IRC log started Fri Apr 14 00:00:01 2000 [msg(TUNES)] permlog 2000.0414 -:- SignOff veblen_: #TUNES (nap olympics!) -:- SignOff washort: #TUNES (Ping timeout for washort[d128.narrowgate.net]) -:- water [water@tnt-9-41.tscnet.net] has joined #tunes hey corey, hcf 12:50am -:- HenZo [henzo@bespin.dhs.org] has joined #tunes hi henzo hey are you a bespin admin or a tunes member? Both real name? ah n/m Hey hcf, can you help me out? I need a Scheme to C compiler. anything you want to talk about? 01:00am -:- SignOff hcf: #TUNES (Leaving) -:- hcf [nef@me-portland-us804.javanet.com] has joined #tunes Hm Guess not. Oh you're back. how's tril, henzo? He's fine It's my fault he hasn't been around lately. he seems busy oh so i can blame You :) I got him hooked on a game called Counter-Strike :) sorry But we've been programming too Unfortunately, not much on tunes * water/#tunes casts a cybernetic curse upon HenZo :( I keep bugging him about tunes, but he actually IS really busy.. bring back Tril and i shall release you from the abominable curse He's got a lot of stuff to do... :( I've been mass busy lately too, so I haven't had a chance to get into anything major. -:- smoke [smoke@15dyn124.delft.casema.net] has joined #tunes there's this DB he keeps promising to have set up Anyway, do you know anything about Scheme->C? i bet i can find one that'd be helpful why do you need one? Tril has like 1/4 done on the DB...hard part is done, he just needs time to work on the rest of it. I need to give someone something I wrote in scheme, but they don't have a scheme interpreter, and they're not even very computer literate. HenZo: http://gatekeeper.dec.com:/pub/DEC/Scheme-to-C/ what about http://www-rn.informatik.uni-bremen.de/software/elk/ ? hcf: I happen to be on that exact link right now. But what do I get out of there? the latest tarball, i'm guessing along with appropriate patches hmm my best guess is that everything there is out-of-date Yeah that's what I assumed. * water/#tunes has spent part of tonight working on pros/cons of slate and scheme as tunes hll (no pun intended) 01:10am HenZo: try stalin * HenZo/#tunes is installing elk on Bespin stalin, huh abi stalin i heard stalin was a _very aggressive_ optimizing scheme compiler at http://www.neci.nj.nec.com/homepages/qobi/software.html yeah.. it's really uptight * HenZo/#tunes abi stalin stalin is a _very aggressive_ optimizing scheme compiler at http://www.neci.nj.nec.com/homepages/qobi/software.html alright. screw this for now. i'm going to bed. bring back tril later guys, thanks for the tips -:- SignOff HenZo: #TUNES (Looking for Tril...) * water/#tunes has a look-see at abigail hm forget that 01:20am -:- SignOff smoke: #TUNES (Ping timeout for smoke[15dyn124.delft.casema.net]) -:- SignOff hcf: #TUNES (Leaving) -:- smoke [smoke@16dyn233.delft.casema.net] has joined #tunes -:- Closing Link: TUNES[206.63.100.13] by turtledove.openprojects.net (Ping timeout for TUNES[206.63.100.13]) -:- Connection closed from irc.us.openprojects.net: Success -:- Use /Server to connect to a server -:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0] -:- BitchX+Deb1an: For more information about BitchX type /about -:- Welcome to the Internet Relay Network TUNES (from varley.openprojects.net) -:- Your host is varley.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from varley.openprojects.net) -:- This server was cobbled together Wed Apr 28 1999 at 12 02:19 EDT(from varley.openprojects.net) -:- varley.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv -:- [local users on irc(26)] 3% -:- [global users on irc(413)] 45% -:- [invisible users on irc(507)] 55% -:- [ircops on irc(20)] 2% -:- [total users on irc(920)] -:- [unknown connections(0)] -:- [total servers on irc(34)] (avg. 27 users per server) -:- [total channels created(356)] (avg. 2 users per channel) !varley.openprojects.net Highest connection count: 58 (53 clients) !varley.openprojects.net Welcome to Open Projects! You are on 4 ca 1(2) ft 14(14) tr. -:- Mode change [+f] for user TUNES -:- Mode change [+iws] for user TUNES -:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org " -:- TUNES [tunes@bespin.dhs.org] has joined #tunes -:- [Users(#tunes:6)] [ TUNES ] [ smoke ] [ water ] [ coreyr ] [ abi ] [ smkl ] -:- Channel #tunes was created at Sun Feb 28 08:48:06 1999 -:- BitchX+Deb1an: Join to #tunes was synced in 6.464 secs!! -:- Mode change [-ws] for user TUNES -:- eihrul [lee@usr5-ppp205.lvdi.net] has joined #tunes -:- nate37 [nate@ppp-206-117-3-119.dialup.pcmagic.net] has joined #tunes -:- SignOff water: #TUNES (Read error to water[tnt-9-41.tscnet.net]: Connection reset by peer) -:- XeF4 [xef4@194.255.106.103] has joined #tunes -:- SignOff nate37: #TUNES (nate37 has no reason) -:- SignOff XeF4: #TUNES (BitchX-1.0c16 -- just do it.) -:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp205.lvdi.net]) re 06:10am -:- Closing Link: TUNES[bespin.dhs.org] by varley.openprojects.net (Ping timeout for TUNES[bespin.dhs.org]) -:- Connection closed from irc.us.openprojects.net: Success -:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0] -:- BitchX+Deb1an: For more information about BitchX type /about -:- Welcome to the Internet Relay Network TUNES (from devlin.openprojects.net) -:- Your host is devlin.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from devlin.openprojects.net) -:- This server was cobbled together Wed Apr 28 1999 at 12 02:19 EDT(from devlin.openprojects.net) -:- devlin.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv -:- [local users on irc(97)] 10% -:- [global users on irc(441)] 45% -:- [invisible users on irc(537)] 55% -:- [ircops on irc(19)] 2% -:- [total users on irc(978)] -:- [unknown connections(0)] -:- [total servers on irc(34)] (avg. 28 users per server) -:- [total channels created(360)] (avg. 2 users per channel) !devlin.openprojects.net Highest connection count: 146 (142 clients) !devlin.openprojects.net Welcome to Open Projects! You are on 4 ca 1(2) ft 14(14) tr. -:- Mode change [+f] for user TUNES -:- Mode change [+iws] for user TUNES -:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org " -:- TUNES [tunes@bespin.dhs.org] has joined #tunes -:- Topic for #TUNES: TUNES, Free Reflective Computing System: http://www.tunes.org/ || Slate Language: http://www.tunes.org/~water/slate-home.html -:- topic set by water [Tue Apr 11 20:53:53 2000] -:- [Users(#tunes:5)] [ TUNES ] [ smoke ] [ coreyr ] [ abi ] [ smkl ] -:- Channel #tunes was created at Sun Feb 28 08:48:06 1999 -:- BitchX+Deb1an: Join to #tunes was synced in 7.083 secs!! -:- Mode change [-ws] for user TUNES -:- Closing Link: TUNES[bespin.dhs.org] by devlin.openprojects.net (Ping timeout for TUNES[bespin.dhs.org]) -:- Connection closed from irc.us.openprojects.net: Success -:- Use /Server to connect to a server -:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0] -:- BitchX+Deb1an: For more information about BitchX type /about -:- Welcome to the Internet Relay Network TUNES (from sterling.openprojects.net) -:- Your host is sterling.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from sterling.openprojects.net) -:- This server was cobbled together Wed Apr 28 1999 at 15 12:37 EDT(from sterling.openprojects.net) -:- sterling.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv -:- [local users on irc(82)] 8% -:- [global users on irc(442)] 45% -:- [invisible users on irc(534)] 55% -:- [ircops on irc(19)] 2% -:- [total users on irc(976)] -:- [unknown connections(1)] -:- [total servers on irc(34)] (avg. 28 users per server) -:- [total channels created(360)] (avg. 2 users per channel) !sterling.openprojects.net Highest connection count: 106 (99 clients) !sterling.openprojects.net Welcome to Open Projects! You are on 3 ca 1(2) ft 14(14) tr. -:- Mode change [+f] for user TUNES -:- Mode change [+iws] for user TUNES -:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org " -:- TUNES [tunes@bespin.dhs.org] has joined #tunes -:- Topic for #TUNES: TUNES, Free Reflective Computing System: http://www.tunes.org/ || Slate Language: http://www.tunes.org/~water/slate-home.html -:- topic set by water [Tue Apr 11 20:53:54 2000] -:- [Users(#tunes:4)] [ TUNES ] [ smoke ] [ coreyr ] [ smkl ] -:- Channel #tunes was created at Sun Feb 28 08:48:06 1999 -:- BitchX+Deb1an: Join to #tunes was synced in 6.277 secs!! -:- Mode change [-ws] for user TUNES -:- abi [nef@bespin.dhs.org] has joined #tunes -:- smoke_ [smoke@16dyn233.delft.casema.net] has joined #tunes /scan 07:50am -:- water [water@tnt-9-41.tscnet.net] has joined #tunes -:- thomas [thomas@193.217.63.152] has joined #tunes hi tom hi what's new? I was just about to ask that same question does anyone here know if it's possible to compile a defclass and defmethod in one file in one pass in common lisp? well, eihrul's working on a new (complete) slate implementation heh i get the impression that i have to load the defclass first before being able to compile the defmethod, which is inconvenient that's nice, water water: in what language? sm: in cl at first water: aha I've been looking at an interesting lisplike syntax/semantics, where identifiers can be objects as well as the object their refering to. 08:30am hm is this your own idea? specifically, a object named for example '( do a b )' would correspond to a method with two parameters. well what if you want to partially-evaluate a method? the idea occured in a work-project a work-project on what? well, for example, if I write '( do c d )' it would be translated into the effect of the method, where a and b are substituted with c and d. ( don't know how that could answere it, really :) oic the first example was a declaration it was just the name of the method. well suppose i had (do a b c) and i wanted to specialize it on a and c only? in it's definition, it would refer to a and b - and those references are the once that are substituted. you would , for examle define '( do a b )' as '( do a b c )' huh? i said a and c ok! to make a new method that only has b as its parameter define '( do a c )' in stead. ok, then ( do b) is defined as ( do a b c ) that wouldn't work it seems in the declaration, the symbols are the names of the parameters, But... in order to specialize the declaration, you have to bind some of the parameters with values but not all of them I know, this part is not finnished yet. the semantics are not complete. -:- vvater [water@tnt-10-42.tscnet.net] has joined #tunes 08:40am i don't see how you can do that with the system you mention (without totally re-writing the method) (i got cut off for a minute) I'm not trying to create this specific language, just want to observe some effects and problems related to it. well, i'm just trying to point out the limitations ;) I know, the whole matching of identifiers and doing the correct substitution is a problem - but an interesting one .) how do you make symbols into objects? is this like my slate objects' slots being objects as well as parameters? -:- SignOff water: #TUNES (Ping timeout for water[tnt-9-41.tscnet.net]) -:- vvater is now known as water yes, you could say that. In TOOL, parameters are slots/attributes. And that design decision is probably final well you know i set parameters by passing messages, right? yes, I suppose you suppose? I'm haven't looked to closely at slate 08:50am i can give you an overview if you'd like sure i'll try to keep up ok. are you familiar with objects in self? alittle, prototyped using "slots" right well a slate object is a set of slots that refer to other objects that act as methods even state in slate is just a method that returns a value ok well, objects are also namespaces that you can navigate like directories in a file system you can clone objects in order to pass values around so, objects are immutable yes but slate also has meta-objects that define how each object's behavior gets invoked how do they fit in? or even whether it gets invoked (maybe depending on the senders identity) well each object has a 'parent', enclosing namespace '..', and 'meta' slot the meta-object actually implements the message-handling via a lookup and apply set of phases and each object's mo can be customized for its own use, even for special situations does this make sense? so the meta slot is like the invisible, implisit method that says 'invoke' / evaluate or send right yes, that make very sence! of course you can switch mo's or meta-behaviors dynamically 09:00am btw this mo idea came from moostrap originally yes, because you've given them a name so they can be refered to yep it's actually very much like TOOL. one thing that makes meta-behavior powerful is the ability to specify the meta-behavior of a slot's result so if you clone an object, you can make it into a dumb reference or another identical independent object, or something in between (clone is another slot) the difference may only be, that in TOOL I emphasize the use of first order objects. huh? how does slate not do this? sending first order objects is a subsitute for sending pointers. um slate does this I don't know if it doesn't, just that I think it is important cool as i said above, when you clone, meta-behavior customization can turn the clone into a dumb reference there's also another small aspect to slate so I can make for example a reference to a first order object - that is, put a first order object into a reference every object acts like a function, in that all objects have a result slot that can get invoked implicitly sort of in slate, you clone to get a reference yes, meaning you clone the whole representation of the object but in order to preserve the original object's identity, you have to force the reference to never modify itself (unless you re-target the reference i suppose) actually, not quite if i cloned the whole object explicitly, then it would have seperate state instead my references are more like objects with only 'parent' set to the refferred object and 'meta' set to the reference-like behavior but if they have the same state _and_ are immutable, then they must be read only one moment! the state stays in the original object, although the environment might implicitly hold state at the reference in order to cache values (say on a distributed system) no, in slate you get more options for immutability than just "read only" immutability in slate is just a result of meta-behavior * water/#tunes waits for thomas 09:10am -:- hcf [nef@me-portland-us729.javanet.com] has joined #tunes re i think i have this tutorial outline figured out water: tutorial? yeah a slate tutorial cool water: i dont fully understand the above conversation but it sounds nice. :) hm then maybe i should make two versions one for people who grok lisp and/or self already 09:20am i'm back and another for your run-of-the-mill procedural coders water: :P tom: check my comments above * thomas/#tunes reads corey: anyone interested in tunes should check out those languages yes, sure, you can always define different message passing schemes and policies meta-behaviors do you consider meat behaviour the way and object interprets/recieves a message, or the way it sends a message? water: 2 tuts sounds good receives, although sends can be intercepted if you use a little trick i learned how? if you make a clone into a proxy-like object in the namespace local to the sender, you can intercept sends a wrapper yeah except proxies can wrap from various directions of course hm that covers most of the concepts just notice that slate slots using meta-behavior of the slots themselves can be made into all the types of variables or constants you need despite prior discussions we may have stuff in common :) oh wait, i forgot the literal systems do you have a slate syntax? well right now its SEXP-based (object message) with everything reducing to messages at the moment (woo hoo!) I think thats' the way to go yeah well it's not easy 09:30am some things took a while to reduce to message-passing and i'm even looking right now at reducing ( and ) to messages >:) ok maybe that's going overboard, but it's still an interesting prospect brb k water: how well would a system like this perform? well it simply *needs* dynamic compilation to work properly i.e. static compilation will not be a simple option but if eihrul works it out right, he'll use run-time type feedback to improve on the self figures (which are 50%speed of compiled and optimized c) without a static compile 09:40am hm it seems like its only a matter of time now before slate is an actual tunes hll candidate 09:50am water: where do you think your philosophy differs from where you think tunes is going? tunes is programmer-centric they don't generalize the quotienting requirement of Tunes HLL to ontological relativism they also have no idea what substructural thinking is about and they also don't generalize tunes beyond computing qed 10:10am * water/#tunes adds a section to the tutorial addressing Lisp fanatics -:- beholder [beholder@dialin-187-67.sudbury.primus.ca] has joined #tunes hey! water: Hey what's news? (i'm trying to round up tunes people, btw) wow, there isn't usually ppl here during the day :) well this is my weekend water: Good, to discuss your new implementation ideas? yes, among other things water: Well, it's nice to see you here again :) so how available are you lately for tunes? beh: i'm here most every day water: Truthfully, not very. I start a new job on Monday which will eat away my 9-5, and after that I have a group of people I teach unix stuff to every night (they just kinda followed me home one day ;) :/ 10:30am ditch 'em! :) Ok, for you I'll ditch them a few nights a week :) well you could help Tril with the tunes web DB Is that project still going? apparently he has the hard work done but there's still some stuff left to take care of, probably mostly simple things if you would contact him about it, i'd appreciate it Well, I'll get the details from him, but in my current state I can't make any real promises. But I'll do my best :) as for Tunes itself, bill tanksley's getting into Joy as a LLL and i've got the Slate spec nearly ready Do the two relate? and eihrul's working on slate implementations hm well joy is like forth except its mathematically formalizable (combinators) so it's definitely a LLL, not something you can program with as an end-user and slate is turning out to be quite a multi-paradigm kind of language Do you believe slate is one step closer to a working arrows system? yes for reasons i haven't explicitly stated Yes, I noticed that, can you state them now? well how much do you grok slate? (btw i'm writing up a slate tutorial as we speak) Well, not very well :) I haven't seen the site sice the initial post, which was only a vague description alright well i made up a little explanation just a little earlier today, check the logs water: I'm on the site at the moment, would that be better to explain? also i explained it late yesterday in a slightly different way hm most likely not the site needs an overhaul lately, since it grew ad hoc (which was the best i could do at the time) but reading the site after the explanation in the logs might be beneficial Ok, what's the log URL again? abi logs logs are #tunes logs at http://www.tunes.org/files/irc Thanks, checking it out Woah, someone is gonna have to form a directory structure outta thoese log files, that's a huge listing :) 10:40am talk to tril about that beholder: u want http://www.tunes.org/files/irc/2000.0414 hcf: Got it and perhaps 0413 water: Where you're talking to thomas? yes both beh: yep -:- coreyr is now known as corey Damn phone keeps ringing... grr.. Hmm, object namespace hierarchies? yeah like a file-system That's contrary to arrows :) the hierarchy is illusory, though 10:50am Of course So this language could be easily implemented? Sorry, I mean more easily implemented than arrows? i think so, yes oh yeah, definitely easier than just going straight for arrows How far are you along, anything I can try? hm our current lisp evaluator is way out of date eihrul's cooking up a new one but the old code is still on line Hmm, if it helps any binENG is making a Forth interpreter using lisp for a school project. I'm not sure if it has any relevence to you project though. yeah i heard forth interpreters are quite simple in lisp Yeah, I told him to do a C compiler... but for some reason he balked at the idea ;) heh I mean if you want a challenge... hehe have you gotten to the web docs now? Still reading conversations, quite a few interruptions here... too much life k What is a slot? uh well objects have methods and vairables in ordinary oo, right? yeah in slate, variables are methods, and i call both slots 11:00am call 'em sub-objects if you want to Oh... hmm, I'm reading through thinking it's something more complicated that I'm having trouble comprehending... hehe ;) Language has so many problems... yeah the ideas are pretty simple (i came up with the slate idea in '95, btw) Not slate, the English Language, or any other human language for that matter just after i started reading the tunes language hm would you be interested in <>? :) Nahh, I'm more into (^) than anything at the moment ;) *? Hehe, sorry stupid joke ;) oh abi <> <> is Modality or a diamond symbol or your Arrow-like human language idea or additional blurb at http://www.tunes.org/~water/ abi: (^) is beholder's own knock-off language that advertises itself as being just as good as <>, but in reality it's made by Canadian immigrants in slave labour camps heh So how's the tutorial going? The spec sounds interesting. hm not bad. it's mostly built around the recent conversations as well as stuff i've told to newbies to explain the basics of tunes programming ideas Oh, I guess I missed it before.. it IS a meta-language ;) 11:10am * water/#tunes decides to hyperlink to tutorials of other languages, and quote them as well yeah in multiple ways, actually Yeah, not just syntatically (sp?) syntactically right -:- NetSplit: fontana.openprojects.net split from tolkien.openprojects.net [11:12am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [fontana.openprojects.net] So it is still "code" based... unlike arrows. People will still have to resort to typing code even in abstract domains right? yeah but the code is just a sequence of symbols What? -:- Netjoined: fontana.openprojects.net tolkien.openprojects.net -:- smoke [smoke@16dyn233.delft.casema.net] has joined #tunes -:- smoke_ [smoke@16dyn233.delft.casema.net] has joined #tunes it doesn't depend on text each identifier is just a tag for accessing an object which i intend on abstracting from english textual basis And each identifier COULD be a string of characters comprising a human word then. Ok, understood actually the sequence part is illusory as well so in the futurist scene, one could picture a person manipulating arrow objects with their hands in a vr environment and be modifying code? How so? but you get the ides (object message) syntax is just a link between values you can connect such things together into the whole system as a kind of graph corey: Not modifying code, modifiying where the arrow points, it might not have to be code I imagine. yeah which is why people will receive a Slate Object :) corey: Arrows is a unified system where everything is data and everything can point to everything else it's also a way of thinking corey: In a meaningfull way hopefully :) grokking the arrow concept involves being able to manipulate exactly the right arrow to get a desired effect water: Yeah, I guess that's the "meaningfull" part :) You can point to anything, that's a pretty long noose to hang yourself with ;) right but then in an arrow system there are too many to effectively see at once the trick is to know how to manipulate your view of things Which is the ontology concept. * water/#tunes nods * beholder/#tunes still remembers the many conversations :) 11:20am Are you going for the "visual programming" concept for the first implementation of slate? That might take extra time... perhaps oh wait no not the *first* implementation more likely the third Ok, sounds reasonable yeah we're looking at developing the usability in the same sequence as self In the same sequence that self was developed? skipping some unnecessary stuff of course yes, roughly What lang are you writing this lang in? lisp, right now Common Lisp? i'm going to put together a squeak implementation as well yeah, CL What CL interpreter are you using? I mean, which/who's imeplementation oh i don't think it'll matter we'll keep it pretty portable at first Yeah, that's the concern. I hear even lisp is not totally portable... sad really most of this is up to eihrul and whoever else joins in Yeah, coders choice. Good call. eihrul recommended cmucl to me so i would infer that How long do you think it's going to take? Is the implementation details solidified yet? eih said it would take a week or two to have an evaluator that demonstrated all the language features then it would take at least a month or two to move to a full library of primitives and an actual dynamic compiler Hmm, very quick. of course i'm going to encourage and help him as much as possible to get something really usable by year's end, when i have to go back to sea :( (object message) wow, very simple :) water: Do you bring your laptop to sea with you? namespace access makes it a bit more complicated than it looks yes (\global\functions\print "Hello, World") ? kind of sorry, syntax isn't fully cemented yet (the sugar, that is) How would you do the same thing I just demonstrated in your syntax? 11:30am well there are several ideas i'm looking at pure sexp-style: (global (functions (print : "Hello, World"))) ":" hides some namespace access btw since strings have their own namespace :) yes like "use default method" type thing no not like that more like "clone this namespace here and access the following slot from it" -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us729.javanet.com]) Hmm, so before an object can be used it must be cloned? another variant on the syntax decomposes the SEXP a bit yes global / functions / print : "Hello, World" \ \ Hmm, that example seems to be a lot of typeing... got any ideas that are shorter? -:- _ruiner_ [blah@ppp468.wi.centurytel.net] has joined #tunes with the last two being optional i think Never mind, that one was better