00:10:17 [QUIT] air quit: http://www.qzx.com/ :: sleep 00:59:50 lalhira joined #tunes 01:00:52 [QUIT] lalhira quit: lalhira has no reason 01:29:06 [NICK] JALH changed nick to: Mike_L_ 01:50:51 water joined #tunes 01:51:18 oh yeah, i'm relaxed :) 02:46:23 [QUIT] JALH quit: [BX] Time wasted: 2h 30s 05:41:38 eihrul joined #tunes 05:42:00 re 06:43:05 [QUIT] JALH quit: Ping timeout for JALH[host213-123-40-106.btinternet.com] 07:26:33 water wakes up 07:27:56 ok 07:28:10 i went clubbing last night 07:28:14 anyway 07:28:50 how would you like to start today? 07:29:07 hmmm, maybe with an all-canadian breakfast 07:29:26 heh 07:29:49 i was thinking more about discussion and such 07:29:52 ;) 07:29:54 (that joke does not mean what you think it does, but id igress) 07:30:37 up to you 07:31:10 i just woke up, so i don't have a preference yet 07:31:44 coreyr joined #tunes 07:32:03 hey cor 07:34:38 hm 07:35:16 i have an idea about the meta-level object graph representation 07:35:31 although i don't think it's a final solution 07:36:18 anyway, it seems possible to build a notation for node names that is centered around the MO's first-order object 07:37:44 (i keep getting interrupted) 07:38:12 sorr 07:39:08 so anyway, since most of the objects that we want to consider will be local to the object, and the MO's view of things should be relative to that object, i'm considering that the "pathname" of the other object nodes in the graph denote their identity 07:39:48 although it wouldn't be a pathname, just a vector (right word?) of the lookup values 07:40:01 (pause for questions) 07:40:38 what is the practical benefit here? :P 07:41:00 heh... 07:42:13 well earlier i was having some problems with the idea that these meta-level graph representations would have to depend on some arbitrarily-generated ID which i believe would have led to a lot of unnecessary problems 07:43:27 does that make sense? 07:43:57 i'm most concerned here with being able to express the necessary graph rewrite rules within the language 07:44:29 a simple graph ADT per se would not be enough 07:46:02 did that help? 07:46:31 again, not really 07:46:36 which is not the fault of your solution 07:46:46 but the fact that i don't really understand what the hell the problem is 07:47:16 well, how would *you* suggest expressing the graph-rewrite rules? 07:49:03 eihrul wonders if water is getting too caught up in graph rewrite as a particular interpretation of languages as opposed to graph rewrite as a language. ) 07:49:21 even if we simply generated unique IDs, it would be an unnecessary abstraction layer imo 07:49:24 otherwise, we're not thinking of the same slate :) 07:49:25 P 07:49:44 i'm talking about the "apply" phase, dude 07:50:06 well, that would be expressed in slate code... 07:50:17 grr 07:50:23 duh 07:50:47 and i don't particularly favor graph rewrite as a way of programming large systems :) 07:51:01 this is not the usual graph-rewrite 07:51:50 mostly because the overall system would *never* get reduced to a normal form 07:52:23 i never though graph rewriting need necessarily produce a normal form, but that's a separate issue 07:52:33 and how the hell can slate code express anything if it can't be applied? 07:53:04 it's not like we have cool list-structure constructs like FP languages do 07:53:30 it's not like we can say "do this" because slate automagically has them by definition 07:53:38 because it doesn't 07:53:43 it can... 07:53:50 but it shouldn't 07:54:16 what is wrong with having certain objects whose initial implementations are expressed outside the language? 07:54:23 at some point, you need them 07:54:26 you CAN'T avoid them 07:54:28 or maybe this whole HLL requirements printout i have should just be thrown out ;) 07:54:39 *sigh* 07:54:52 i know this just as well as you 07:54:59 the HLL requirements are like the Constitution 07:55:12 depending on whose reading it, you'd think it was a different document 07:55:15 do NOT use that metaphor, please 07:55:22 heh 07:55:43 i think it is appropriate 07:55:45 i'm saying this would be the way to bootstrap slate-in-slate for instance 07:56:04 i don't care if it is or isn't 07:56:16 well, you could just as well use the slate "primitives" to express the primitives themselves... i don't see how that can't solve the bootstrap issue 07:56:32 but i definitely know i'm one of the few people who can follow even a half of what Fare says in it 07:56:45 well, you said Fare can't even follow what he says in it :P 07:56:53 because what's primitive and what's derived may change! :) 07:57:00 yep 07:57:18 because what's primitive and what's derived may change! :) 07:57:19 because what's primitive and what's derived may change! :) 07:57:36 well, nothing says you can't use the primitives to express a different set of primitives... 07:57:42 in fact, we want it to be able to change, right? 07:57:43 which once activated, are now the default set of primitives :) 07:58:03 but what of having the open MOP? 07:58:15 (the MO's protocol) 07:58:55 are we to say "don't screw with apply, because that's the business of the core developers"? 07:59:34 this language should allow people to change those graph- rewrite rules that are part of evaluation 07:59:47 in ways that aren't too complex 08:00:01 which ordinary graph ADT's are 08:01:14 (ya know, it's kinda scare how "harmonia" and "tunes" are related both as words and the cs research projects they are) 08:01:25 s/scare/scary/ 08:01:50 now is what i'm saying making sense? 08:02:14 how does not using "graph rewrite" or whatever formalism you want to throw at somebody aid in this goal? 08:02:15 coreyr: is anything i'm saying making sense to you? 08:02:28 *not* using? 08:02:33 it only elevates the means by which to change these things to a level beyond the user 08:03:01 i thought we already went through this 08:03:31 apply is not some complexish graph rewrite mechanism... 08:03:37 it is merely a damned method defined in a meta-object 08:03:41 we want to be able to express that when things get muted, sometimes we want entire parts of the context to get implicitly cloned, for example 08:03:46 what is so mystical about this? 08:04:05 ok, if apply is defined in the MO, tell me how it would work, smart-ass 08:04:48 especially since we need lots and lots of primitive apply methods just to make slate work 08:04:56 a message is send to the MO to do an apply... 08:05:04 s/send/sent 08:05:16 as well as parametrizable apply methods to generate new ones on demand 08:05:25 water pauses 08:05:55 i don't see what the issue really is here 08:06:38 how about the fact that objects are indistinguishable except from the state of their surroundings? 08:07:23 so to know how an object will behave when applied, you'd have to be able to express that apply method in a uniform kind of way 08:07:50 so you could *exploratively* look into it and learn how each object works 08:09:13 btw this graph-like adt thing would just be a kinda of system-wide mirror 08:09:24 "adt"? 08:09:34 abstract data type 08:13:08 is any of this starting to make sense yet? 08:15:59 if you're just talking about having a model of the code in the system, i thought this was already a given? :P 08:16:13 "a given"? where? 08:16:35 well... by being able to ask code what it does, what it is... 08:16:51 oh yeah that magic f***ing wand? :) 08:17:04 well, if you're not talking about that 08:17:08 then what the fuck are you talking about? 08:17:23 what you just said 08:17:39 i'm simply saying it's not a given until i proposed a way to do it 08:17:47 ok... 08:18:09 if you had simply said "representation of code" from the beginning, this would not be an issue 08:18:18 lol 08:18:31 a graph is too general a thing with no inherent meaning without a context :P 08:18:41 so a graph in and of itself is meaningless to me 08:18:44 so the term "this accomplishes the apply phase" to you means something else somehow??? 08:18:47 but a representation of code is meaningFUL 08:19:07 i don't see a representation of code as accomplishing the apply phase 08:19:15 i see it as describing how the apply phase is accomplished 08:19:19 no, not the representation 08:19:26 the code itself 08:19:29 this reifies it 08:19:37 the code itseld and its representation are two different things! 08:19:43 sorry, i didn't distinguish properly (grammatically) 08:19:46 s/itseld/itself 08:19:53 well 08:20:12 that's a viewpoint that can be taken on dynamically in the system 08:20:23 llast time i checked, my processor doesn't execute graphs 08:20:41 but you should be able at times to treat the representation as the code 08:20:43 P 08:20:52 it doesn't execute Lisp either! :) 08:21:02 or C or assembly 08:22:47 yes 08:22:52 source code != code :) 08:23:15 but it can be *seen* that way 08:23:43 which is one of those tunes requirements that that be possible at all levels 08:25:10 hm 08:25:26 so are we now much closer to being "on the same page"? :) 08:25:49 maybe now you can explain multiple dispatch to coreyr ;) 08:26:24 eh? 08:26:35 he had a question about it 08:26:43 right in the middle of our discussion 08:26:53 but it seems he's afk or something now 08:28:45 [QUIT] JALH quit: bbiaf (min) 08:33:52 eihrul wonders if he is just insanely dense or English just sucks. 08:34:04 the latter i suppose 08:34:21 i would never deny it, at least 08:35:53 eihrul hmms. 08:36:01 okay, that makes 4 homework assignments i've forgotten already 08:36:05 this is probably bad 08:36:24 s/probably// 08:37:21 eihrul sighs. 08:43:39 hcf joined #tunes 08:43:50 hey hcf 08:43:59 hey 08:46:49 water: u aware of #lisp? 08:50:25 yep 08:50:35 dunno who's on it, usually 08:50:48 oic 08:54:37 abi: #lisp logs are at http://vengeance.et.tudelft.nl/~smoke/log/ 09:05:52 XeF4 joined #tunes 09:06:13 hey xef4 09:08:01 hey 09:33:36 [QUIT] water quit: The Tao went that-a-way! 09:39:12 [QUIT] coreyr quit: bbl 11:49:29 Kyle_L joined #tunes 11:49:36 [TOPIC] hcf: TUNES: Free Reflective Computing System http://www.tunes.org || Slate Programming Language http://www.tunes.org/~water/slate-home.html || http://lambda.weblogs.com || http://www.brics.dk/~pado2/ || http://www.aopsys.com 11:56:40 is the BNF on http://www.tunes.org/~water/slate-syntax.html current? 11:56:52 probly not 11:59:50 for any current info, u have to ask water 12:03:04 I wonder if some good could come from sitting down fresh and attempting to implement a (crude) reflective computing system in 72hrs 12:03:19 note: I am not suggesting others do that 12:09:21 [QUIT] JALH quit: Read error to JALH[host213-123-48-174.btinternet.com]: EOF from client 12:21:42 eihrul joined #tunes 12:26:31 Tril joined #tunes 12:26:31 [MODE] ChanServ set mode: +o Tril 12:28:35 Brianna joined #tunes 12:29:11 test 12:30:15 p0ng 12:31:36 Logs for this channel #TUNES are now in /pub/tunes/irc, old logs moved to /pub/tunes/irc/old 12:31:49 They are also duplicated in ~lar1/logs/tunes 12:31:53 abi: FOM is Foundations Of Mathematics at http://www.math.psu.edu/simpson/fom/ 12:32:09 restarting Brianna again 12:33:12 Brianna joined #tunes 12:33:19 ok 12:33:26 She'll start on bootup now too 12:33:30 abi: til is Transparent Intensional Logic at http://www.phil.muni.cz/fil/logika/til/ 12:36:17 hcf_ joined #tunes 12:36:20 [QUIT] hcf quit: Read error to hcf[207-172-225-9.s9.tnt1.pld.me.dialup.rcn.com]: Connection reset by peer 12:36:39 [NICK] hcf_ changed nick to: hcf 12:45:04 icac, newest factoids: 12:45:05 q-dag is Query DAGs: A Practical Paradigm for Implementing Belief-Network Inference at http://www.cs.cmu.edu/afs/cs/project/jair/pub/volume6/darwiche97a-html/jair-f.html 12:45:05 q is a VHLL at http://www.bothner.com/~per/software/index.html 12:45:09 analysis and compilation of OOLs is at http://www.csd.uu.se/~thomasl/wpo/oo-compilation-papers.html 12:45:09 xotcl is eXtended Object TCL at http://nestroy.wi-inf.uni-essen.de/xotcl/ 12:45:09 fom is Foundations Of Mathematics at http://www.math.psu.edu/simpson/fom/ 12:45:12 til is Transparent Intensional Logic at http://www.phil.muni.cz/fil/logika/til/ 12:45:12 pride and prejudice is Pride and Prejudice: Four Decades of Lisp at http://kmi.open.ac.uk/people/snw2/papers/prejudice/prejudice.html 12:48:30 abi: HANK is a cognitive modelling language/environment at http://kmi.open.ac.uk/projects/hank/ 12:58:40 water joined #tunes 12:59:01 hey all 12:59:10 hi 12:59:11 hi tril, kyle 12:59:31 how are things going for you? 12:59:32 water: how recent is the BNF on http://www.tunes.org/~water/slate-syntax.html? 12:59:38 not at all 12:59:49 XeF4 breathes a sigh of relief 12:59:59 heh 13:00:11 things are okay 13:00:36 i'm still doing language research 13:01:05 aha.. is the Uncommon Lisp Object System still available? 13:01:17 xef4: most likely not 13:01:44 i gave my computer to charity after i quit school 13:03:03 it wasn't very good anyway 13:04:14 [QUIT] ink quit: Read error to ink[adsl-63-204-135-0.dsl.snfc21.pacbell.net]: EOF from client 13:04:23 whoa 13:04:32 laurent martelli has a dot-com? 13:04:59 but not very useful 13:05:03 (wrong window) 13:09:40 so is anyone interested in writing up a new slate.lisp? 13:11:52 hm 13:12:13 i really need to write, but i also need to relax 13:25:55 [QUIT] water quit: The Tao went that-a-way! 13:54:42 [QUIT] abi quit: dying by hcf's request 13:55:19 abi joined #tunes 14:26:33 abi: poplog is also at http://www.poplog.org 14:26:33 okay, hcf. 14:35:55 [QUIT] radtke quit: Ping timeout for radtke[dlri9-213-33-45.bnu.zaz.com.br] 14:38:49 XeF4 considers appropriating the 'Uncommon Lisp Object System' name for his own project 14:40:18 abi: LogiMOO is an Extensible Multi-User Virtual World with Natural Language Control at http://www.cs.unt.edu/~tarau/research/99/lm.html 14:40:40 i like Computing System Designed On A Train better 14:41:04 the term Uncommon Lisp is already too krufty :) 14:51:30 [QUIT] JALH quit: g'night 15:55:14 yes. but it lends itself to good acronyms 15:55:31 U LOSE? sounds bad to me. 15:55:51 "ULOS - the way out of conventional OS design" and similar idiocy 16:05:32 eihrul joined #tunes 16:30:16 [QUIT] hcf quit: Leaving 16:31:05 [NICK] kc-store changed nick to: kc5tja 16:38:48 hi, does anyone have complete archives for mailing lists recently? I want to check if I am missing any posts. 16:38:58 ie. since august 25 16:39:54 err, I mean auguest 17 17:43:52 eihrul joined #tunes 17:48:40 arbit joined #tunes 18:12:28 hcf joined #tunes 18:32:06 [QUIT] arbit quit: Read error to arbit[209.6.184.68]: EOF from client 18:44:16 abi: elex is a multi-language scanner generator at http://www.ozemail.com.au/~mpp/elex/elex.html and elex for prolog is at http://odur.let.rug.nl/~vannoord/Elex/ 18:44:30 abi: fsa utils are Finite State Automata utilities at http://odur.let.rug.nl/~vannoord/fsa/fsa.html 20:13:37 jerome joined #tunes 20:21:41 hello 20:22:03 hello Tril 20:23:35 water joined #tunes 20:24:30 hi all 20:24:55 hi water 20:25:20 hi water 20:25:29 jerome: anything we can do for you? 20:25:54 ? 20:26:12 like answer questions 20:26:22 no thanx 20:27:01 hmm 20:27:25 i'm in a bit of a rut for anything but research 20:39:35 i hate being in a rut 20:40:48 is anyone in the mood to talk? 20:41:06 maybe 20:41:32 i take it you're still busy with bespin work 20:42:17 well, not really... I did what I can on the mailing lists 20:42:38 read my last mail to tunes? 20:42:56 just now, yes 20:43:25 do you have any slate mail that is missing from the archive? maybe there isn't any.. I'm not sure. 20:43:39 there hasn't been for some months now 20:45:52 all the archives for slate that I have are at : http://lists.tunes.org/cgi-bin/wilma/slate#browse and http://bespin.org/pipermail/slate/ 20:47:15 iirc that's all of them :/ 20:47:38 all right. 20:54:43 water` joined #tunes 20:55:28 [QUIT] water quit: Killed (NickServ (GHOST command used by water`)) 20:55:35 [NICK] water` changed nick to: water 20:55:35 [QUIT] Kyle_L quit: Leaving 20:57:23 what do you want to talk about? 20:57:44 anything about tunes, really 20:57:55 something substantial 20:58:25 i've been pouring over the hll req's and semantics pages in hardcopy form and writing notes on them for the last week 20:59:02 with very specific points and questions and noting each bullet that i don't have a concrete answer for yet 20:59:57 but i can't seem to go public again with this stuff 21:00:07 i feel like no one gives a damn 21:00:23 and i've been depressed somehow 21:00:56 "somehow" meaning i don't understand why, and can't seem to change the situation effectively 21:01:25 heh, people here obviously are affected by my words ;) 21:01:32 Well, sounds a lot like Fare, although I have an idea why he is depressed... (No progress in TUNES) 21:02:05 he's stuck on his own way of expressing things in tunes 21:02:24 i can't talk with him without him replying in his "phd way" 21:02:54 which no one really groks usually, and i only grok what i can determine is part of our common research background 21:03:43 speaking of which, if that guy actually explained where he got all the HLL terms from and provided direct links to research papers supporting those ideas, we'd be in a hell of a lot better shape 21:03:47 I think he needs feedback/help with his business plan.. 21:04:10 i think you know how i feel about his business plans 21:04:24 I forgot 21:04:26 business plan(s) 21:04:44 allegedly he's involved in two ventures now, or something 21:05:05 arbit joined #tunes 21:05:14 applying money to tunes in this way could be really bad 21:05:40 the pressure to "have something we can call tunes" might jeopardize the nature of what is produced 21:05:41 applying money to tunes in this way means he can hire people who will get things done :) 21:05:43 re arbit 21:06:14 IMHO as long as the money is provided "no strings attached" and they can work as long as necessary to get what they desire... then it's OK :) 21:06:18 honestly, i doubt it 21:06:41 no one but Fare really has a firm grasp on what his writings mean 21:07:00 that's unfortunate. 21:07:11 hola 21:07:29 funding ? 21:07:39 exactly, even a professor who does research into reflection could not make heads or tails of Fare's ideas 21:07:57 Tril: hmm, you mean we can now get paid to not produce anything tangible? 21:08:12 water: there's a reason i mentioned the Constitution :) 21:08:12 (witness the ASPIC 2000 conference reaction and those of earlier conferences to Fare) 21:08:13 arbit: Fare is trying to start a company to pay some people to work on TUNES 21:08:28 ASPIC? 21:08:37 eih: pliant 21:08:52 annual symposium on pliant i? c? 21:09:01 something like that 21:09:09 in french, i believe 21:09:12 abi: aspic2000 21:09:12 aspic2000 is at http://www.ehess.fr/centres/cams/person/pom/aspic.html or the first conference on Pliant 21:09:19 i didn't know fare presented at that? 21:09:41 he did 21:10:05 hcf: its down 21:10:17 great 21:10:34 what's down? 21:10:44 the aspic site 21:10:51 um no it isn't 21:11:33 there's not much on it, but it loads fine 21:11:37 works for me too 21:11:56 there he is :) 21:12:07 the page source code on the pliant looks like tcl? 21:12:09 who? 21:12:49 Tril: CMU has sucky routes :) 21:12:57 [QUIT] hcf quit: Leaving 21:13:04 hcf joined #tunes 21:13:09 is Fare in that picture? 21:13:29 icac, www.tunes.org/~fare/tmp/Pliant/aspic2000report.html 21:15:23 Yeah, I'm pretty sure the guy in the front on our left is Fare. 21:15:35 compare to http://fare.tunes.org/pictures/fare1.jpg 21:15:51 wow 21:17:26 we sort of look like each other 21:17:51 anyway 21:17:58 separated @ birth! 21:18:07 don't even go there 21:18:48 bbiam... talk amongst yourselves... mingle even ;) 21:19:46 [NICK] morton changed nick to: com1 21:24:29 [NICK] nate37 changed nick to: morton 21:24:35 [NICK] morton changed nick to: nate37 21:25:35 nate37: you are morton? are you the one I banned from this channel for making fun of hcf? and I just gave you a bespin account because why? :-) 21:25:36 arbit bbiab: food 21:26:19 Tril : heh no, he was saying that he couldn't change his name back to morton, so i showed him i could 21:26:50 you just chanegd your name to morton right now. who said that? 21:27:04 com5 did 21:27:10 (the real morton) 21:27:21 oh.. in #osdev? 21:27:24 yep 21:27:35 hm 21:27:42 you can check the logs if u don't belive me 21:28:18 [NICK] ink|doing_stuff changed nick to: ink 21:28:25 later, you 1337 mp3 dudes 21:28:51 gnite 21:29:36 water repeats to himself: "you must start writing up new docs" "you MUST start writing up new docs" 21:29:41 P 21:31:16 unfortunately, there still are important things i haven't been able to decide upon 21:31:21 like slate syntax 21:31:33 which bugs the hell out of me sometimes 21:33:37 you know, i remember saying that if slate doesn't work in a year that i would drop the whole thing 21:33:51 (i.e. by january 2001) 21:34:07 i think i still agree with that 21:39:42 hcf: how's diktuon2 coming? 21:40:21 dunno 21:40:27 water: do you know diktuon1 is up ? got an account? 21:40:34 yes i do 21:40:46 coreyr made you one, then 21:40:58 yep 21:41:27 cool 21:42:33 i wish i could figure out either how to make slate's syntax natural to understand or what syntax slate would need to be understood 21:42:50 semantics are all designed? 21:43:03 semantics get more clear by the day 21:43:23 i'm outlining the mop currently 21:43:39 but the code model bugs me like crazy 21:44:05 rewrite rules are cool, it's the slate syntax itself that just seems wierd 21:44:13 brb 21:45:01 is there a prototype of the syntax one could peruse? 21:47:55 i doubt it 21:51:40 or maybe just a brief explanation 21:51:56 my car was just on fire.. im waiting for it to cool before i head to the store 21:57:00 ??? 21:57:23 "on fire"? 21:57:26 oh well 21:57:28 only slightly.. 21:57:31 as for the syntax.... 21:57:32 water: arbit wants syntax for slate? 21:57:33 it's better now 21:58:12 slate expressions are a series of names to travel around namespaces 21:58:40 this is because a slate environment is a web of objects 21:59:04 and the "parts" of the object have names, and are just links to other objects 21:59:39 now, the way you get objects to *do* things is by leaving handles for application 21:59:50 an example.... 22:00:15 for the OO-style, i might want a point object with two state components "x" and "y" 22:00:40 so i would have a "myObject" link from the environment to the object 22:01:05 and "x" and "y" would link from that object to the ones providing a kind of state 22:01:17 hm this is not going well 22:01:36 abi slate tutorial? 22:01:36 slate tutorial is at http://www.tunes.org/~water/slate-tutorial.html 22:01:41 im following.. 22:02:19 the idea is that slate expressions are independent of any notion of whitespace other than the "." usage (for application) 22:03:08 so it's the basis of the evaluator's interactivity that you can partition any slate expression into arbitrary parts and still be able to inspect things 22:04:31 the tutorial explains a little more, but it also needs much work 22:05:06 i'm reading it... 22:05:07 for example, to get x's value, you don't just type in "x", you type in "x^.." 22:05:14 k 22:05:49 ^=result 22:09:49 that tutorial is pretty outdated now 22:10:10 i see.. very interesting 22:10:48 anyway, the apply phase takes the environment object graph and replaces a part of it with a new part 22:11:16 that's how the meta-object determines how an object behaves 22:12:39 to clarify, there is one meta-object per object in slate that describes how that object works and even implements it 22:13:19 and slate does not have classes as a built-in concept, but there is a parent slot which accomplishes the inheritance part of classes 22:14:37 hey are slots objects? 22:15:00 they are links to other objects, yes 22:15:15 iow objects' slots are other objects 22:15:44 one tough nut to crack is a tunes idea 22:16:38 that all of the kinds of literal objects (numbers, characters, strings, logic) are all regular, user-level objects and that the system can have any of them be primitive and others derived automatically 22:17:02 for example... 22:17:37 that point object with x and y in it conceptually depends on numbers being supported before tuples of them 22:18:10 but i could also make a bunch of tuple objects and assign the whole object to them 22:18:49 but expressing the idea that the tuples parts have to "act like numbers" is not easy to do 22:19:29 in normal languages, it's hardwired, and you have to manually change it 22:19:38 blah 22:19:53 i'm going to go relax for a bit 22:19:59 im trying to grok.. sounds very interesting 22:20:10 ok 22:20:39 the slate pages have a references page with a couple of url's to relevant papers in postscript format 22:20:57 that describe the lookup-apply idea 22:21:16 i'll check 'em 22:21:39 yeah ./slate-references.html 22:22:05 i need to add a bunch of papers to that list 22:23:40 my car's cooled down. lets see if itll drive 22:24:07 arbit bb 22:24:21 [QUIT] hcf quit: Leaving 22:42:53 [QUIT] Tril quit: bye 22:44:55 [QUIT] water quit: The Tao went that-a-way! 23:26:41 [QUIT] _ruiner_ quit: destroy what destroys you 23:57:15 nate37 left #tunes 00:15:20 nate37 joined #tunes