IRC log started Sun Mar 26 00:00:01 2000 yeah you have plenty of power to run squeak it runs well on my pda [msg(TUNES)] permlog 2000.0326 although it'll scream on my handheld once the jit engine is done 8) ok... listen, I gotta run, but I'll be around... hehehe later guys ok -:- SignOff TFox: #TUNES (Leaving) wow it's amazing how clueless computer techs can be about programming issues it's disgusting to think that he gets paid anything decent is everyone else here lurking? 12:10am water yeah I'm sort of here. I'm just flitting about my PC coding, reading, mudding, &c whatever Bah. I'm not trolling at least. That has to be worth something? * water/#tunes gives ult a cookie for not being a troll enjoy * ult/#Tunes smiles and puts the cookie in his pocket 12:30am -:- SignOff ult: #TUNES (Ping timeout for ult[user-37kba0k.dialup.mindspring.com]) -:- fire` [no@209-68-229-46.dialup.cust.tfb.com] has joined #tunes whoa it's hickserv hey i saw your interview who prompted that? the interviewer he just showed up on the channel i see that's cool so how's it been going? there's lots of issues to work through tril finally is getting involved again yeah i noticed his increased activity on the mailing list but slate keeps getting more and more concrete increased = one post? that's more than he has posted in a while sure slate seems interesting thanks it's nice that you borrowed from self it's my favorite language :) yeah a lot of people like it i forget are you the one who actually has a sparc? heh no although i can do a little sparc assembly hm that must have been eihrul or someone or at least could well have you been able to run self on anything (like a mac)? -^lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all. We're in the process of dealing with wingate exploits; if you have code to handle these situations, please drop by #openprojects. Thanks. no i just admire it from afar bah i've been playing with squeak lately okay that's good how are you with coding in it? not that great i need to learn more smalltalk have you tried using the browsers? yes btw, there are some new morphic tutorials (very good imo) on the squeak swiki alrighty i think i have the url somewhere would you like it? i read two of them oh ok no i know where they are 01:30am i knew i shouldn't have taken that 5 hour nap this afternoon now i can't sleep heh good you can keep me company not that i'm that interesting btw we can talk in #dolphin for noisy stuff -lilo(lilo@varley.openprojects.net)- [GlobalNotice] If you have code that's designed to handle wingate exploits, we need to look at it. Please email code or pointer to support@openprojects.net. Thanks. oh yeah lar1 owns the old sparc i tried to research beta but i still don't understand what patterns are it's not too hard you just take a data structure and give it input and output declarations and then you allow all the blocks to act like this since you have data+procedures = objects, you can have them inherit and all that i see im glad you've somewhat simplified some information on arrows np i need to html-ize arrow information when i get the time how will slate relate to arrow? hehe i'm going to roll arrow over slate arrow's going to start out as a novel kind of type system there are still some issues i haven't resolved, but slate makes implementing arrow much more natural then i could have done any other way cool well, of course i got involved with tunes for this very reason has Fare gotten anything done on the hll? not really heh doesn't suprise me unless you count rscheme as the hll-prototype, which isn't even his work scheme is nice but it isn't tunes worthy you think so? why? it's too procedural hm how is slate not like that, though? (i already have ideas on this, i just want to hear your ideas) well the syntax is lisp like but isn't it more object oriented like self? yes that's true scheme objects aren't very expressive there are no ways to access members 01:40am shhhh... Fare may hear you :) actually there are :P lisp-based object systems like CLOS (have you used it?) are pretty powerful oh no i haven't yes but my objections to them are based on the lisp foundations instead of foundations which help more with other styles of managing dynamic systems i like the idea of unifing concepts...that is one of my agendas yeah that's why i chose the slate idea over clos btw the slate idea i came up with in college about 4 years ago immediately after i looked at the tunes, beta, self, and clos systems i'm just puting some ideas together and gaining a little programming experience yeah you still need experience helping this guy put together a half-life mod and sell it to the world opponent network have you written any big programs yet? oh ^ | :) how big is it? that way i can see some code written by professionals and write my own yeah well...i don't know an accurate way say but i also signed a non disclosure agreement so... i don't think it violates an nda to talk about code size that you wrote i haven't written any code thusfar not even scripts? there really isn't much to write bah you should work on some code project we're still waiting to get some stuff back from the coder code it yourself! the other coder rather uh i don't think you really understand the circumstances water what circumstances are we talking about? 01:50am i just joined up and before i can get anything done we need to get something back from a guy who has sorta lost some steam some sort of code sounds like a crappy project to join well experience is experience yeah but you don't get to code yes i do what do you do exactly? heh beats me so far oic you can't code until you receive code? i haven't really had the chance to talk to the guys lately...with school and track hm * fire`/#tunes yawn i think i will hit the sack now ok cya later -:- SignOff fire`: #TUNES (Leaving) 02:00am abi henry baker is at ftp://ftp.netcom.com/pub/hb/hbaker/home.html -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us728.javanet.com]) 02:20am -:- Fufie [stig@tunnel-44-31.vpn.uib.no] has joined #tunes -:- ult [ult@user-38lcn52.dialup.mindspring.com] has joined #Tunes -:- dalvarez [ircusr@212.68.72.98] has joined #tunes abi: seen nate37 nate37 was last seen on IRC 1 days, 11 hours, 58 minutes and 54 seconds ago, saying: nice talking to u.. bye [Fri Mar 24 15:56:36 2000] water: i only own an sgi, no sparc 04:00am oh (dozing) doh 05:00am -:- SignOff ult: #TUNES (Leaving) -:- ult [ult@user-38lc67u.dialup.mindspring.com] has joined #tunes -:- SignOff smoke: #TUNES (Excess Flood) -:- smoke [smoke@16dyn164.delft.casema.net] has joined #tunes -:- SignOff smoke: #TUNES (Excess Flood) -:- smoke [smoke@16dyn164.delft.casema.net] has joined #tunes -:- SignOff smoke: #TUNES (Excess Flood) -:- smoke [smoke@16dyn164.delft.casema.net] has joined #tunes -:- SignOff smoke: #TUNES (Excess Flood) -:- smoke [smoke@16dyn164.delft.casema.net] has joined #tunes -:- SignOff water: #TUNES (Ping timeout for water[tnt-9-95.tscnet.net]) -:- SignOff ult: #TUNES (Read error to ult[user-38lc67u.dialup.mindspring.com]: Connection reset by peer) -:- SignOff Fufie: #TUNES (akk) abi: Joy? Joy is at http://www.latrobe.edu.au/www/philosophy/phimvt/j00syn.html or a forth like functional programming language 09:10am -:- Ghyll [karltk@193.217.201.254] has joined #tunes -:- water [water@tnt-9-96.tscnet.net] has joined #tunes re hey dalvarez thx * water/#tunes is considering buying dalvarez's sparc just to run selfon self on, even how much? i think after 6 years i should finally try it out :) dunno i got his msg after i fell asleep because there was a sun i found for $565... but it seems to have been sold, doh 09:40am * water/#tunes queues up the list of papers he wants to read abi: ultratechnology? i haven't a clue, dalvarez 10:00am on ultratechnology there are some IMHO very interesting pages with informations about a hardcore MISC processor resembling a Forth-like model, which now has been prototyped and is currently debugged. take a look if interested yeah i've looked at it before the most amazing thing with it is that it has 7k transistors only 10:10am -:- SignOff Ghyll: #TUNES (Ping timeout for Ghyll[193.217.201.254]) * water/#tunes wonders if/when Tril will complete the re-implementation of the Tunes glossary/Reivew by the time Tunes is released... maybe just a slate release would get him to finish it well, then it would disprove my assertion we couldn't have that, could we? :) -:- nate37 [nate37@ppp-206-117-27-144.dialup.pcmagic.net] has joined #tunes hello hey nate just got back from my trip not more than 5 minutes ago... *tired* :) and yet you come to irc :) my parents are out eating breakfast, they'll be back soon 10:40am anyways I have to go take a shower, be back on later. tonight k -:- nate37 [nate37@ppp-206-117-27-144.dialup.pcmagic.net] has left #tunes [] hm i just got a postscript file whose only page number was "1 of 0" 10:50am that's a collector's item... heh 11:00am hm an object's immutability status determines how well it can be distributed without overhead at the extreme end, the integers *never* get changed, and they are distributed implicitly with the evaluator at the opposite end, we have interaction-heavy processes that must keep their interactions on the same node of a distribution network to avoid unnecessary overhead hmmm, good point but... what about the meta-object? even if an object is immutable, is its meta-object? hm meta-space migration? if you distribute the object, do you have to distribute the meta-objects involved as well? yes i suppose immutability includes the meta-tower consider that Top *has* to be truly immutable if one is using slate (i'm not counting mobius-products here) actually even with mobius it might work the same way, but that's an issue for later doesn't at first seem so... but yes, i suppose it would have to be well Top won't be every object's mo even in the core language 11:20am most of the issue with Top's protocol design is simply about making sure its simple to get the entire core language's meta-behavior in a few steps of reasoning s/simply// * eihrul/#tunes hmms. on distribution of slate objects alone you could write a PhD thesis :) shhhhh.... Fare might be listening :) what does PhD actually stand for ? ah.. philosophiae doctor, right 11:30am ok here's a thought... compiling away the meta-level could be unified with an open compiler architecture although a compiler architecture that directly supports slate-style mo's seems odd well, how so? which statement? "seems odd" well, i just can't think of a compiler that acts that way explicitly compiles away slate style mo's? sure, or more precisely expresses part of its structure as mo's hm maybe compilers for apertos oh :) another fun part is going to be generalizing traditional compiler optimizations to apply to a system that is very, very dynamic with very few of the concepts used in those optimizations actually known a-priori :) yeah well that's going to take time and getting ideas from self stuff like algebraic reassociation, etc yes even branch analysis, etc... (if you have a conditional branch, you can't analyze the path unless you know the result of the condition) well, there's nothing stopping someone from using a very static style in slate * eihrul/#tunes nods. just need to track dependencies when that's the case, it might be that the stuff we add to the self compiler could reduce to the normal optimizing c stuff tracking dependencies won't be too hard in slate :) well i think that they won't well, if you read some of the self optimization papers... any time they inline'd something in machine code they had to set a dependency so they could invalidate the cache'd machine code under such time as the inline'd object changed hm yes 11:40am that idea can be modelled by slate objects using cloning and mo's of course, this requires a lot of object overhead to be dropped after the compile (down to one bit, possibly :) possibly 11:50am hm slate allows for more run-time optimizations by not requiring a specific representations for its objects it seems * eihrul/#tunes senses a degree proposal for him in the works. :) as opposed to self go for it well, can one do that freshman year? :) you can try where are you going, btw? Carnegie Mellon cool don't let them tempt you into becoming a lisp bigot ;) still have yet to get into the CS department... i of all people was put on the waiting list :) well tell your academic advisor about your ideas tell them "i'm not just another coder waiting to make more windows programs!" :) -:- SignOff dalvarez: #TUNES (Ping timeout for dalvarez[212.68.72.98]) seriously do whatever you can to get what you want going to have to... bleh every time i read that interview i think about changes i'd like to make to it 12:00pm s/read/consider/ it'd be great if i got interviewed about an arrow distro someday :) perhaps hm i think this is the 30th paper or so i've read today of course i mostly scan over it and focus on relevant stuff 30th? you just woke up... that was 2.5 hours ago yep almost 3 now (according to the logs) it's a good thing that i'm sorting these papers at 10 papers per hour, 25 average pages per paper, that is in excess of 250 pages per hour :) most aren't 25 pages... more like 15 it's the 80/20 rule all over again 12:10pm abi EqLog is an equational logic programming language at http://www-cse.ucsd.edu/users/goguen/sys/eqlog.html anything interesting? it's much like maude but with much less code and development where were brain dumps of abi stored again? http://www.tunes.org/~nef/abi/db/ ah hm i wonder if Tunes is supposed to be a modernist approach to computing like the ststus quo? i've never thought of it explicitly like that, but i realize now that i hate the modernist approach, and i think Fare likes it very much although i'm no post-modernist * eihrul/#tunes shrugs. you grok modernism, right? well, i'm no historian :) bah have only heard it used wrt art/literature in which case it was all about signals... whereas post-modernism was all about noise 12:20pm not just that enlighten me then of how it applies to Tunes bleh i can't cut-and-paste text from postscript ok i transcribe "belief in adeqquacy of the hierarchy, formalization and control to achieve desired ends without error." yeah that sums it up => modernism? * water/#tunes nods okay, so in other words... centralization post-modernism doesn't work only because of the language used well centralization at any level i happen to think that <> makes for a nice alternative to post-modernist trash relativism of sorts? of sorts, but without the mamby-pamby lack of power in communication well, yes... that is what i hate about post-modernism... -:- SignOff smoke: #TUNES (Reconnecting) -:- smoke [smoke@16dyn164.delft.casema.net] has joined #tunes abi taf is Tossing Algebraic Flowers Down The Great Divide at http://citeseer.nj.nec.com/did/7754 hmm, what is taf about? abi taf is also about the gap between theoretical computer science and the social aspects of information processing okay, water. i hate summarizing :) well, someone has to... because the title surely doesn't :) 12:30pm the title IS a summary that's my point even my summary is the same thing it is, but it exists on a plateau no slope... to you, maybe others would find even my summary a bit hard to grasp unless you had known the context of the title, it would be hard to identity it anyway s/ity/ify hm at least i didn't hype tunes in the interview 12:40pm -:- ult [ult@user-38lc6c3.dialup.mindspring.com] has joined #Tunes heh a functional, object-oriented language called "foo" 12:50pm there isn't one afaik, but there should be :) you going to rename slate? heh no i doubt haveing the tunes/arrow hll be called "foo" would be good having, even well, foo is a very polymorphic word heh it's tempting but i'd need better reasons than that 01:00pm perhaps "foo" is the core of the slate language :) or maybe slate - mobius = foo so that foo = HLL- effectively wow google meshes nicely with odp 01:10pm damn it, there's a foo proglang i suppose it was inevitable :) nope false alarm erg n/m someone got to it first but it's not "functional oo" what is it? http://hackerlabs.sourceforge.net/projects.html not much actually you've been preempted by 50 KB of C code... -:- SignOff ult: #TUNES (Ping timeout for ult[user-38lc6c3.dialup.mindspring.com]) no, i haven't ;) i can argue that foo is an abbreviation besides i doubt that this person will mind 01:20pm -:- hcf [nef@me-portland-us725.javanet.com] has joined #tunes hey hcf hey 01:40pm hm there's some decent stuff between eihrul and i in the logs icuc i always scan the logs nate37: thx for the noise u contrib'd water: u'v looked at avail right? yes but i don't recall specifics on it * water/#tunes looks at avail again ok -:- SignOff smoke: #TUNES (z) * water/#tunes dl's the avail source to have a look-see i didn't notice the squeak implementation before 01:50pm avail? abi: avail? avail is at http://www.ericsworld.com/Mark/ interesting -:- Fufie [stig@tunnel-44-31.vpn.uib.no] has joined #tunes hm visualworks, not squeak yet eeg, explicitly typed smalltalk... sort of i'll stick with slate :) it's not the language that's interesting, it's the implementation * eihrul/#tunes nods. it sounds very self-like, afaik it's just a compiler written in smalltalk it looks to still use byte-code as well... which is bad... where did you figure that from? hrmm, perhaps not :) no i found them look in reflection-Methods numbered primitives => bytecodes doh and you call yourself a programmer. for shame ;) well... i don't have www-grep :P 02:00pm neither do i... i just knew where to look ;) but then a lot of smalltalking has to have an effect on that i suppose grep has made me lazy... yeah Interpreter.cpp also makes the bytecodes explicit this points out some potential difficulties with making a dynamic compilation based language using smalltalk compilers actually, no this will change once the author ports it to squeak yep, boot-strapping * water/#tunes returns to the papers 02:10pm bah! these papers are impossible to categorize usefully 02:20pm -:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp79.lvdi.net]) -:- Ghyll [karltk@mp-217-219-71.daxnet.no] has joined #tunes hey ghyll evening water: know if anybody's done/doing a x86/linux port of selv ? water: s/selv/self/ -:- eihrul [lee@usr5-ppp14.lvdi.net] has joined #tunes try #linpeople oh they're doing it but it's not ready yet afaik ah, ok. it does run on the mac, though as if I have one.. well, don't you know any mac users? there are no mac users in norway :) oh heh or, rather, I've only seen a handful. yeah i forgot that you're in norway my chances of bumping by a sparcstation is considerably higher. well you could emulate a mac, and then run self on *that* ;) 02:50pm hmm.. well.. I could also just it remotely on my uni's sparc boxes. probably just as slow :) yeah 03:00pm :P what a backlog of reading to do that is why we start a massive nation-wide campaign for the 48 hour day... here, relax with http://www.soda.co.uk/soda/constructor/ eihrul: I would be happy if we'd just get one more hour.. ghyll: cool, it's like thinglab thinglab ? url ? abi thinglab thinglab is at http://www.2share.com/thinglab/ThingLab%20-%20index.html or code at ftp://st.cs.uiuc.edu/Smalltalk/Squeak/Goodies/ThingLab/ its a squeaky thing not very fast yet, and the code is getting drastically-modified it's basically a constraints compiler * Ghyll/#tunes mutters something about never running java in netscape oh great i injured it :) it's just sort of flopping and twitching now that's what it does best..