IRC log started Fri May 12 00:00:01 2000 [msg(TUNES)] permlog 2000.0512 -:- SignOff smoke: #TUNES (Ping timeout for smoke[16dyn161.delft.casema.net]) -:- smoke [smoke@15dyn51.delft.casema.net] has joined #tunes -:- SignOff water: #TUNES (Ping timeout for water[tnt-10-23.tscnet.net]) -:- water [water@tnt-9-233.tscnet.net] has joined #tunes re 01:20am -:- dalvarez [dalvarez@tristan.sc.cs.tu-bs.de] has joined #tunes hey dalvarez :) anything on your mind? sure. but nothing actually relevant for tunes :/ what lisp interpreter is used for tunes dev? anything supporting common lisp specs or the scheme specs and tunes dev is not unified right now, as no designs have been approved why don't you try coding up a HLL environment and see how far you get ;) s/get/get?/ would you like URLs? abi: where is lisp? i think lisp is ((simply) (perfect)) abi: lisp? somebody said lisp was ((simply) (perfect)) damn 02:10am http://www.lisp.org/ it is not that hard to guess abi: scheme well, scheme is a dialect of Lisp that stresses (an ((overly) minimalist) concept) of elegance or good for teaching or at http://www.schemers.org or see rnrs 02:20am -:- witten [witten@un.torsion.org] has joined #tunes hey witten hey how goes slate? well eihrul's working on the compiler (mobius) and i'm working out the finals of the syntax and the core libraries cool well we're also loaded down lately by work and school yah of course although i go on vacation tomorrow, so i can spend more time on getting stuff right did you look at the tutorial work? nope not yet icuc http://www.tunes.org/~water/slate-tutorial.html 02:30am i'm such a sucker for interesting languages i could add a lot of stuff, but many ideas are still not finalized, although the semantics of the language and libraries has been agreed on well that's definitely something well, you're in the right place ;) heh unfortunately, syntax (aka factorization of functionality so it fits the language model) is a difficult part of the language design yeah, i hear ya well if you have any comments or questions on the stuff in the tutorial, please let me know i will eww, parentheses :) -:- SignOff dalvarez: #TUNES (bbs) :P suck it up :) hmm, making assignment a method. interesting.. 02:40am yeah it makes stuff really tricky to handle under the hood, though i can imagine i'm working out the necessary stuff lately, which isn't incredibly hard, but it determines lots of little minutia concerning the libraries and oh yes, the number and string and logic systems will be in user space cool we're still figuring out what's the "tunes way" to make this work right, syntactically -:- SignOff smkl: #TUNES (Read error to smkl[glubimox.yok.utu.fi]: EOF from client) like plugging in primitive access to machine registers to the user-level object systems yeah, that's tough got any suggestions? or maybe you know of a system that does something like this in an interesting way? i'm trying to make a system that does something vaguely similar, actually i'm running into problems with the "return the object" vs. "return the result of the object" which i see you've solved, explicitly yeah we spent some time working that one out on IRC (many argumentative evenings there) hehe i think i'm going to do it explicitly, too, but the opposite way yeah i've been considering making the 'result' call explicit oh, it's not explicit now? perhaps that option could be left in user-space well, i mean forcing the programmer to call 'result' explicitly oh i thought that's the way it was currently just a syntactic issue only -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes yeah 02:50am no, currently, "/" prevents the 'result' slot from getting invoked question.. do your evaluations return slots? ah ok cool then it's the same way i was going to do it more or less (it makes it act more like a file system) yah yes all evaluations return slots from somewhere then how do you extract the value from a slot (like, say for assigning it to something) heheh because if you evaluate the slot, you'll get a slot the slot _is_ the value; read on in the tutorial which is sort of a problem :) no it isn't :) i read it how do assign one slot to another? it's right near the end well, the ":" slot sets the parent of the sibling "result" slot to the value (!) hm so result inherits the appropriate behavior from the value.... inheritance subsumes assignment yeah there's a whole interactive session in the tutorial at the end where "x" is explored within "myPoint" yah (keep in mind i haven't even gotten to meta-objects yet ;) which is a large part of what makes this system work yeah, i remember that since the basic notion of a file system *is* the bare slate object system... it takes a lot of customization to have such a system act like a real programming language what kind of customization? 03:00am well, everything from namespace security to mutability to message-forwarding to inheritance control ah which covers a lot of things that you wouldn't ordinarily think fall into those categories so you're keeping the core language really minimal? yeah so fscking minimal that it makes lisp 1.5 look like bloatware bahah well even though the syntax-level stuff is still up in the air, the basic object model and meta-object model is well-understood enough by both eihrul and i that he can implement most of the compiler and be sure that it will be translatable to slate later on fairly simply hm verbose... oh well 03:10am well at least someone understands it :) hopefully i will iron out the remaining issues during the next 2 weeks and be ready to post the rest of the ideas into the tutorial after that, i'll totally update the site and start doing some PR into the tunes mlist cool i'll probably ask for another update to the evaluator that eihrul wrote yeah the one that's posted right now barely supports the tutorial and that's it afaik -:- XeF4 [xef4@194.255.106.43] has left #tunes [] well, i better sleep ok cya nite -:- witten [witten@un.torsion.org] has left #tunes [] 03:20am -:- SignOff water: #TUNES (The Tao went that-a-way!) -:- antoine [antoine@psytrance.egenetics.com] has joined #tunes -:- water [water@tnt-9-233.tscnet.net] has joined #tunes hello antoine well, it seems Fare is up to his old tricks... becoming a sychophant of someone who's got experience (above is a comment concerning a new mailing list thread) (what follows are comments for the log to be discussed later) if Fare really intends and is capable of turning Tunes into VC-fodder, then there may be some problems, as the vast majority of Tunes members and lurkers are attracted by claims on the Tunes web site that Fare has no clue how to back up his current attitudes suggest that he totally rejects anything but Lisp ideas now, so that there may be a split of some kind soon of course this hinges on his being able to either finish his PhD effort (which seems less likely than otherwise due to his lack of formal credibility in his writing ) or his being able to salvage VC money if he doesn't manage to gain a respectible PhD hm i think i'll stick with completing my own designs and ideas and formalizing them anyway, i'll be back at 6pm PST today to discuss this and respond via mail to Fare and perhaps others -:- SignOff water: #TUNES (The Tao went that-a-way!) 03:50am * antoine/#tunes is away: pontificating -:- antoine [antoine@psytrance.egenetics.com] has left #tunes [] -:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #Tunes -:- hcf [nef@me-portland-us420.javanet.com] has joined #tunes -:- SignOff hcf: #TUNES (Leaving) -:- doc777 [attillas@mg053-82.mailgate.net] has joined #tunes http://www.tunes.org/~water/slate-home.html' 07:50am -:- doc777 [attillas@mg053-82.mailgate.net] has left #tunes [] -:- dirt [lysergicac@stargate.intelligenesis.net] has joined #tunes -:- water [water@tnt-10-138.tscnet.net] has joined #tunes >:D hi all :))))))) -:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp44.lvdi.net]) is anyone here? (corey, dirt, kyle, sami, smoke, tom?) 08:30am hm there's a new arrow/slate fan 09:10am Reto Schmid who works here: http://www.ifi.unizh.ch/req/ 09:20am mmmm MathMorphs :) 09:50am -:- _out [aa@200.201.30.81] has joined #tunes hey, _out -:- SignOff _out: #TUNES (Ping timeout for _out[200.201.30.81]) -:- _out [aa@200.201.30.81] has joined #tunes 10:10am re out <_out> hi is there anything i can do for you? answer questions or discuss or explain something? <_out> not yet, just staying around by now ah lurking 10:20am -:- SignOff _out: #TUNES (Ping timeout for _out[200.201.30.81]) if only the irc protocol had a L-flag :-) heh 10:30am -:- NetSplit: clarke.openprojects.net split from sterling.openprojects.net [10:39am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [clarke.openprojects.net] -:- Netjoined: clarke.openprojects.net sterling.openprojects.net -:- water [water@tnt-10-138.tscnet.net] has joined #tunes -:- SignOff dirt: #TUNES (Leaving) -:- hcf [nef@me-portland-us400.javanet.com] has joined #tunes re hcf hi javanet/rcn r having "issues" so their users cant access sites like www.tunes.org what's javanet/rcn? oh your isp yeah, rcn owns jn so what r u and ur web guy planing for ur pages? 10:50am well we discussed a few things last week but until today i had no time to go meet with him i'll be hanging out with him a few times this weekend so we'll probably get something done about it at the very least, i'll give him my bespin account info and let him hack it into something he's comfortable with what do you think of the recent postings to the mlist by Fare? which were those? (note, i cant reread them atm) "Re: Ken Dickey on a new OS" oh apparently, there's someone interested in tunes that Fare is interacting with (perhaps on usenet?) woa, hmm, i can access uk.tunes.org tho :) heh that's abit odd it seems not bound geographically, i can access www.cs.washington.edu, yet not bespin hm 11:00am i havnt seen any posts to usenet by fare recently see, tunes.org should be a frontend to a system that lets us know what ppl r doing instead of having to directly talk to fare also, tunes should (imo) have an official attempt at .. this problem developed in '95 iirc dealing w/ interesting poop from usenet and other communications hm and ppl should take part in these things on a regular basis * water/#tunes eyes the channel lurkers with some ire well, BillT i know is active, as well as Jecel http://www.mars.dti.ne.jp/~umejava/object/oolinks.html links to slate wow too bad http/html didnt have an easy way to determining what pages link to one of urs yeah it makes sense, this group seems really interested in self this guy, www.andrewcooke.free-online.co.uk/andrew/index.html doesnt link to u i dont think but they don't even mention smalltalk or squeak but did mention slate some comp.lang.misc no, i've met him before he's a usenet daemon i got a mention on CLM? yeah, by him recently? hold i'm looking myself http://x40.deja.com/=dnc/[ST_rn=ps]/getdoc.xp?AN=602512328&CONTEXT=958154801.1621557270&hitnum=0 i gave u a link to this post b4 oh i don't recall, though 11:10am i give up for now, bbl -:- SignOff hcf: #TUNES (Leaving) k -:- dirt [lysergicac@stargate.intelligenesis.net] has joined #tunes re dirt hey water is there anything you'd like to talk about? hmm don't think so -- thx for asking though ok it just seems you're here quite often when we have logs online that can be scanned true and a serendipitous thing it is too it really should get organized better, but Tril seems to be busy lately because if i'd stuck to scanning logs, i wouldn't have found the #python channel :-] heh ok sorry, i just feel like the discussions here are more ethereal than i'm comfortable with * smoke/#tunes would like a talk about the future of guis.. are there any experts in the gui field in tunes? np hm mostly due to my lack of hard core comp-sci background 11:20am smoke: well, i work with morphic quite a bit and have had lots of ideas about guis i read on the mailinglist again this type of remark: "worry about the core first, the gui will come later" (quoted very freely) yeah i think that's a scary remark why? the gui will be shaped by the programming language i don't believe in a core that can be so powerful as to embrace 'everything' later on heh smalltalk sound familiar? hm? um squeak supports several guis because the vm only supports the bit-buffers ah i meant the other way around i guess which way around is that? start from thinking about a user interface to come to a core i propose redoing the smalltalk approach besides, interface issues are already part of a different project from HLL for a good reason why? hm. hold. i know morphic doesn't suite smalltalk perfectly, but that's like throwing out the whole environment just to get a new gui perhaps it's better to start from the core at the lowest level and gui at the highest level to converge to a mid-layer.. wonder if that would be the HLL then... (sorry for thinking out loud) well, the Self fellows did a bit of that they did prototype-based OO because the interface would be better that way hm yes OO is a very useful concept for the user you should try MathMorphs btw abi: mathmorphs? i guess mathmorphs is a project combining Mathematics and Smalltalk at http://www.dm.uba.ar/MathMorphs/ hey, interesting yeah direct-manipulation stuff hm that url is dead, i'll feed abi the new one http://mathmorphs.swiki.net/ would it be possible to make a computer generate usable guis btw? i dislike the current trend of (braindead) humans designing guis by hand well you'd have to have a formal algebra of interface ideas like the idea of buttons for example oh wait you don't mean widget-generation, you mean layout? well, basically everything 11:30am dude that would be ridiculously difficult i mean there are research projects that show how to do some things properly with automatic layout etc, but it's not mature stuff at all for example, to control a factory with some conveyor belts and robots and things.., you'd have to pinpoint what data could be altered, and since the controlling program (could) define(s) how the data relates, a gui could be built out of that automatically this would require that the programming language explicitly models (and shows) the relation of data relationships even, sorry yeah which sounds too much like AI well, the machine doesn't have to do any thinking hehe yeah that's what you think until it screws up your gui haha :) "hold i'll just place this button right ... here.. for your convenience, hold..." seriously most of those projects reach a point like that could you point me to some researchers who're working on this? well you can mine cora for papers on interface stuff abi: cora cora is probably at http://www.cora.justresearch.com/ there are some good ideas there thanks what does CORA stand for? i forget (oh duh, there's an About CORA link) * smoke/#tunes shuts up :) thanks for the hints, water.. np there are many websites for ui projects btw are you familiar with vpl ideas? it's something i'm very interested in for slate, but no one has successfully pulled it off except for ProGraph visual pls ? yeah hm well, what i've seen so far didn't raise my interest :( what have you seen? squeak, smalltalk, and microsoft tools :) bah! haha :) those aren't visual, you dolt! :) they make visual things, but they are text-based hey it says so on the cover! :P abi: vpl? vpl is Visual Programming Language grr oh you mean reaaaal visual programming without text at all? yes, although labels count like, drag this `thing' into this other `thing' and then *kazboom* you've got a `thing' ? (hm that would be audible programming also) so morphic is sort of visual (if you use only morphicwrappers) yeah that's an oversimplified view it's not just mouse-action-based for example, there's program activity visualization ideas 11:40am hm i'd like to quote this "we invented text because we got sick of drawing mammoths and frightened people on cave walls" heh hm well, i do know some about scientific visualization mm this is about visualizing algorithms and data structure live i've used AVL or so, with which you could connect datastreams to create an output image water: `live' ? water: while it's running? yep ehm or also after-the-fact, as a profiling tool hard to visualize (with benefits) hm maybe i still have the urls i'm reading about Fabrik yeah that's what squeak is looking at in the near future hm sounds rather limited.. (i've explained that before) yeah i know it's not the greatest hm - so what's your interest in vpls then? as a basis for a ui well, both ways really hm yes taking the best ui concepts and rolling them into a uniform vpl syntax it feels like the current uis are a subset of something larger, which we haven't found yet damn some of these links have died hm i can't seem to combine a command line interface with riding a bike, mounting a horse and clicking on buttons in MS Windows well when i was in college working on what is now slate, it had a visual syntax oh wait 11:50am abi: wouter? hmmm... wouter is the creator of many interesting programming languages at http://www.ecs.soton.ac.uk/~wvo96r/proglang/index.html some of those are vpl's keep in mind the only good production vpl is Prograph abi: prograph? prograph is, like, a visual data-flow language with oo extensions at http://www.pictorius.com/prograph.html i first used prograph on my friend's powermac a few years ago the .com sounds scary don't tell me it's mac only :) no it's mac and wintel and yes the company was formerly a sane one :) anyway, it's a really nice development system, especially for oo i'll have a look at their webpage most dataflow vpl's are like prograph, usually with extensions or different layout ideas ok i love those 'faster than c/c++/java' figures :)) before testing either, i know developing with haskell is twice as fast as with prograph :) well, i intend on developing an entire visual interface algebraic system using slate heh prograph doesn't have a conservative PL at its baes? base. no, it's dataflow ie a kind of functional language with control-flow perks aha i was very intrigued by a 2d language i came upon one day befunge? :) the author explained that it was very useless yes! befunge it seemed very useful to me * water/#tunes knows of all the interesting proglangs for what? but one would need a 2d processor too, and those are hard to craft no, you can emulate the 2d part well, for parallel processing to start with yes, but emulating would make it less powerful than it could be yeah if one could contrive a 2d processor, it would bring new ideas well i'd rather use a kind of functional language so i don't have to stick with a grid your arrow system was multi dimensional, no? * water/#tunes nods cant a 2d celluar automata do processing, or am i totally mixing ideas? 12:00pm no this is a bit different yeah i'd love to fiddle with arrows one day.. great concept but yes you're right, corey well if i get some professional researchers to help out (which someone contacted me the other day) * smoke/#tunes imagines himself optimizing fourty arrows in a specific area of an until then not understood network s/me/me about/ (fiddle fiddle) heh well, i'd better be reading some instead of wasting your time.. thanks for the fish :) np i like it when you talk i pick up things so ask some questions! :) you know i started doing cs research in '94, when tunes started i spent that year in a farmhouse in the desert reading and dreading middle school. :) coreyr: if you want to hear *me* talk (which i doubt), visit #offtopic or #demoscene :) which desert was that? the high desert, eastern oregon oh 12:10pm hm no one has stuff to discuss? 12:30pm -:- SignOff dirt: #TUNES (Read error to dirt[stargate.intelligenesis.net]: EOF from client) -:- dirt [lysergicac@stargate.intelligenesis.net] has joined #tunes re -:- eihrul [lee@usr5-ppp46.lvdi.net] has joined #tunes hey eih lo -:- SignOff dirt: #TUNES (Read error to dirt[stargate.intelligenesis.net]: EOF from client) 01:20pm -:- dirt [lysergicac@stargate.intelligenesis.net] has joined #tunes -:- aligod [Aligod@98C8B1E2.ipt.aol.com] has joined #TUnES hi aligod hi water !slate-mp3 is there anything i can do for you, ali? no, no mp3's here :) hmm Slate wtf is that? it's a programming language currently in development it's a kind of cross of lisp and smalltalk, if you've heard of those people with lisps making small talk? heh