IRC log started Wed Feb 9 00:00:04 2000 [msg(TUNES)] permlog 2000.0209 -:- SignOff air: #TUNES (http://www.qzx.com/ :: sleep) -:- SignOff Plundis: #TUNES (Ping timeout for Plundis[130.238.23.252]) -:- Plundis [plundis@130.238.23.252] has joined #tunes -:- smokie [tw026024@zaalf06.twi.tudelft.nl] has joined #tunes -:- SignOff smokie: #TUNES (zoef) -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- NetSplit: varley.openprojects.net split from forward.openprojects.net [02:23am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net] -:- Netjoined: varley.openprojects.net forward.openprojects.net -:- abi [nef@bespin.dhs.org] has joined #tunes -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-20.vpn.uib.no]) -:- Netjoined: sterling.openprojects.net clarke.openprojects.net -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- smoke_ [smoke@15dyn71.delft.casema.net] has joined #tunes -:- SignOff smoke_: #TUNES (Ping timeout for smoke_[15dyn71.delft.casema.net]) -:- NetSplit: varley.openprojects.net split from irc.linux.com [06:09am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net] -:- Netjoined: varley.openprojects.net irc.linux.com -:- abi [nef@bespin.dhs.org] has joined #tunes -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- NetSplit: fontana.openprojects.net split from tolkien.openprojects.net [06:14am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [fontana.openprojects.net] -:- Netjoined: fontana.openprojects.net tolkien.openprojects.net -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- NetSplit: irc.linux.com split from devlin.openprojects.net [06:20am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [irc.linux.com] -:- Netjoined: irc.linux.com devlin.openprojects.net -:- abi [nef@bespin.dhs.org] has joined #tunes -:- Plundis [plundis@130.238.23.252] has joined #tunes -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- Fare [fare@s195.paris-119.cybercable.fr] has joined #tunes -:- smoke is now known as avoozl -:- NetSplit: wang.openprojects.net split from irc.linux.com [06:59am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [wang.openprojects.net] -:- Netjoined: wang.openprojects.net irc.linux.com -:- Plundis [plundis@130.238.23.252] has joined #tunes -:- SignOff Fare: #TUNES (:Connection reset by pear) -:- avoozl is now known as smoke -:- rares [rares@wtrb-sh4-port141.snet.net] has joined #tunes -:- SignOff rares: #TUNES (KVirc 1.0.0 Millennium BETA 3) -:- hcf [nef@me-portland-us131.javanet.com] has joined #tunes -:- Fufie [stig@tunnel-44-28.vpn.uib.no] has joined #tunes -:- SignOff smoke: #TUNES (gdb zombie) -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- thomas [thomas@193.217.63.152] has joined #tunes hi fufie? i think fufie is in favour benevolent dictatorship or (too) used to imperative languages yes? hi 11:10am -:- SignOff smoke: #TUNES (Ping timeout for smoke[15dyn71.delft.casema.net]) -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- NetSplit: fontana.openprojects.net split from adams.openprojects.net [11:51am] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [fontana.openprojects.net] -:- Netjoined: fontana.openprojects.net adams.openprojects.net -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- SignOff smoke: #TUNES (One day sheep will rule the world) -:- smoke [smoke@15dyn71.delft.casema.net] has joined #tunes -:- eihrul [lee@usr5-ppp139.lvdi.net] has joined #tunes -:- rares [rares@wtrb-sh8-port215.snet.net] has joined #tunes -:- SignOff hcf: #TUNES (Leaving) -:- smoke is now known as avoozl -:- NetSplit: varley.openprojects.net split from devlin.openprojects.net [01:49pm] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net] -:- Netjoined: varley.openprojects.net devlin.openprojects.net -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- abi [nef@bespin.dhs.org] has joined #tunes -:- Plundis [plundis@130.238.23.252] has joined #tunes -:- Fufie [stig@tunnel-44-28.vpn.uib.no] has joined #tunes -:- thomas [thomas@193.217.63.152] has joined #tunes -:- avoozl [smoke@15dyn71.delft.casema.net] has joined #tunes -:- eihrul [lee@usr5-ppp139.lvdi.net] has joined #tunes -:- NetSplit: irc.linux.com split from adams.openprojects.net [01:52pm] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [irc.linux.com] -:- Netjoined: irc.linux.com adams.openprojects.net -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- Fufie [stig@tunnel-44-28.vpn.uib.no] has joined #tunes -:- eihrul [lee@usr5-ppp139.lvdi.net] has joined #tunes -:- NetSplit: varley.openprojects.net split from adams.openprojects.net [01:54pm] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net] -:- Netjoined: varley.openprojects.net adams.openprojects.net -:- abi [nef@bespin.dhs.org] has joined #tunes -:- Fare [rideaufr@esmeralda.enst.fr] has joined #tunes -:- avoozl is now known as smoke` -:- SignOff abi: #TUNES (Ping timeout for abi[bespin.dhs.org]) -:- abi [nef@bespin.dhs.org] has joined #tunes -:- SignOff rares: #TUNES (KVirc 1.0.0 Millennium BETA 3) -:- water [water@tnt-9-127.tscnet.net] has joined #tunes anyone here? yes cool what's up? not much just yet just got home in fact i haven't resolved the syntax problem completely yet :( why aren't you at school, btw? school is out oh yeah different time zone well have you had any ideas about how to handle slate at all? 03:00pm * eihrul/#tunes ponders why applyTo isn't handled by the meta-object... hm whereas lookupIn is handled by the meta-object applyTo is interpreted by the looked up object itself... which has me thinking the extrema of both manners... it could be done a different way, i suppose either they're both interpreted by the object itself or they're both interpreted by the meta-object -:- SignOff smoke`: #TUNES (Ping timeout for smoke`[15dyn71.delft.casema.net]) though #1 in some way is strangely appealing... * water/#tunes nods well, what you're suggesting happens to just be a refactorization of the "implementation" of lookup and apply phases in those papers * eihrul/#tunes nods. they both *seem* equivalent though #1 also *seems* problematic :) well then we should pick the one best suited to our concepts hm * eihrul/#tunes ponders how one would send the 'lookupIn' message to an object... well, #1's can be had from #2 by factoring out a unique bmo for that object -:- smoke` [smoke@15dyn71.delft.casema.net] has joined #tunes it almost seems as if the prototyping-versus-classing argument :) remanifest hm though, it would seem odd to send an 'object' the message 'lookupIn' yeah, bmo's seem like classes in this respect (because that name would also be used for the mop) not necessarily er... those slots would have to act Toppish... hm ok er, perhaps not... it's just they could never be explicitly accessed :) 03:10pm heh that'd be quite a problem the fact that 'lookupIn' and 'applyTo' have names as any normal selector becomes a hindrance if an object describes it's own meta-behavior... well, they'd be implicitly accessable... (in that they get invoked by the default mop) * water/#tunes nods that brings up the issue of naming though, one would have the same problems if they linked an object's meta to itself... (which is what this perhaps generalizes to) (and perhaps the same problem exists with named slots with implicit behavior in slate as-is now) atleast, clashing i think that i still prefer bmo's even if they are class-like implicit behavior? as in? behavior that is specified by the default... not by the user e.g.? the 'meta' and 'parent' slots oh or data slots, i suppose, too though, it's not quite as evil as self-mo'ing objects which Top is * eihrul/#tunes also ponders that you could perhaps generalize mbo from the other... hm? in that the default lookupIn and applyTo slots could contain methods that implement the 'meta' slot and what-not sure * eihrul/#tunes is just trying to linked the two together in some manner... s/linked/link though, 'lookupIn' and 'applyTo' would have to have somewhat unique names if objects describe themselves to avoid problems hm i.e. _lookupIn, _applyTo... borrowing from Self... hold on a minute, brb though... perhaps this ties in a 'map' with a 'meta-object' too... since that's where these things would most likely end up residing (because they rarely change) whereas right now, a 'map' is another particular meta-object that remains separate... hm 03:20pm the leading underscores thing, though, is for primitives * eihrul/#tunes nods. these however could be overridden the 'default' mop would be implemented in those two primitive methods... (but if someone wanted an object with unique meta-behavior, they could override those particular slots) and override the default mop... even though, the underscores more serve to differentiate it it from the regular 'lookupIn' and 'applyTo' slots... so that the user could freely define those without fear of clashing... and those slots perhaps could not be subject to themselves... well, the simplest thing would be to just leave them in the bmo (though, i'd have to think on it more to see if that's true :>) well, in the worst case, you'd use them merely to implement the mop... (just an idea) honestly, i haven't been thinking in this direction at all mostly considering the consequences of our particular 'object-function' union for syntax and semantics btw, 'applyTo' might be considered to be implemented by an object which "contains" both the obj and bmo and is dynamically generated by the bmo 03:30pm though i think just moving applyTo into the meta-object is less obfuscated... but how would the syntax of the mop look then? you mean the rewrite of a message-send? yes hold, lemme write it out ((object meta) (apply_To ((object meta) lookup_In selector object) object)) hm ok and you can even create the illusion of the other by just supplying a default apply_To that implemented the other version... not sure i follow well, you'd make an apply_To method that just takes the 'apply:' argument, sends the 'applyTo:' message to it with the argument 'To:' which duplicates the functionality of the other... 03:40pm k hm your syntax actually requires arguments to be bound to the selector before the message is sent which i happen to like it doesn't even conflict with partial-eval rmm... that was a mys-type :) ((object meta) (lookup_In selector object)) it was a good one :) the rest was no more than you'd already specified... well i'm not sure i like the "arglist" idea anyway well, that was in no way meant to be the slate syntax... unless there's a syntactically useful way to specify binding some but not all of the arguments because syntax is in flux right now... though... that does make me wonder... given: ((object meta) lookup_In selector object) should lookup_In, selector, and object be independent messages sent to the receiver... or should they be a composition of messages? :) uhh are you serious? depends what you mean by serious... that's a big step syntactically-speaking more like RPN than oop well... unary messages already behave as a composition of sorts in smalltalk... yes true though they associate in the other direction but smalltalk "knows" how many arguments a message has by the selector name i.e. counting colons * eihrul/#tunes nods... well, unless specified by a ()... you assume unary hm? i see (object unary) as (object (unary)) hm that takes paretheses beyond just grouping, though 03:50pm well, what do parens do as of now? :) just grouping, afaik -:- smoke` is now known as avoozl well, how does that extend beyond grouping? hm well, as you suggested, it tells how many args a selector has not as i see it the explain s/the/then i see it as merely cascading the specified messages to the object (and in no way interpreting it as an arity indicator) e.g.? e.g. (object foo bar baz) cascades the messages foo, bar, and baz to object in no specific order why no specific order? well, so far... though an ordering would probably be needed somewhere yes it would especially if you have side-effects... well if you use msg's to do dataflow (functional-style) then that's also true as in (mySwitch turnOn turnOff) yes, side effects :) not necessarily well... perhaps not side effect so much as state and state=value though, where i was questioning whether to make it composition... is that should (mySwitch turnOn turnOff) be interpreted as: ((mySwitch turnOn) turnOff) or: (mySwitch (turnOn) (turnOff)) ordering is implicit in the first one... er rather, explicit :) this type of example is trivial, though, without arguments to the selectors, that is 04:00pm (mySwitch (turn toWhat)) >:) so how do you bind the arguments syntactically? that's the question... how? how do you see an interpreter interpreting that statement? well, since selectors are objects, too, we could just send them accessor messages (in terms of lookup + apply) -:- SignOff Fare: #TUNES (Ping timeout for Fare[esmeralda.enst.fr]) i see it as bind+lookup+apply yes, but for this particular example... how would it be reduced by the interpreter :) where (type-wise) selector+bind->selector which example statement? (mySwitch (turn toWhatState)) which happens to be unbound hm this also aids in the specification of a slate interpreter... i know well, explain what you mean by "reduced" what steps would the interpreter take (in terms of lookup + apply) to evaluate that statement oh great i mis-parsed your example :( well, i see a problem with the way it is the selector is implicitly sent a message to a set a slot of unspecified name though, i didn't necessarily need it in my mental interpreter... trying to remember how i didn't though :) 04:10pm well, this is a stretch, but argument binding could be acheived by specifying pre-emption of the message-send with a re-written selector using the accessor to change the selector's state yes, that's a stretch but not if you do it abstractly, perhaps i'm trying to avoid resorting to special syntax * eihrul/#tunes is trying to reduce that particular mySwitch example... i don't mind complicated syntax, as long as it's very uniform and doesn't break any concepts i guess i mean complicated expressions, given simple syntax ugh... i hate imperativish code :) as in? as in the evaluation steps i'm transcribing... oh 04:20pm * eihrul/#tunes ponders... well, it doesn't require any implicit slots... what doesn't? my evaluator so describe it it's probably crude, confused, and what-not... but it's what i came up with :) and is written in a pidginish smalltalk... darn it, no CRs i'm still in win32, you know :) just newlines... notepad will interpret them somewhat correctly er, wordpad notepad was the thing that couldn't interpret them, doh oh ok though... i'm not sure about the actually apply part and how it would actually extend to multiple messages -:- _ruiner_ [DIY@ppp188.wi.centurytel.net] has joined #tunes -:- SignOff _ruiner_: #TUNES (destroy what destroys you) what are ' and ^ ? given a selector foo, foo' denotes the valued retrieved by lookupIn of foo, and foo^ denotes the result of applying a result to foo' (the reply) though... the last part is probably hoky and needs to be revised :) 04:30pm er hm * eihrul/#tunes is evaluating an expression of many messages... * Fufie/#tunes expresses longing for his bed (push Fufie bed) (say "Goodnight") ugh... destructive macros! you fiend not sure i grok this not sure i do yet either * eihrul/#tunes is trying to wire down the semantics... oh good, so i'm not alone ;) yes, the apply part breaks down with multiple messages, as expected this blurb is a stack trace, right? in a manner of speaking, yes though it's using a wetware stack... right * eihrul/#tunes ponders how many mips he rates... huh? bogomips? yes or appropriate benchmarking measure of choice... 700 on mine * eihrul/#tunes guesses he is approximately 6.02 * 10 ** (-23) bogomips. oh you mean your brain ;) yes brains are for geometry and logic, not arithmetic yes, that's why they rate so poorly (wrt evaluation) nah 04:40pm hm bbl k in about 2.5hours -:- water [water@tnt-9-127.tscnet.net] has left #tunes [] -:- avoozl is now known as smoke` * eihrul/#tunes ponders... hmmm it actually does work nicely if you consider it as composition... * eihrul/#tunes cackles. -:- SignOff smoke`: #TUNES (z!) 04:50pm -:- washort [washort@d119.narrowgate.net] has joined #tunes -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-28.vpn.uib.no]) -:- NetSplit: wang.openprojects.net split from merril.openprojects.net [05:41pm] -:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [wang.openprojects.net] -:- Netjoined: wang.openprojects.net merril.openprojects.net -:- Plundis [plundis@130.238.23.252] has joined #tunes -:- air [brand@p0wer.qzx.com] has joined #tunes -:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp139.lvdi.net]) -:- eihrul [lee@usr5-ppp169.lvdi.net] has joined #tunes hmmm -:- SignOff washort: #TUNES (changing servers) 06:20pm -:- washort [washort@d119.narrowgate.net] has joined #tunes -:- Kaufmann [newbie@dial417.infolink.com.br] has joined #tunes yo 07:10pm yo what's up? working on a slate-ish interpreter -:- rares [rares@wtrb-sh3-port37.snet.net] has joined #tunes cool you decided on sexp for the syntax, right? yes message-passing semantics? i'd guess so :) * eihrul/#tunes is still trying to understand the code *he-wrote.* mops are a mind-fuck, that's for sure :) that's why we like them -:- hcf [nef@me-portland-us732.javanet.com] has joined #tunes WTF is mops? meta-object protocol abi: mops? mops are Meta-object Protocols, see http://www.parc.xerox.com/spl/projects/mops/ okay... pray the very first method call in this interpreter works... to c'thulhu, fare, or whatever gods you worship... 07:20pm aaaaaah yes... bugs duh i should've known I thought you were talking about like brooms and stuff forgive me * Kaufmann/#tunes is a dimwit * eihrul/#tunes cackles. silly eihrul, how many LOCs? Lines of Conversation? 99 mops the ultimate People's Front of Judea tribute -:- ult [noone@user-37kba4t.dialup.mindspring.com] has joined #tunes hey ult hrrrm including a parser? kaufmann: it just interprets lists right now... :) the advantages of sexp... ah yes so what, mostly a big cond form? 07:30pm nope.... not even close :) cond forms RULE! but eihrul has if()then(if()then()) so on and so forth don't need cond when you have pretty much uniform language semantics... :) Hey eihrul, wot damtypes are the darned nyloc? ;) uh, necromantic, slashing, chaos, most prominently... which one is the most common one though? I'm gonna get a rune. slashing... eihrul, well, lemme see the code then * ult/#tunes nods wait - it's 99 LOCs, but what's the CPL average? brb er 99 LOC/CPL? lines of code/characters per line? yeah whoa... I just realised I'm going insane. Cool! I'm quite insane. Lots of fun. a few weeks ago they installed a security camera on our building's elevators, which have mirrors on them. Now I can't even look at myself in a mirror without getting the feeling that someone is watching me Maybe someone is... gee thanks eihrul, well? heh. Maybe Eihrul finally summoned Hastur! 07:40pm perhaps i have the migraine... weird tho that u got one too... eihrul, what would the factorial definition look like in this language? hah don't ask :) why not? because i don't know! I'm the kind of guy who judges a programming language by the elegance of the factorial definition well K u my man when we have one... we'll let you know eihrul, well, then how do you expect your interpreter to work? "Gee, I'm writing an interpreter for this great new language... I don't know what it's like, but it's supposed to be really cool..." :) that's every fsckin oss project on day one :) sheesh -:- overfien [overfien@134.39.254.213] has joined #tunes yoh whats up in my own little HLL, it's: factorial(1) <= 1 factorial(x | x :: |N) = x * factorial(x - 1) * rares/#tunes is stalking a bid he made on an amiga 1000 overfien, dunno urgh in my own little HLL, it's: factorial(1) <= 1 factorial(x | x :: |N) <= x * factorial(x - 1) Kaufmann: dude... the language is largely unspecified at this point :) i'm just writing a prototype interpreter... eihrul, no wonder it takes up only 99 LOCs :) -:- SignOff ult: #TUNES (Ping timeout for ult[user-37kba4t.dialup.mindspring.com]) -:- ult [noone@user-38lcmoc.dialup.mindspring.com] has joined #tunes HASTUR HASTUR HASTUR HASTUR! the trick is to take the proto interp and build the real thing from it I'm working on a functional/logic reflective HLL, based on rewriting logic but featuring optimising compilation, coupled with an intelligent runtime system 07:50pm and just how long is it going to take you to implement all that? :) umm and a second ago you were groaning about spy mirrors I see LOL rares eihrul, hopefully not more than a few months Fufie has already agreed to help me do work on the prototype for this language's type system but implemented on Scheme you're going to build an optimizing compiler in a few months? :P eihrul, well, everything but the optimising compiler. :) I've got to shut down now. Buh-bye hah... my interpreting is meta-regressing ad-infinitum! -:- SignOff Kaufmann: #TUNES (kathyanne gets kick'd/ as leaves fall on still waters/ she is here no more.) 08:00pm * eihrul/#tunes cackles... -:- water [water@tnt-10-50.tscnet.net] has joined #tunes hey all hey, water... :) hey * eihrul/#tunes has a little bit of mop'ing going on... i just got a really good book on new stuff in rewrite logic eih: cool water: i used _lookup and _apply, because that was easiest for now :) now i'm trying to replace most of the bootstrap code with primitive method objects ok see that'swhat makes me believe in this project if booting is possible, anything is what is that? what is what? what makes you believe...? holy cow, squeak's gonna rock by the end of this year well I keep getting confused at the edge of the singularity so to speak where you have to get the thing up and running but in such a way that your boot system doesn't muck up the elegance of the rest of the system * eihrul/#tunes cackles. water: what's happening with it? disney's gonna throw in a lot of code that they've been working on in secret for multimedia and enhancing the morphic ui apparently soon, too i had to use 'in... sadly 'in? er 'in' oh why? oh wait i know just for the mop... 08:20pm until we have Top and the primitives for modifying it, we can't mark slots for input the author of this rewrite book definitely explores the regions of declarative programming beyond just logic xor functional T-45 huh? my Amiga -:- SignOff overfien: #TUNES (Ping timeout for overfien[134.39.254.213]) 08:30pm anyway, hopefully whatever language features we put into slate will likely benefit from whatever morphic gets upgraded into no wonder no one in high school learns LISP it would scare the fuck out of Disney heh dude i can't imagine anyone learning lisp in highschool the way of teaching lisp completely would miss your average teenager dude they should learn it in kindergarten while they're playing with legos well not everyone could pick up logo, iirc i doubt 6yr olds have the concepts matured required to deal with higher-order fp you'd be surprised (and i've taken courses on developmental psych) they're more resourceful than their parents resourceful does not equate to reasoning power well in a nationwide test when givew the following puzzle only the kids could solve it: yes? :) you lie on your back an have to pick up an object on your left side with your right hand without turning your body clockwise nor counterclockwise the kids got it the parents couldn't 08:40pm what does this have to do with higher-order fp? it's a matter of seeing a set of rules within a set of rules and a 6 year old can deal with the idea of an box in an identical box ok i can see that but i don't think it equates, just the same in fact I think the capability gets worse for every ten years you add :) no we just bury it under convention adults still understand games and pretending, don't they? :) there should be a Barney program for parents lol adults? games? pretending? Quake? Doom? Lawsuits? Media? Hysteria? Christmas shopping? who k nows hmmmm rares? i heard rares was addicted to politics