68.72.98] has joined #tunes
<ircusr> greetings
<Downix> hey
06:20am
<Downix> almost done
<ircusr> ?
<Downix> wrong window
* Downix/#tunes rocks to Limp Bizkit
06:30am
<Fufie> rock
06:40am
-:- SignOff Downix: #TUNES ([BX] Back wit anutha one of doz BitchX-rockin' beats!)
-:- binEng [Anders@j141.ryd.student.liu.se] has joined #tunes
-:- hcf [nef@me-portland-us813.javanet.com] has joined #tunes
<hcf> ho hum
<binEng> di dum ;)
-:- SignOff ircusr: #TUNES (bah not much chat in here)
07:20am
<eihrul> lo
<binEng> hi
<eihrul> gah, i unsubscribed from the squeak mailing list
<eihrul> and i'm still getting mail
<binEng> bad luck :)
08:10am
<eihrul> you sift through 100 mails, 9 of which you actually want to read! :P
* binEng/#tunes is away
08:20am
-:- fire [no@209-68-229-157.dialup.cust.tfb.com] has joined #tunes
<eihrul> lo
<fire> hi
<eihrul> what's up lately? you still dinking around with lengua?
<fire> yeah :)
<fire> my brother got a book on programming language design
<fire> so im reading up
<eihrul> Fufie: btw, an object is not a function with a state.... since a function already has a state (or else you wouldn't need a stack)
<eihrul> cool
<fire> i've been playing around with the idea of a language without side effects but it seems impractical
<fire> what do you think?
<eihrul> well, what about printing...
<eihrul> if you didn't have *some* side effects
<eihrul> you wouldn't get any feedback!
<fire> yes
<fire> and if you wanted to store info
<eihrul> you'd need side-effect free hardware
<fire> etc...
<Fufie> eihrul: boiling down to objects being functions? ;)
<fire> well i am thinking of how only some things can use side effects for certain purposes
<fire> with it cleanly in lengua
<eihrul> Fufie: well, as i see it, an object is just a function where its state persists after the function has executed
<Fufie> side-effect free universe as well
<eihrul> Fufie: which can be exhibited by free variables in Common Lisp
<eihrul> which i've actually exploited to implement my CL demo :)
<Fufie> I think closures are another good example of objects as functions
<eihrul> too bad CL doesn't provide any way to access the current closure from within the closure :)
08:30am
<eihrul> (which was the hardest thing i had to simulate)
<Fufie> is the CL demo released to the public?
<eihrul> rmm, its nothing fancy, and really sloppy, but its public domain i suppose :)
<Fufie> I'll have a look later
<Fufie> right now I am in the middle of a release
<eihrul> it just makes a hash table
<eihrul> where each element is a lambda that accesses a variable from the enclosing lambda
<eihrul> (with the effect that the hash table keeps the variables free)
08:40am
<eihrul> Fufie: hey?
<eihrul> do you still have that walker?
* eihrul/#tunes could use it for something.
<Fufie> yes.. give me a few minutes here
<eihrul> thank you
<eihrul> (i need it to optimize in the thing demo :>)
<eihrul> to see if anyone's actually using the variables and if i need to actually define them or not
09:00am
<Fufie> I found three walkers but never got around to decide which one to go for.. 
<Fufie> I'll put them in a directory for easy download for you
* eihrul/#tunes bows.
09:10am
<Fufie> www.ii.uib.no/~stig/eih/
<Fufie> I have yet to do more than sporadic checking though
<eihrul> "A simple code walker," yeah right!
<eihrul> :P
<Fufie> simplicity is overrated ;)
09:20am
<fire> bah
<fire> simplicity is next to godlyness
<eihrul> Fufie: hmm, i have two lisp books left... "Common Lisp, Ed 2" or "Paradigms..."... which first? :)
<Fufie> the last one
<Fufie> CLTL2 is nice, but mostly as reference material
<Fufie> PAIP rocks
<Fufie> fire: then my statement should be as expected from an agnostic
<eihrul> k
<eihrul> fire: there's various ways to offer simplicity
<eihrul> and many are worse than others....
<eihrul> s/worse/*much* worse
<Fufie> simplicity is often confused with elegance.. something elegant may look simple.. but usually is the result of hard work
-:- AlonzoTG [Alonzo@216-164-128-26.s26.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes
<eihrul> Fufie: then would not elegance be conceptual simplicity? :)
<eihrul> or at least the derivative of simplicity
<Fufie> I am not sure I agree with myself enough to answer that :)
<eihrul> i.e. the simplicity of a given piece of code
09:30am
<eihrul> as it contributes to the simplicity of the work as a whole
<Fufie> but when looking at a sleek sports car or a beautiful marble statue it looks simple.. it sppeals to us.. but it is not simple at all
<Fufie> appeals
<eihrul> i never said simplicity is always simple to create!
<eihrul> simple is merely simplicity as it appears to the viewer, in the way i use it
<Fufie> yes.. simplicity is in a way the looks of something elegant
<eihrul> neglecting abstractions, anyway
<Fufie> but something simple doesn't have to be elegant.. it tends to be crude and unfinished
<eihrul> (like the body of the car, the shape of the statue)
<eihrul> Fufie: and so often that is very true
<eihrul> which is why i said:
<eihrul> there are various ways to offer simplicity, and many are *much* worse than others :)
<Fufie> elegant code often looks simple.. in a painful way.. you look at it and realise how simple the given problem can be solved and really wish you came up with.. it's like a glove that fits perfectly
<eihrul> yep
<eihrul> but i think of it the other way around, or that its atleast a circular definition
<Fufie> and the programmer who is unaware of the tao will think that it is simple. a programmer who is at one with his tao knows that it is the result of many many failed attempts
<eihrul> simplicity may contribute to elegance
<Fufie> something elegant is the result of love and caring from the designer or programmer,... and skill of course :)
<eihrul> since elegance is just a generalized notion of overall simplicity and other aspects
<eihrul> Fufie: i think you're using the meaning of simple too literally
<eihrul> i see there being many *forms* of simplicity
<eihrul> contributing to a given work
<eihrul> you have overall simplicity, simplicity at the incremental level, conceptual simplicity, etc
<Fufie> I am merely andbarely trying to provoke a reaction :)
<eihrul> none of these meaning the other is necessarily present
<eihrul> and these together tend to contribute to the 'elegance' of code
<eihrul> and just because a given idea is simple
<Fufie> ack.. all the fuss needed to get a release out is frustrating
<eihrul> doesn't mean its necessarily simple to develop (which is another kind of simplicity)
09:40am
<eihrul> most every simple concept we have was hard to develop :) or took a really long time to do so
<eihrul> eh?
<Fufie> I think the quest should be for elegance and simplicity will come along, rather than chase for simplicity and end up with something crude made in visual basic ;)
-:- Ghyll [karltk@mp-217-219-77.daxnet.no] has joined #tunes
<eihrul> well, i consider elegance to be defined as the right mix of simplicities, man
<eihrul> so it doesn't matter which way you approach it
<eihrul> so long as you know the right recipe for the various components :)
<Fufie> you should aim for elegance.. aiming for simplicity might give you something crude and ugly
<eihrul> i don't chase for maximal simplicity in everything
<eihrul> that just leads to complexity
<Fufie> complexity is ok
<eihrul> (and inelegance)
<eihrul> not of everything!
<eihrul> if everything is complex, then it can't be elegant
<eihrul> the generalized notion of complexity would mean everything is complex :)
<Fufie> most things I deal with are complex :(
<eihrul> poor fufie :)
<Ghyll> Fufie: some free association: complex -> complex numbers -> imaginary -> imaginary worlds -> fufie
<eihrul> human -> hooman -> star trek
<eihrul> s/human/-> human
<Ghyll> eihrul : haha
<eihrul> speaking of which
<eihrul> i need to proon some incremental complexity from this code :)
<eihrul> er prune
<eihrul> too much irc is bad for the spelling....
<Ghyll> 'proon' reminds me of a nasty word..
<Ghyll> (while on the subject of free association..)
<eihrul> no words are nasty
<eihrul> that's just their connotation :P
09:50am
<Ghyll> eihrul: I find that some words sound nasty. Especially words in foreign cultures where they seem to bend their tongues twice behind their ears to make that obnoxious sound. but you might well argue that neither the sound nor the spelling of a word is the word itself..
<eihrul> which is what i argue...
<eihrul> (i don't like to use 'nasty' when describing a word)
<eihrul> maybe its just because 'bad' or 'nasty' words are such a taboo in the united states
<eihrul> that i happen to not agree with that i would argue such
<Ghyll> eihrul: but don't you make a word irreproachable by such arguments ?
<eihrul> eh?
<eihrul> a word is a word, man
<Ghyll> eihrul: I mean, if somebody finds a word flawed, you can always argue that 'no, it's the connotation that's fucked'
<Ghyll> eihrul: or 'no, you're just not satisfied with the spelling'
<eihrul> so long as you have some way to separate denotation from connotation, i don't see that as a problem
<eihrul> if they can prove that it's merely connotation as opposed to denotation, then change whichever one is broken
<eihrul> but it is more the denotation that i am concerned with
* binEng/#tunes is back
* Fufie/#tunes is also back
<fire> bin!!!
<eihrul> but, many americans (my parents included)
<eihrul> tend to compeltely ignore the denotation and have no understanding what the connotation is for
<binEng> hi :)
<eihrul> and merely punish on unfounded reaction to the connotation
* Fufie/#tunes find some words immoral and bad
-:- lar1 [larman@sdn-ar-015casfrMP094.dialsprint.net] has joined #tunes
<lar1> Hey
<eihrul> if you asked a given person...
<eihrul> more than likely they would have no valid reason why 'fuck' or 'shit' or 'ass' or what-not are immoral or bad words
<eihrul> and even what a 'bad' or 'immoral' word is
<Fufie> a word like "pokemon" is immoral
<Ghyll> eihrul: but isn't that a result of them not knowing what 'connotation' is about, at all ?
<binEng> Hmm, what is it with the four elements and ppl around here... air, fire, water... ice, anyone?
<eihrul> Fufie: no, not the word, its use has been immoral though, yes
<eihrul> Ghyll: yes... its a meta-connotation
<Fufie> so is a word like "microserf", it's not like the workers there have the rights of a serf.. it's more like "microslave"
<eihrul> Ghyll: that's the only one they remember any more, the connotation of the connotation, and not the connotation or denotation themselves
<eihrul> all they know is that it 'probably' infers 'something bad'
<Fufie> eihrul: the immorality of the pokemons spring from the combination of the letters.. it is the word itself which is bad and evil and eat ham with jam on
<Fufie> some letters when put together can create big evil
<binEng> what is so bad with pokemons?
-:- SignOff AlonzoTG: #TUNES (Read error to AlonzoTG[216-164-128-26.s26.tnt1.lnhva.md.dialup.rcn.com]: No route to host)
<Fufie> like h and i and t and l and e and r
<lar1> Ahh!  BinEng!  Are you feeling ok?
<eihrul> Fufie: shrug, another taboo in america
<eihrul> if i merely speak the word 'hitler' to my parents
<eihrul> in respectful or irrespectful context
<eihrul> i am lashed at
<binEng> lar1: hi! Sure I'm ok, any reason why I should not?
<eihrul> too many people using not using denotation (a name of a person)
<binEng> lar1: or was it just to be nice
<eihrul> s/using not using/not using
<lar1> binEng: <binEng> what is so bad with pokemons? I am just kidding with you ;)
* binEng/#tunes tickles lar1 with a feather
<lar1> eihrul:  You are old enough to not have the parent problem, right?
10:00am
<binEng> I hardly know what a pokemon is
<eihrul> lar1: i'm 18, and my parents still lash at me for saying either
<eihrul> and they don't have a single valid reason why
<eihrul> but they do it anyway
* lar1/#tunes always thinks eihrul is older even though he has told me his age a few times...
<eihrul> i think the world would be better off without such fanaticism :P
<eihrul> (though, perhaps not? who knows...)
<lar1> Then we wouldn't has powerful words like 'swear words'... when used in the proper context 'fuck' or 'shit' can carry meaning
<lar1> Although when used causally, they certenly loose their meaning
<eihrul> words aren't powerful
<eihrul> they don't deserve to be
<eihrul> words are a means of communication
<Fufie> afk
<eihrul> and if certain words may not be used to communicate
<eihrul> then they totally lose their expediency
<eihrul> and thus serve no purpose as words
-:- binEng is now known as ice
<ice> bah, registered nickname
-:- hcf has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System <http://www.tunes.org> || slate <http://www.tunes.org/~water/slate-home.html>
<lar1> Arent there cimumstances in this society where words _do_ need to cary power?
<lar1> Persuasion, etc?
-:- ice is now known as binEng
<eihrul> nope
<eihrul> words should not carry power
<eihrul> people should attach power to words if they so-choose
<eihrul> but the word itself should not have some objective meaning 
<lar1> Can sentences carry power?
<eihrul> no
<lar1> So "We are treated like shit."  and "We are treated poorly." mean the same thing to you?  
10:10am
-:- Kaufmann` [newbie@dial348.infolink.com.br] has joined #tunes
<Kaufmann`> Shalom!
<lar1> Shalom!
<eihrul> lar1: no
<eihrul> you are misinterpreting what i'm saying most likely
<Kaufmann`> no what?
<eihrul> Kaufmann`: look at the logs, long convo
<lar1> eihrul:  Yes... can you reword or somthing?
<Kaufmann`> oooh
<eihrul> lar1: i am merely saying that their should be no objective meaning of a given word
<eihrul> and that a person is free to interpret a word however they wish
<eihrul> but, that people interpreting words wrongly is a source of much of my troubles :P
<lar1> Hmm, ok, I see what you are saying
<eihrul> sometimes people take a word to be objectively 'bad'
<eihrul> when in my interpretation, i don't make any allowances for such
<lar1> Have you done work with fractals?
<eihrul> nope
10:20am
<lar1> Anybody?
<eihrul> Fufie: hmm, pruning some incremental complexity has pruned some conceptual complexity :)
-:- SignOff Ghyll: #TUNES (bar)
<binEng> lar1: How so?
<lar1> binEng:  Studing them?
<binEng> lar1: I've looked at it a little, but not enough to say I've actually done work with fractals.
<Kaufmann`> Okay
<lar1> binEng:  Ah... hmmm... would you say that somthing like a watermelon seed is a form af 4th dimentional fractal?
<eihrul> Kaufmann`: ?
<Kaufmann`> what is the difference between Joy and merely endowing Forth with combinators and some higher-level operators? Is it merely conceptual?
<binEng> lar1: uh, can't say
10:30am
<Kaufmann`> Boy, am I a conversation killer or what?
<binEng> looks no better, eh
-:- beholder [beholder@ppp-037.m2-1.sub.ican.net] has joined #tunes
<Kaufmann`> ?
-:- AlonzoTG [Alonzo@207-172-49-235.s235.tnt7.lnhva.md.dialup.rcn.com] has joined #tunes
* Kaufmann`/#tunes smites Alan with a mighty regexp
-:- SignOff hcf: #TUNES (Leaving)
-:- hcf [nef@me-portland-us738.javanet.com] has joined #tunes
10:40am
<Kaufmann`> well, I'm off
<eihrul> why?
<eihrul> you just got here, man...
-:- SignOff Kaufmann`: #TUNES (kathyanne gets kick'd / as leaves fall on still waters / she is here no more.)
10:50am
<AlonzoTG> om
* binEng/#tunes unoms AlonzoTG
-:- SignOff lar1: #TUNES (Ping timeout for lar1[sdn-ar-015casfrMP094.dialsprint.net])
11:00am
-:- Netjoined: devlin.openprojects.net irc.linux.com
-:- fire [no@209-68-229-157.dialup.cust.tfb.com] has joined #tunes
-:- beholder [beholder@ppp-037.m2-1.sub.ican.net] has joined #tunes
-:- NetSplit: devlin.openprojects.net split from asimov.openprojects.net [11:06am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [devlin.openprojects.net]
* binEng/#tunes is watching Monty Pythin (Life of Brian)
>>> AlonzoTG [Alonzo@207-172-49-235.s235.tnt7.lnhva.md.dialup.rcn.com] requested PING 947950184 from #tunes
-:- lar1 [larman@1Cust48.tnt26.sfo3.da.uu.net] has joined #tunes
<lar1> abi: seen air?
<abi> air was last seen on IRC 14 hours, 51 minutes and 16 seconds ago, saying: water: did u add any details? [Fri Jan 14 21:39:35 2000]
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
12:40pm
-:- water [water@tnt-10-201.tscnet.net] has joined #tunes
<hcf> water!
<water> yo. here for a few
<lar1> Hey water!
<water> wow. everyone excited? :)
<lar1> water:  You ever do any fractal work?
<eihrul> wb
<water> i toyed with it, programmatically, and i've worked with nonlinear stuff
<water> i mean, the math.theory
<lar1> Could somting like a watermelon seed be considered a 4th dimentional fracta?
* lar1/#tunes ducks and hides
<water> huh?
<water> yeah i guess it could represent one
<water> what aspect of  a watermelon?
<water> i assume you mean the coloring of the skin
<lar1> No...
<water> then elaborate
<water> quickly, cause i want to go relax in the city
<lar1> from the seed, grows the plant (a fractal inits self, but I degress), from that plant grows watermelson, which inturn make more seeds
<lar1> So is the seen not a fracatal iterated over time?
<water> hm
12:50pm
<water> yes in a way... and its chaotic to determine which seed goes on to the next iteration
<water> although that's impossible to prove either way, afaik
<water> anything more relevant? :)
<lar1> Not from me...
<water> anyone got info/questions/requests/gripes for slate or arrow?
* hcf/#tunes has urls to give
<water> review-worthy or just for me?
<water> i got the feeling tunes-review wouldn't appreciate 'em
<water> just pass 'em
<hcf> i tend to prefer to let u determine worthiness
<eihrul> water: i 'refactored' the code to make it an easier read :)
<hcf> http://www.transframe.com/transframe/papers/unify.htm http://www4.informatik.tu-muenchen.de/papers/BourdoncleMerz_Otio1996.html
<hcf> http://www.ccs.neu.edu/research/demeter http://web.comlab.ox.ac.uk/oucl/research/areas/ap/topics.html http://www.dfki.uni-kl.de/~boley/welcome.html
<hcf> http://www.dimi.uniud.it/~pietro/Papers/paper_arg.html http://www.dimi.uniud.it/SLP/Papers/papers.html
<water> eih: cool
<water> i looked at demeter a long time ago
<water> will look again, though
<eihrul> what's demeter?
<water> i think that i can actually make the arrow papers easy to grok using co-induction ideas
<water> hm. lots of papers to sift through yet again
<water> thx hcf
<hcf> np
<lar1> There is more then one arrow paper?
<water> well, there's a ready outline of the formal theory
<water> most of the writing for that will be busy-work
<lar1> ah
<water> the outline is online if you want it
<lar1> ok, I'll take a look
<water> it's off of my home page... the link is "technical"
<water> anything else?
<water> oh yeah, thx eihrul
<water> hm
01:00pm
<water> i'll try to get the web-group management issues sorted out when i get back tonight, including contacting trill for scripts, etc
<eihrul> for slate?
<water> yep
<eihrul> i still need to start reading all those papers :)
<eihrul> exams kept me busy all week
<water> that reminds me, though, i should contact the author of icct
<water> eih: np
<water> well, later all
-:- water [water@tnt-10-201.tscnet.net] has left #tunes []
-:- lispbliss [lispbliss@cont021p10.ont.micron.net] has joined #tunes
01:10pm
-:- SignOff lispbliss: #TUNES ()
-:- dalvarez [ircusr@212.68.72.98] has joined #tunes
<dalvarez> EHLO everyone
<lar1> Hey
01:20pm
<Fare> abi: icct?
<abi> it has been said that icct is Interaction, Computability, and Church's Thesis at http://www.cs.brown.edu/people/pw/papers/bcj1.pdf
<Fare> dalvarez: hi
<Fare> dalvarez: wanted to talk to me, yesterday?
<Fare> dalvarez: no more tmp?
01:30pm
<Fare> are you now prm?
<dalvarez> hey, Fare
<dalvarez> was away for some time at another tty
<dalvarez> right, not any more tmp, eihrul has told you, right
<Fufie> what happened to tmp?
<dalvarez> tried to reach you yesterday to finish what we had begun a day before where the LAN was unfortunately powered off at midnight
<dalvarez> now we finally find opportunity for that
<dalvarez> Fufie: well tmp means it is tmp
-:- smkl [sami@glubimox.yok.utu.fi] has left #tunes []
-:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes
<dalvarez> anyway else it wouldn't be tmp
<dalvarez> :)
<Fare> wassup
<dalvarez> well, I wanted to contribute to your tunes project cause I am fascinated of the, say "futuristic", concept of it
<dalvarez> yet I'm unexperienced but I'll give my best inside the bounds of my ability
<Fare> ok
<Fare> well, are you familiar with compilation?
<Fare> with writing run-time for systems?
<Fare> with developing web applications?
<Fare> with constraints solving?
<dalvarez> compilation in the sense of putting together different parts of code or the traduction of source to machine processable form?
<Fare> translation
<Fare> although not forcibly all the way from source down to binary assembled code
<dalvarez> I think I understand the basic concepts of that like parsing and optimisation but have no practical exprerience
<dalvarez> ... of any kind
<dalvarez> well, and with web applications, I played a bit with the BSD sockets and stuff as described in R.W.Stewens UNP
<Fare> hum. Is there a subject of interest where you'd like to contribute?
01:40pm
<dalvarez> application porting
<dalvarez> ...while collecting experience with the underlying system
<dalvarez> later I could extend that to system programming
<Fare> unhappily, we currently have no core to which to port applications
<dalvarez> but for first I prefer starting with s.th. easy
<dalvarez> you mean no system at all?
<Fare> although you might like to get familiar with such systems as OCaml or MzScheme or Poplog with which we might share some general features (higher-order core, yet side-effects, and a module system)
<Fare> i mean no system at all
<dalvarez> As far as I understood the concept you're looking for s.th. like a mach mk
<Fare> dalvarez: except that we think that the mach got the abstractions upside down
<Fare> (see papers/Glossary/index.html#microkernel
<dalvarez> I recently downloaded it's source of the digital ftp server together with tanenbaums amoeba
<dalvarez> and have not yet begun to check it out
<dalvarez> for the first tunes might be completely theoretical but sure you have worked out programming interfaces, have you?
01:50pm
<Fufie> dal: it might be a good idea to try out a reflective language to get the idea of reflection
-:- SignOff hcf: #TUNES (Leaving)
<dalvarez> Fufie: I read a small part of the arrow concept about refection and am not yet familiar at all with this expression. Do you mean a language of self modifying code for metaprogramming 
<dalvarez> Fare: it noticed...
<Fare> self-modifying code, yes, but in a cleanly staged way
02:00pm
<dalvarez> what are you primarily using for developement
<Fare> good question
<Fufie> dal: I don't think the arrow system is not quite there as a full-features system to test with.. you might want to try out a lispy language, either Common Lisp or Scheme
<Fufie> s/don't//
<Fare> dalvarez: as you might know, the coding part of Tunes is near zero at the time
<dalvarez> I'm learning Prolog
<Fare> as some put it, it's still vaporware
<dalvarez> Fufie: ??
<Fufie> dal: yes, I guess the best languages to learn for a future tunes-system is common lisp or scheme
<Fare> or maybe also ocaml
<dalvarez> is it realistic by matters of performance?
<dalvarez> and features?
<Fare> ocaml is quite performant
<Fare> more than tunes will probably be for some time
<dalvarez> don't know it
<dalvarez> never heard
<Fare> abi: ocaml?
<abi> ocaml is probably a language from ML family with support for objects and modules at http://caml.inria.fr/ocaml/
<dalvarez> in what fields is it used?
<Fare> see hump.html there for sample applications
<dalvarez> will do so, but right now explain me what is to do
<Fufie> dal: feature-wise I have yet to see anyone beat Common Lisp.. as for performance a Common Lisp system ran the Deep Space One thing of NASA.. that was real-time.. Scheme is small and elegant and may not be too fast, but still is fast enough for a lot of work
<Fare> Fufie: read the paper about Stalin's global optimization techniques?
<dalvarez> ok, I'll take a closer look at the langs
<Fufie> fare: I have read it some months ago.. it is however non-standard Scheme and it assumes several things last I checked.. Scheme has some parts which are bound to be slow :(
02:10pm
<Fare> the most salient feature that enables the optimization is: whole-program analysis
<dalvarez> I subscribed to nearly all mailing lists offered on tunes.org to get a picture of the latest devel stage, but didn't receive not one single msg till now
<dalvarez> does the project lack participiants?
<Fare> no, it lacks participation
<Fare> which is slightly different
-:- SignOff lar1: #TUNES (Leaving)
<binEng> It lacks general direction and momentum
<Fufie> it lacks the core so people can get started
<dalvarez> so the concepts worked out have not yet reached a implementable degree at all 
<Fare> s/able/ed/
02:20pm
<Fare> a lot are implementable.
<Fare> None have been implemented.
<dalvarez> why not define some interfaces according the concepts and apply them to an existing mk, so you'll be able to change the implementation later
<binEng> I don't have the impression ppl know what they could sit down and code
<Fufie> bineng: a core is needed to experiment with.. otherwise we'll all try to kill each other's gedanken projects.. which is about as tasteful as religious wars in a bowl of fish soup
<binEng> yuk
<binEng> yes, I can agree a base for experimentation is needed
<Fare> dalv: one reason is, we think the conceptual core of the system is NOT the underlying runtime (much less a uK), but the code-translation management
-:- AlonzoTG [Alonzo@207-172-126-162.s543.tnt5.lnhva.md.dialup.rcn.com] has joined #tunes
<binEng> Why not stick to Joy or Self or something?
<Fufie> bin: neither is sexp-based
<binEng> So use Lisp
<Fufie> eihrul and water and hcf's slate thing seems more interesting then
<Fare> binEng: joy is a pure language, etc, etc
<Fare> binEng: about as usueful as a core Haskell w/o all the support
<Fufie> bin: I'd love to use Lisp, but Common Lisp might not be ideal.. EuLisp might be better..
<dalvarez> when do you plan to release a useable (*util*) version?
<Fare> binEng: I consider Self as a proof-of-concept, but hardly a language I would work with
<Fare> (would you work in pure pi-calculus?)
<binEng> Look, we're only talking about a base for experimentation here..
<eihrul> Fufie: what's novel of eulisp as compared to common?
<Fare> dalvarez: good question
<Fufie> dal: this depends.. if Fare is succesful with his vm-agenda things might happen and at the other side of summer, things bight work somewhat..
<dalvarez> in 2000 ?????
<Fufie> eihrul: it is simpler, yet more advanced.. it didn't need to be backward compatible.. it is also modular and has a hierarchical module-system
<Fufie> s/bight/might/
<eihrul> Fufie: how 'free' is the implementation?
02:30pm
<Fufie> eihrul: very.. it has not really been implemented in other things than a reference interpreter.. leaves more room for tunes to define it's own stuff.. changing CL might be more of a controversy than saying we use EuLisp and it's great.. 
<eihrul> abi: eulisp?
<abi> it has been said that eulisp is http://www.bath.ac.uk/~masjap/EuLisp/eulisp.html and http://www.aiai.ed.ac.uk/~jeff/lisp/eulisp.html and ftp://ftp.maths.bath.ac.uk/pub/eulisp/
* Fare/#Tunes considers mzscheme as a cross-compiler
* Fare/#Tunes is kind of impressed by the Rice people whenever he meets them or read their papers
<Fare> s/Rice/Rice PLT/
* Fufie/#tunes also likes Rice PLT.. decent chaps and MzScheme is cool
<eihrul> Fufie: ah, so eulisp is just a language spec?
<Fufie> and the plt mailing list came to life again :)
<Fufie> eihrul: basically yes
<Fare> eulisp has a sample implementation, that basically sucks and lacks some advanced features
<Fufie> it would be charming to make a european lisp
<eihrul> Fare: doesn't mean an implementation couldn't be built to support them
* Fufie/#tunes presumes the french mob would love to make an anti-american effort ;)
<eihrul> Fufie: hell, the americans would love to make an anti-american effort
<Fare> eihrul: of course it doesn't mean that
<Fare> Fufmann: anti-american?
* Fufie/#tunes would love an OS which provided decent locale support from the start instead of the loser-thing called gettext
* Fare/#Tunes would love that too; actually, my mom being pressing me for a VN database, it'd be a main spot in Tunes
<binEng> what's gettext?
02:40pm
<Fufie> gettext is an evil hack made to add rudimentary support to apps written in C
<Fare> bE: a horrible kluge
<Fare> it's not even a hack.
<Fufie> kluge is the right word, yes
<Fare> it's an ugly retro-fix for utter thoughlessness
<binEng> why? what?
<Fare> Yet, coming from non-metaprogramming imperative programmers, it's kind of a feat, already.
<eihrul> hey, they're not non-metaprogramming imperative programmers, they're just metaprogramming functional programmers who haven't started yet....
* Fufie/#tunes gives eihrul a twinkie
<eihrul> they just need to be converted...
<eihrul> maybe tunes could handle such a difficult translation? :)
<eihrul> not only would it translate programming languages, but it could possibly translate programmers...
* Fufie/#tunes decides to use PST to make the translation date as MET is about to become the 16th and the release date is the 15th ;)
<Fare> only use UTC
<Fufie> I can make it with UTC as well
<eihrul> Fufie: btw, twinkies suck
02:50pm
<Fufie> oh sorry.. only wanted to reward your optimistic view on the world :)
<dalvarez> I'll use the few remaining minutes till network power-off to clear a last thing ( before I get that nasty TRYING TPE\nTRYING AUI - shit again): how do I get a member-account on tunes.org ?
<eihrul> well, it isn't optimistic... its just if i can change, so can they :)
<eihrul> Fufie: and in retrospect, its horrible to think of myself as an ex-non-metaprogramming-imperative programmer
<eihrul> its much nicer to think of a metaprogramming functional programmer still in gestation...
<Fufie> t
<eihrul> hmm, David Ungar gets around...
<eihrul> dalvarez: ask fare or something
<dalvarez> eihrul: that's what I'm doing
<eihrul> well, then stop pinging me! :P
<eihrul> you pervert
<Fare> dalvarez: I can create one for you at once
<Fare> dalvarez: got ten minutes?
<dalvarez> eihrul: thought I'd been lagging
<eihrul> Fare: what are the perks of becoming a member anyway?
<Fare> perks?
* Fare/#Tunes dict perk
<dalvarez> Fare: 6 to be exactly
<eihrul> Fare: what are the benefits?
<binEng> eihrul: hehehe... (evil laugh)
<Fufie> eihrul: cvs access to the tunes area ;)
<eihrul> wow, i must have membership then!
<eihrul> so that i can learn from the large and extensive tunes cvs code base :)
<Fare> eihrul: perks are you can look contemptuously at all those who just "don't get it"
<binEng> coool
<dalvarez> eihrul: don't like to waste the few instructions to get the ICMP processed or what the hell?
<dalvarez> Fare: so it's useless by any means?
<smkl> did you do an ICMP or CTCP ping?
<Fare> no, you also get CVS write access to tunes.org
<Fare> so you can modify the pages, and add code
<Fare> (and get mail&shell account)
<dalvarez> that'll be unneccesary for first
<Fare> (not to be abused -- read the policy)
<eihrul> maybe i'll join, some day...
<dalvarez> mail & shell would be nice
<dalvarez> I like gimmicks
<dalvarez> 3 mins
<Fufie> policy?
<binEng> yeah what's that policy?
<binEng> never heard of that
<Fufie> eihrul: in the same cvs tree there is several other related projects of tunes.. I think I use a couple of M
<Fare> /usr/local/bin/policy -- basic no-net-abuse agreement
<dalvarez> the net will be powered on a few minutes later cause it's WE, I'll go and return later to avoid "Carrier lost" system spasms
<dalvarez> leave
<eihrul> Fufie: like?
03:00pm
-:- dalvarez [ircusr@212.68.72.98] has left #tunes []
<Fufie> bespin:~> ls /cvs
<Fufie> CVSROOT    TunesWiki  retro      tril       tunes
<Fufie> CVWiki     fare       sds        ttp        xcom
<eihrul> okay, why not...
<eihrul> Fare: join me up
<smkl> 31631   /cvs/sds
<Fufie> smkl: obviously some people write code ;)
<Fare> eihrul: where?
<eihrul> Fare: tunes
<eihrul> Fufie: what is sds?
<abi> rumour has it sds is a framework for developing programming tools such as source level documenters, source browsers, source metric collectors, and similar
<Fufie> http://sds.yi.org/
<Fufie> (I am updating it with a new release 0.5.10 any minute)
-:- ircusr [ircusr@212.68.72.98] has joined #tunes
<ircusr> back again with network up and running and practical infinite time (it's just the hubs are powered off at midnight automatically to have s.th. to disagree with and keep moral down) 
<smkl> 24 hours practically infinite?
03:10pm
<ircusr> smkl: I don't pertain to the class of people who completely lack natural requirements like sleep
<ircusr> smkl: that means I'll have to stop talking sometime
<ircusr> it was so damn busy some minutes ago - where have they all gone?
<ircusr>   
<eihrul> did you just beep?
<ircusr> yeah
<eihrul> i humbly request you stop that
<ircusr> k
<smkl> GGGGGGG
<ircusr> humbliest I ask your forgivement (what a word)
<binEng> I didn't hear anything
<ircusr> smkl: me too
<binEng> it is with my sincerest humbleness that... ;)
<ircusr> the humbleness is mine, mylady :)
<binEng> ah yes, pardon
<eihrul> had i the ops to kick yas now....
<Fufie> I thought I was the only lady here..
<binEng> umm, me?
<eihrul> </sarcasm>
* binEng/#tunes attaches a sarcasm warning sign to eihrul
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
03:20pm
* Fufie/#tunes smiles.. done.. 
* eihrul/#tunes needs the ability to read two papers at once.
<eihrul> how would one go about attaining that?
<binEng> easy. You have two eyes (I assume). Use both.
<ircusr> then you'll still need a refective brain
<eihrul> yes, but i have but one locus with which to process it...
<ircusr> or timesliced
<Fare> eihrul: I can read two papers at once (but not understand either)
<eihrul> Fare: that is a requirement of above-stated activity
<eihrul> being able to read both papers concurrently and in so much detail as one could be possibly read
<Fare> eihrul: do time-sharing through preemption
<eihrul> will yield a process slower than just reading each paper individually!
<eihrul> at most, i'd like it to take 125% as long as a single paper
03:30pm
* eihrul/#tunes needs true parallelism.
<Fare> no, just set the time-slice for preemption to a few hours
<Fare> or maybe just 45'
<Fare> (plus 5' breaks for administrative tasks)
<eihrul> if only the brain were truly reflective
<eihrul> i'd like to change much of its implementation
03:40pm
-:- SignOff ircusr: #TUNES (sweet dreams)
<Fare> eihrul: but after you did, it wouldn't be _you_ who would be there to appreciate (or regret) the result
<eihrul> unfortunately, no
<eihrul> it all depends on the question of what we consider the self and whether it can persist after being shut off....
03:50pm
<Fare> eihrul: if you can read french, try my meta-novel
<eihrul> nope
<binEng> what's a meta-novel?
<eihrul> i'm limited to various programming languages, english, and spanish
<Fare> bE: a meta-novel is a text about a novel
* eihrul/#tunes thinks Fare enjoys using the word 'meta.'
<eihrul> although it is quite a fun word to use
<binEng> Fare: I can imagine, but how do you... implement it?
* Fare/#Tunes enjoys using the _concept_ meta
<eihrul> oops, yes
* binEng/#tunes enjoys the concept of concepts
<Fare> bE: you mean, implement the meta-novel, or implement the novel?
<eihrul> binEng: you mean a meta-concept
<binEng> eihrul: Wouldn't that be _the_ meta-concept in such case?
<eihrul> yes
<eihrul> appropriate article substituted, of course
<binEng> eihrul: So you know Spanish? Could you say something I could surprise my sister with? She's learning Spanish.
<eihrul> rmm, i know spanish
<eihrul> but i don't happen to be the most spontaneous of people... i need a prompt
<binEng> eihrul: C:\
<Fufie> $
<eihrul> i prefer # myself
<eihrul> but alas, use it very infrequently
<binEng> a shame
* Fufie/#tunes prefers not to use #
<eihrul> how often does one actually need to use #?
<eihrul> most of the time you only need $
<binEng> eihrul: Tell her there's no future in learning Spanish.. that Russian is in fashion instead. ok?
<Fare> bE: vosotros sonos unos payasos
<binEng> eh?
<Fare> bE: żestudias o trabajas?
<eihrul> binEng: in southwestern united states, there's a big future in it, unfortunately
<Fare> eihrul: why "unfortunately"?
<binEng> I don't follow you Fare. ;)
<eihrul> Fare: unfortunately for his statement
<eihrul> though, it'd be nice if instead of moving to another flawed language we could move to a less flawed one
<Fare> Let's return America to the very americans, and boot those foreign-language speaking invaders off our land!
<binEng> eihrul: I don't pretend I'm American
<Fare> America to indians!
<binEng> eihrul: neither does my sister
<binEng> Fare: I second that
04:00pm
<eihrul> Fare: sure, but then you have to return it to the other animals that the very americans took it from
-:- Downix [down@d-gnaps-70.ici.net] has joined #tunes
<binEng> Back to the subject! Now, what do I tell her?
<eihrul> and then they to something else
<eihrul> and regression to who knows where...
<Downix> Ok, who the hell decided how to set up Windows 95?
<binEng> Surely what Fare said can't be it?
<Fare> hell, the jews had people pay for their genocide. Let's claim that they pay 1 symbolic dollar back to the genocided people of Canaan for their genocide (with 5% composite interest per year over 4000 years)
<binEng> eihrul: well? prove your linguistic knowledge
<eihrul> binEng: tell her that there is some movement somewhere to execute with prejudice any spanish speaking people
<binEng> eihrul: Whatever, but in Spanish plz, that's the whole point.
<binEng> eihrul: or.. say that one learns Spanish much better on the Internet and that I managed to learn the whole thing in half a week. ;)
<eihrul> hay no futuro en espanol, entende la idioma rusa
<eihrul> (not quite sure how to say russian language, but close enough :>)
<binEng> eihrul: ah, ok add the latter I said, but a BTW in front
<eihrul> not sure of the spanish equivalent, but literal translation could be "por la manera" :)
<Fare> hay un movimiento en alguna parte para ejecutar con prejuicio todos los que hablan espanol
<binEng> Fare: What did that mean?
<eihrul> si, los espanoles necesitan mueren!
<Fare> abi: spanish is a language with dictionaries at http://www.anaya.es/diccionario/diccionar.htm
<eihrul> binEng: there is a movement en some place to execute ith prejudice all that speak spanish
<binEng> great
<Fare> eih: s/necesitan mueren/se deben morir/
<eihrul> well
<eihrul> er... yes, brain fart
<eihrul> should have been necesitan morir
<eihrul> but as to whether you use need or should is a matter of perspective :P
<Fare> todo el mundo necesita morir, pero cualgun al momiento propicio
<eihrul> sure, you can look at it that way
<binEng> ok, it's good now... you can stop now
04:10pm
<eihrul> (progn (yawn self) (assert (make-language :medium 'oral) 'boring))
<Fare> 3053 time
<eihrul> this early?
<eihrul> its probably nearly night in france; you must wait till the morning!
* binEng/#tunes notices that it actually *is* night where he is
* Fufie/#tunes goes to bed.. good night
<binEng> night
<eihrul> oh, well then its definitely too early
<eihrul> s/its/it's
<Downix> ok, who's awake?
<binEng> I am! (I think)
<Downix> alright
04:20pm
* Downix/#tunes is pondering
<binEng> (but not for much longer)
<eihrul> Downix: what, that the amiga needs to live again and that its the best architecture you've ever seen save your own?
<Downix> eihrul:  How'd you guess?
<eihrul> deja vu?
<Downix> pretty much
* Downix/#tunes is pondering about how much work it would be to get the OS to be ported
<Downix> This is one of the most bizzare options....
<Downix> I feel like I'm being tempted.  Either stick to my plan as/is and stick to Eddas... or licence the AAA and Hombre from Amiga Inc.
<eihrul> umm, you said you've already put in one man year into eddas
<eihrul> so wouldn't it be throwing money down the drain?
<Downix> eihrul:  I could continue Eddas still, but the Amiga licence could get us a profit in the short term.
<Downix> eihrul;  But by doing that we could end up looking like we're just an Amiga company
<eihrul> do both then
04:30pm
<Downix> there's the image problem
-:- air [brand@p0wer.qzx.com] has joined #tunes
<Downix> Plus then we'd have to deal with keeping their old OS running while we work on Eddas, running core's Clementine, lar's OS or air's Brix
<eihrul> it is more likely you'd be running clementine than the latter two
<Downix> very likely
<eihrul> since core has a lot more work into his and actually has a company backing it
<Downix> But then we're going to have to have an OS guy keep AmgiaOS up to date (a major pain in itself)
<Downix> We'd have 3 solutions, a lot of cross-competition
<Downix> AAA, Hombre and Eddas
* eihrul/#tunes shrugs.
<eihrul> well, do they atleast all fit a particular different need?
<Downix> No
<Downix> Hombre is AAA's replacement
<Downix> and Eddas is a totally different beast made to fit into their market
<Downix> But it would buy me time
<eihrul> well, you could cater to both customers :)
<Downix> Dr Hepler said that using a MIPS CPU in Hombre was quite possible, so we could set up Eddas as a replacement for Hombre'
<Downix> very true
04:40pm
-:- Ghyll [karltk@mp-217-240-61.daxnet.no] has joined #tunes
<binEng> Bye
-:- SignOff Downix: #TUNES (Ping timeout for Downix[d-gnaps-70.ici.net])
<Ghyll> bye
<Ghyll> (whomever's leaving)
<binEng> me ;)
-:- SignOff binEng: #TUNES ( <k!14>)
-:- Downix [down@d-gnaps-191.ici.net] has joined #tunes
-:- SignOff Downix: #TUNES (Ping timeout for Downix[d-gnaps-191.ici.net])
<eihrul> Ghyll: hrmm, new to tunes?
04:50pm
<Ghyll> eihrul: somewhat. new to #tunes. Not read the compulsory tunes.org/ material either. Hmm. So.. 'Yes' :)
<eihrul> any tunes related project you're working on?
<Ghyll> eihrul: more or less.. http://sds.yi.org
<eihrul> ah, with fufie?
<Ghyll> eihrul: I'm told by James Little it's a little like Prism. 
<Ghyll> eihrul: but I've not read up on Prism yet.
<Ghyll> eihrul: Yup. 
<Ghyll> eihrul: I'm Fufie's henchman.
05:00pm
<eihrul> prism was supposed to be some weird source code translation system, i think atleast
<eihrul> or rather is
<Ghyll> My impression was that it was supposed to be a framework for code-tools; kind of like the infrastructure needed to support a compiler.
<Ghyll> I think the idea was to have compilers, documentation generators, debuggers, editor and all the rest of the menagerie talk the same language.
<eihrul> well, i've probably looked at it even less than you :)
<Ghyll> I for one would be very relieved if I didn't have to write/modify/hack front-ends for every damn languange we want to support. 
<eihrul> let's ask abi
<eihrul> abi: prism? what is it?
<abi> eihrul: wish i knew
<eihrul> abi: prism?
<abi> prism is a way to program with multiple specialized, interoperable languages. or at http://www.teleport.com/~sphere or YABL
<eihrul> Ghyll: just teach the world the mother-tongue, and no problems
<eihrul> i should say mother-tongue(s)
<Ghyll> I wish one would suffice.
<eihrul> well, the world could do without Java and C++ atleast
<eihrul> that'd be a start
<Ghyll> fortran, too.
<eihrul> then you could move on to eliminating cobol, basic, ^ fortran
<eihrul> visual basic of course... many more
<eihrul> pascal can go as well
<eihrul> any language less expressive than common lisp :)
<Ghyll> unless you want to throw away all the current hardware, you'd still have to leave one language to write the device drivers.
<eihrul> hey, i didn't say assembly
<eihrul> nor did i say C
<Ghyll> I noticed. 
<eihrul> C has limited usefulness
<eihrul> so long as you stay within its *domain*, no problem
<Ghyll> C is a nice evolution of the assembler.
<eihrul> sure
<eihrul> but things like C++ which try to make it more than it can be :)
<Ghyll> It's just marginally easier to write than macro assembler.
<eihrul> hell...
<eihrul> the macro assemblers have better macro systems than C!
<eihrul> nasm for example
<Ghyll> I find it ironic that the US defence made Ada when c++ was coming ("Be all you can be")
<eihrul> hmm, if you added meta-programming to an assembler
<eihrul> you could probably have something very close to a compiler :)
<eihrul> well, cleaner than self modifying assembly :)
<Ghyll> you have a very crude form of meta-programming with the macros.. 
<Ghyll> only compile-time, tho.
<eihrul> well, compile time meta-programming would be enough to make it a fairly usable compiler for certain languages
<Ghyll> but if you want to write self-modifying programs, you probably shouldn't do it using assembly.
05:10pm
<Ghyll> why not just use ML, then ?
<eihrul> Ghyll: well, i mean that it would be the compiler, man :)
<eihrul> atleast during bootstrapping stage
<eihrul> (before you have your compiler in your target language)
<Ghyll> you're really into the "be all you can be"-thing, aren't you ? :)
<eihrul> not really
<eihrul> i'm very much lazy
<Ghyll> and you're per chance a programmer ?
<eihrul> yes
<eihrul> programmers => lazy 
<Ghyll> exactly.
<Ghyll> (I shouldn't have bothered asking..)
<eihrul> we make the computer do most everything for us :)
<eihrul> the ultimate laziness!
<eihrul> hell, sometimes we even make our computations lazy! (haskell)
<Ghyll> or spend more time having them doing whatever it is need to do automatically, than it would require to do manually..
<Ghyll> lazy is good. 
05:20pm
<Ghyll> gnight
-:- SignOff Ghyll: #TUNES (bar)
-:- mibin [mibin@62.11.104.255] has joined #tunes
05:40pm
-:- SignOff mibin: #TUNES (Read error to mibin[62.11.104.255]: Connection reset by peer)
-:- mibin [mibin@62.11.104.255] has joined #tunes
-:- Random^^ [random@208.159.80.180] has joined #tunes
-:- Random^^ [random@208.159.80.180] has left #tunes []
-:- mibin is now known as mibout
-:- Nanny [user9545@ndavinci.cc.uottawa.ca] has joined #tunes
-:- Nanny [user9545@ndavinci.cc.uottawa.ca] has left #tunes []
-:- Nanny1 [user5383@ndavinci.cc.uottawa.ca] has joined #tunes
-:- Nanny1 [user5383@ndavinci.cc.uottawa.ca] has left #tunes []
-:- AlonzoTG [Alonzo@216-164-130-162.s543.tnt2.lnhva.md.dialup.rcn.com] has joined #tunes
-:- NetSplit: lackey.openprojects.net split from sterling.openprojects.net [07:53pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [lackey.openprojects.net]
-:- Netjoined: lackey.openprojects.net sterling.openprojects.net
-:- Fare [fare@quatramaran.ens.fr] has joined #tunes
<eihrul> abi: eulisp?
<abi> somebody said eulisp was http://www.bath.ac.uk/~masjap/EuLisp/eulisp.html and http://www.aiai.ed.ac.uk/~jeff/lisp/eulisp.html and ftp://ftp.maths.bath.ac.uk/pub/eulisp/
08:30pm
<eihrul> Fufie: are you there?
-:- hcf [nef@me-portland-us829.javanet.com] has joined #tunes
<eihrul> lo
<hcf> hi
08:40pm
-:- SignOff eihrul: #TUNES (Read error to eihrul[usr5-ppp24.lvdi.net]: EOF from client)
-:- eihrul [lee@usr5-ppp24.lvdi.net] has joined #tunes
<eihrul> abi: eulisp
<abi> eulisp is http://www.bath.ac.uk/~masjap/EuLisp/eulisp.html and http://www.aiai.ed.ac.uk/~jeff/lisp/eulisp.html and ftp://ftp.maths.bath.ac.uk/pub/eulisp/
08:50pm
-:- GAlexand [galexand@co-22-36.collins.indiana.edu] has joined #tunes
-:- water [water@tnt-9-124.tscnet.net] has joined #tunes
* AlonzoTG/#tunes greets "Salutations water!" :-)
<water> hey
09:50pm
<hcf> sup?
<water> got a 3rd os
<AlonzoTG> I just drew a graph of system complexity versus ease of use 
<AlonzoTG> which?
<water> beos4.5
<AlonzoTG> groovy.
<water> 8)
<AlonzoTG> I heard they have an excelent kernel.
<water> yeah, it supports all my hw and mounts all my fs's
<AlonzoTG> but I still don't like how complex the underlying system is.
<AlonzoTG> I mean all those files... yuck!!!
<water> well, the api seems very programmer friendly
<AlonzoTG> It has a lot going for it.
<AlonzoTG> It has dumped all the unix legacy and made a really clean system.
<water> geez. *another* squeakCE optimization update
<water> if they make it any faster, there'll be no reason to drop morphic from the image
<water> btw, who's g. alexander?
<water> ok n/m
10:00pm
<eihrul> there can never be too many optimizations :)
<eihrul> (so long as they're hidden, of course)
<water> bah. if the vm grows beyond 500k, that's too big even for me
<water> even *i* don't like squeak that much
<eihrul> selfCE? :)
<water> lol
<water> yeah, that would not work without some sort of overhaul of the implementation
<water> otoh, r5rscheme runs fine on ce (without full graphics support, of course)
<eihrul> welp, finally got around to starting on self implementation papers :)
<water> ah
<water> the progref has a lot of the basic useful info on implementation, btw
<eihrul> looked at eulisp
<water> k
<GAlexand> i'm a random person with an interest in tunes but not enough knowledge to justify opening my mouth, so I intend to lurk if no one minds.
<water> heh
-:- rares [rares@wtrb-sh3-port6.snet.net] has joined #tunes
<water> ok you're welcome to lurk, but please don't do it 24/7, we have too many of those
<rares> hey w
<GAlexand> heh.
<water> hi rares
<eihrul> water: i don't know a person in here who doesn't say something atleast once...
<eihrul> however infrequently
<eihrul> even abi says stuff
<rares> abcdefg...
<water> i've seen people lurk for two weeks without ever saying a thing
<water> i even grep'd the logs to make sure
<rares> abi; abi
<rares> abi: abi
<abi> i am the little girl borg with a big brain or #tunes' infoslut
10:10pm
<hcf> water: any slate news?
<water> hm not just yet
<water> btw, "magnolia" is a pretty cool movie
<rares> thought about the message is anobject thing
<rares> there's a way to hack it
<eihrul> water: btw... there's two rewrite rules on the same line in your grammar
<water> hm?
<abi> well, hm is that fundamental then?
<water> lol
<water> abi forget hm
<abi> water: I forgot hm
<eihrul> letter -> small­letter | cap­letter identifier -> (small­letter | `_') {letter | digit | `_'}
-:- NetSplit: carter.openprojects.net split from irc.linux.com [10:14pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [carter.openprojects.net]
<water> oh yeah typo
<eihrul> unless that's some new agreed upon feature of pseudo-bnf :)
<water> heh no
<eihrul> hrmm, classes as a runtime optimization, nice :)
<water> yep
10:20pm
<water> i'm also going to call meta-objects meta-spaces to keep the analogy fairly intuitive and physical
<water> but that's mobius, not slat
<water> slate
<eihrul> ah, so this would easily eliminate the hash table costs
<eihrul> if nothing else
<water> sure
<eihrul> (wrt size constraints anyway)
<eihrul> and also would allow you to store lots of other gc/optimization information
<eihrul> wow, next page is on that :)
<water> yep
10:30pm
<eihrul> though in self, are such meta-objects exposed?
<eihrul> or can only the vm see them?
-:- lispbliss [lispbliss@cont01p44.ont.micron.net] has joined #tunes
<water> hm
<water> i think only in the vm
<water> since i don't have self on my computer
<rares> see you can;t get rid of the lower level design it keeps seepiung deeper but never goes away
<water> rares: huh?
<rares> the meta-objects
<water> what the hell are you saying?
<rares> i'm tired, but what i'm saying is there's always a static part of the design
<rares> some sort of base upon which the magic happens
<water> not always
<water> moostrap had some interesting ways to dynamically compile away meta-objects
<eihrul> rares: well, afaik, these meta-objects are silent
<eihrul> water: ...
<rares> i'm tired like i said
<rares> this is the time when my ghosts arrive
<rares> and I start rambling month old discussions
* water/#tunes opens another GhostView window just to look at rares' ghosts
<eihrul> though, the object segregation thing seems almost controversial :)
<water> well, you can migrate objects between meta-spaces
<rares> hey
<water> that's the whole idea
<rares> no naked ghost pics
<water> hehe
<eihrul> water: well, i mean growing byte arrays downward and objects upward kinda biases it towards some gc algorithms that don't need to grow in both directions :)
<water> oh that
<eihrul> er that do, i mean
<water> yeah, it's got some problems
<rares> objects and meta-spaces data-pages data-spaces, oops sorry
* water/#tunes ignores rares for a bit
<rares> heh
<rares> well i'm gone anyhow
<rares> cyall
<water> i have to agree there :)
<water> cya
<rares> i have to wake up bright and early for protocol day
<water> "protocol day"?
10:40pm
<rares> yes Unix networking self-taught crasgh course
<water> ah
<rares> NNTP to FTP to HTTP to ICMP
<rares> i'm a bunna have some fun
<rares> gunna even
<eihrul> learn about mops, not unps :)
<eihrul> they're protocols
<rares> um what?
<eihrul> so you won't have to rename the day, either
<rares> no Sun sux bad attitude
<rares> open charlatans
<rares> okay my speech has become barely bisyllabic must sleep
<rares> *poof
-:- SignOff rares: #TUNES ([x]chat)
<water> darn, too many programing systems on my fat32 partition
<water> hmm goodbye ms office :)
<eihrul> make slate such that you only need one :)
<eihrul> (and so that i only need one for that matter :>)
<water> yeah well the ones i have are beta, clisp, and squeak, so that's definitely true
<water> unfortunately, c refuses to go away just like that
<water> afk
10:50pm
<eihrul> we can overcome c! :)
<eihrul> at the worst, there's tal! :)
<water> there's "tal"?
<water> what's tal?
<abi> well, tal is typed assembly language at http://www.cs.cornell.edu/talc/default.html
<water> yuck
<eihrul> </sarcasm>
<eihrul> but as a backend, assembly isn't too shabby :)
* water/#tunes slaps eihrul around with The Art of Computer Programming
<eihrul> mix was rather nifty :)
<eihrul> much more elegant than the x86, for instance
<eihrul> in its 5 bit byte madness....
<eihrul> though, the advantage is you can count a single byte on one hand!
<water> afk again
11:00pm
<eihrul> very tricky with the tagged pointers as well
<eihrul> but tricks are fun
<water> i'd like to get a good implementation with some more versatile tricks, myself
<water> darn it, i completely forgot to look over the links hcf got me today
11:10pm
<water> btw, one of the goals of slate is to dynamically be able to spawn threads and handle concurrency and perhaps even asynchronous messaging
<eihrul> yep, only thing tricks are good for :)
<eihrul> yeah, concurrency is something largely lacking in current generation languages
<water> i think the "spawn object as thread" idea will have to be a primitive
<eihrul> unless you can find some other notion of concurrency
<eihrul> that formalizes that operation better
<eihrul> 'fork' is adequate enough for some cases
<water> well, beta's threads are cooperative. they can't be pre-empted
<eihrul> but is quite crude in others
<eihrul> agg, no preemption?
<water> nope
<water> plenty of synchronization options though
<eihrul> well, what about true parallelism?
<water> and beta's threads act very well as co-routines
<water> i believe beta's threads are os threads, so any parallelism allowed by the os is inherited by the program
<water> the manual says so, anyway
<water> it would also be nice to have a program divide into protected-memory spaces
<water> especially if the environment should support experimental programming
<eihrul> not hardware protected i hope? :)
<water> heh. probably not
11:20pm
<eihrul> if you had a truly-over-the-metal safe language
<eihrul> you'd probably get some speed benefits just by the fact you wouldn't have to abuse the mmu as much
<water> the real trick would be to keep all these "extra" semantics open for modification
<water> and reduce the core language to a code-generator
<eihrul> well, what's wrong with just a slate2some-machine-language translator? :)
<water> heh
<water> well, i need to focus on the language semantics integration part first
<water> (which is what i'm reading up on now)
* eihrul/#tunes is reading up on implementation.
<water> good for both of us
<eihrul> allows for better parallelism in efforts
<eihrul> though not quite:
<water> how so?
<eihrul> (fork make-spec) (fork read-on-implementation) (wait make-spec) (fork make-implementation) (wait make-implementation)
<water> heh
<eihrul> everything's pretty up to (wait make-spec)
<water> don't worry about the spec, just keep toying with the design
* eihrul/#tunes nods.
<water> you'll learn enough from toying
<water> well, within clisp, anyway
* eihrul/#tunes uses cmucl.
<eihrul> but point is moot i suppose :)
-:- SignOff GAlexand: #TUNES (sleep or something)
<water> btw, i'm considering re-introducing the class concept as a declarative spec for objects which can be dynamically 'inherited'
<water> sort of an abstract type system, but mostly at the library level, not the language level
11:30pm
<eihrul> to what end?
<water> mostly to help characterize behaviors without having to explicitly compare object structure
<water> a lot of languages do the same thing in a less organized way
<eihrul> static typing revisited?
<water> hm i hope not
<water> more like making an exception-handling system
<eihrul> well, if you can declare the characteristics of an object
<water> no, i'm declaring the requirements for class-membership
<water> there's a diff
<eihrul> sure... but what use is this if you must examine the object structure to find out class membership if your point was to avoid that? :)
<water> well it could be done systematically
* eihrul/#tunes is still wondering how this would be used.
<water> don't worry, this isn't an issue until later
<water> although i'll pull up the paper that gave me the idea for reference
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us829.javanet.com])
<water> ok here's an example
11:40pm
<water> you could test the state of an object and have it inherit a method for handling that state from a class
<water> like an overflow in a buffer
-:- hcf [nef@me-portland-us108.javanet.com] has joined #tunes
<eihrul> so classifying an object by its state...
<eihrul> as opposed to classifying it by the structure of its state
<water> yeah, or potentially its meta-state
<water> er... implementation state
<water> i hate using "meta" too often
<eihrul> now, why couldn't you just give class A a method for handling overflow within class A?
<water> well, this apparently allows you to factor the code
<eihrul> perhaps a better example would be a signed integer?
<water> ok
<eihrul> which switched between the class of positive integers or negative integers depending on the state of its sign bit
<eihrul> s/switched/switches
<water> hmm
<water> yes that works
<eihrul> though even in that case
<water> maybe a different example would be to have a class of binary operators
<eihrul> its just as simple to define methods to handle both negative and positive cases
<water> right
<eihrul> though... if you want to specialize a given operation
<eihrul> based on the *value* of a given argument
<eihrul> as opposed to the *type* of a given argument
<water> well, "simple" isn't the same as "refactored"
<water> e.g. smalltalk code may be simple, but not simple to refactor
<eihrul> well, why not just send me paper?
<eihrul> i'll shove it in my deque....
<water> it's the cecil-spec.ps
<eihrul> where at?
<water> pg. 18-21
<eihrul> abi: cecil?
<abi> rumour has it cecil is object oriented programming language at http://www.cs.washington.edu/research/projects/cecil/www/cecil-home.html
<eihrul> paper #33 :)
<eihrul> erg...
<eihrul> make that paper #46
<eihrul> more optimization papers :)
11:50pm
<water> heh
<eihrul> though my post script collection is growing nicely
<eihrul> i have atleast 200 papers now
<water> heh
<eihrul> how many in yours?
<water> i lost count a while ago... i still need to get my laptop hd back online
<eihrul> find | wc --lines 
<water> it's distributed, how's that :)
<water> hm... 250 on this machine
<water> but my laptop has 1.6gb of papers, i know
<eihrul> only 62 MB myself
<water> that's more than what my 250 papers takes up
-:- lispbliss [lispbliss@cont01p44.ont.micron.net] has left #tunes []
<eihrul> makes me wonder if you could use multimethods and some macro hackery
<water> i still don't quite grok what multimethods are for
<eihrul> to do much of the specialized compilation techniques of self hackishly above common lisp :)
<eihrul> specializing on more than the first arguments to a method
[msg(TUNES)] newlog 2000.0116
IRC log ended Sun Jan 16 00:00:01 2000