IRC log started Wed Apr 5 00:00:01 2000 [msg(TUNES)] permlog 2000.0405 -:- SignOff hcf: #TUNES (Leaving) -:- smoke [smoke@15dyn123.delft.casema.net] has joined #tunes -:- SignOff nate37: #TUNES (nate37 has no reason) -:- SignOff ult: #TUNES (Ping timeout for ult[user-38lcn4c.dialup.mindspring.com]) -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-53.vpn.uib.no]) -:- Fufie [stig@tunnel-44-53.vpn.uib.no] has joined #tunes -:- dalvarez [ircusr@212.68.72.98] has left #tunes [] -:- nate37 [nate37@ppp-207-151-70-215.dialup.pcmagic.net] has joined #tunes -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-53.vpn.uib.no]) -:- Fufie [stig@tunnel-44-53.vpn.uib.no] has joined #tunes -:- smoke_ [smoke@vengeance.et.tudelft.nl] has joined #tunes -:- SignOff smoke_: #TUNES (One day sheep will rule the world) -:- nate37 [nate37@ppp-207-151-70-215.dialup.pcmagic.net] has left #tunes [] -:- smoke_ [smoke@vengeance.et.tudelft.nl] has joined #tunes -:- SignOff smoke_: #TUNES (brb) -:- smoke_ [smoke@vengeance.et.tudelft.nl] has joined #tunes -:- hailtotheking [Anonymous@203.115.25.3] has joined #tunes <hailtotheking> what is TUNES? <smoke_> hailtotheking: an OS to be <hailtotheking> i just went to the website <hailtotheking> thanks -:- hailtotheking [Anonymous@203.115.25.3] has left #tunes [] 06:10am -:- SignOff smoke_: #TUNES (woosh) -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-53.vpn.uib.no]) -:- XeF4 [xef4@194.255.106.40] has joined #tunes <XeF4> !seen tcn <XeF4> Does anyone know if C is forbidden from the mainstream retro tree? 08:40am -:- AlonzoTG [Alonzo@207-172-49-95.s95.tnt7.lnhva.md.dialup.rcn.com] has joined #tunes <smoke> how do i write '(n over k)' in TeX ? 09:50am -:- SignOff smoke: #TUNES (Reconnecting) -:- smoke [smoke@15dyn123.delft.casema.net] has joined #tunes -:- SignOff thomas: #TUNES (Ping timeout for thomas[193.217.63.152]) -:- thomas [thomas@193.217.63.152] has joined #tunes -:- SignOff XeF4: #TUNES (installing a new hdd) -:- SignOff AlonzoTG: #TUNES (Ping timeout for AlonzoTG[207-172-49-95.s95.tnt7.lnhva.md.dialup.rcn.com]) -:- AlonzoTG [Alonzo@209-122-249-59.s567.tnt8.lnhva.md.dialup.rcn.com] has joined #tunes -:- Fufie [stig@tunnel-44-53.vpn.uib.no] has joined #tunes -:- SignOff AlonzoTG: #TUNES (Ping timeout for AlonzoTG[209-122-249-59.s567.tnt8.lnhva.md.dialup.rcn.com]) -:- eihrul [lee@usr5-ppp16.lvdi.net] has joined #tunes -:- XeF4 [xef4@supernova.sevenheavens.com] has joined #tunes -:- SignOff XeF4: #TUNES (Leaving) -:- tcn [Tom@cci-209150250075.clarityconnect.net] has joined #tunes -:- water [water@tnt-10-124.tscnet.net] has joined #tunes <water> hey all <tcn> hey -:- SignOff tcn: #TUNES (ircII EPIC4-2000 -- Accept no limitations) -:- SignOff smoke: #TUNES (Read error to smoke[15dyn123.delft.casema.net]: EOF from client) <eihrul> hey -:- SignOff eihrul: #TUNES ([x]chat) <water> hm why'd tcn leave? <water> how odd -:- tcn [Tom@cci-209150250047.clarityconnect.net] has joined #tunes <water> re <tcn> yeah, my modem died <water> k * water/#tunes catches up on email <tcn> 20 msgs, 99% crap 02:50pm <tcn> ever use "wmx"? <water> what's wmx? <tcn> nice little window manager <water> nope <tcn> nice kbd controls, sorta like the linux console. Don't have to touch my mouse :) <water> well that's a nice idea, since it makes the code interface-independent <water> (theoretically) <tcn> what do you use, gnome? <water> mostly kde <tcn> oh yeah, I used to use kde <tcn> all the bugs worked out? <water> (mostly for klyx etc) <water> yeah it's pretty stable * water/#tunes contemplates building a slate sim out of the new work on prototype objects in squeak 03:00pm -:- ult [ult@user-37kbass.dialup.mindspring.com] has joined #Tunes <water> ok time to write up some more slate stuff for the mlist * ult/#Tunes bumps water <tcn> see ya -:- SignOff tcn: #TUNES (ircII EPIC4-2000 -- Accept no limitations) <water> ult: eh? * ult/#Tunes shrugs 03:20pm -:- SignOff ult: #TUNES (Read error to ult[user-37kbass.dialup.mindspring.com]: Connection reset by peer) -:- water has changed the topic on channel #tunes to: Tunes - Free Reflective Computing System - http://www.tunes.org/ || Slate - http://www.tunes.org/~water/slate-home.html -:- hcf [nef@me-portland-us700.javanet.com] has joined #tunes <water> hey hcf <hcf> hey <water> no responses to my emails recently, but massimo still seems intelligent <water> what did you think of his response to atg? <hcf> i'l reread <hcf> (lag) <water> hm more drivel about databases <water> they respond too quickly to each other for such a topic. they don't seem to think things out 04:20pm <water> no comment? <hcf> i guess not 04:30pm -:- menozero [gpensa@mi1-1043.dialup.tiscalinet.it] has joined #tunes <water> hello meno0 <menozero> hi. <water> is there something we can do for you? <water> for instance, do you have questions about tunes? <water> hm <menozero> well, I've been lurking at Tunes since 1994. <water> likewise, actually <menozero> I was even on the ml. <water> well i didn't join until a year ago <menozero> My project was about an editor. Still working on it. <water> you use dylan, then? <menozero> but I'm not active. never been. lonely guy. <menozero> *YES* I love Dylan. <water> too bad <water> ok <water> well i have questions about dylan <menozero> right now I'm working with Dylan on my XML editor. <water> you see, i'm working on an HLL prototype called Slate and i was curious about dylan's macro system, since both dylan and slate are oo-friendly <water> but the dylan docs don't seem to explain them well enough for me to understand what they have to offer over lisp macros <menozero> Dylan doesn't offer much over Lisp. <menozero> but it's a nice macro system for an infix lisp. <water> dylan's oo-based, though <water> right or wrong? <menozero> yes. very OO. <water> so how do its macros work, essentially? <menozero> macros and OO are not very tied. <menozero> I don't understand... <water> well, i understand lisp's macros, but they only apply to function-like ideas or multi-methods it seems <menozero> the difference between Dylan (and some lisp) macros and C/M4/TeX/... macros is [you know]. <water> it seems you have to be really careful (moreso than otherwiase) when making a macro that deals with pure oo <menozero> Dylan use multimethods. like CLOS. 04:40pm <water> not proper methods? <menozero> I don't think I've seen much literature on macros. <water> no there isn't much besides the lisp stuff <menozero> I don't understand your problem. an example would be nice. <water> hm <water> well i have web pages on slate that you could look at <menozero> macros just expand some stuff. you say "foo" and it's like "bar1; bar(foo); bar2" <menozero> if carefully designed (wich is hard) they are not dangerous for OO prog. I think. <water> yes but that's limited to some implicit notion of lisp <menozero> slate.url <water> see the topic <water> abi: slate <abi> i heard slate was a unifying/hybridizing of self/beta/lisp at http://www.tunes.org/~water/slate-home.html <menozero> I've seen macros used to define some complex data type together with its methods. <water> well macros depend on a single namespace as far as i've noticed <menozero> for example "define command delete-line "C-k" "Delete the current line." end;". <menozero> macros use the same ns as other symbols. <water> yeah well in slate my objects are all namespaces <menozero> module ex/import just names of things, mostly vars, funs, class and macros. <water> well modules in slate are redundant <menozero> does slate have an implementation? 04:50pm <water> there's an evaluator written in lisp that covers a small subset of it <water> but the language and environment specification aren't worked out completely yet <menozero> (reading about slate) <menozero> I fell in love (!) with Dylan for some reasons. <menozero> Dylan was made to be a practical pl. <water> as opposed to... ? <menozero> lisp <water> que tal smalltalk? <menozero> Lisp is almost perfect, Smalltalk, Self... but when it comes to compiling... <water> uh sure <water> they take much less time to do so? :) <menozero> while Dylan implementations are not so nice, they could produce almost C++-like libraries using the right philosophy. <menozero> which is lisp. <water> anyway <menozero> yes, I know and I don't want to flame. <menozero> really. <menozero> I've tried, but never implemented, to create *my* programming language. <menozero> it was based on a concept/relation encyclopaedia. <water> :P <menozero> functions, variables, classes and everything was based on c/r. <menozero> which is something similar to what tunes was trying (and still?) to do. <water> um no <menozero> (philosophical similar) <menozero> (almost) <water> no <water> really, i mean it <menozero> (ok) <water> i spend most of my time when communicating with tunesers trying to eradicate concept fusion <menozero> (guess it's time to see how much my wrong idea of 1994 tunes has evoluted) <water> such as "my idea is just like tunes!" 05:00pm <water> about half of tunes mlist threads are based on such fallacies <menozero> on the ml everyone had its idea of what tunes was. <water> yes i know that's why i didn't join for so long <water> i've read tunes docs about once a month for the last 5 years or so <water> i think i know more about tunes than Fare does, imnsho <menozero> (ok, don't want to be rude, sorry) <water> well i just feel frustrated that tunes does not attract enough clear-headed people <menozero> [back to my pl] the problem is how to compile any very-high-level-programming-language in a normal-low-level-machine <water> since no one else is working on ideas of similar value <menozero> in less than a week <water> too vague <water> and why don't you eschew dynamic run-time feedback? <water> it's not even a fully-mature way of optimization, and it's already proven very useful <menozero> it's just about how much run-time you want. <menozero> if you use a bytecode, it will never (os not) be as fast as C. <water> bytecode sucks <water> untie the concept of dynamic run-time feedback from bytecode, please <menozero> right. <menozero> I think it's still a static/dynamic or compile/run-time problem. And Dylan is a good instance of what we could use. <water> use for what? <menozero> while smalltalk/lisp/self/java will never. <menozero> "use" for usable. <water> bs <menozero> fast enough. <water> you like tunes? 05:10pm <menozero> [these are just my humble opinions.] <menozero> obviously I never understood tunes. :) <water> i don't doubt it <menozero> I never said I did. <water> what do you think of merlin, then? <menozero> what's merlin? self-related? can't remember. <water> forget it <menozero> maybe never known. why don't you tell me, instead of been slightly rude? <menozero> if you don't want to talk, just say so. <water> sorry, i'm pissed off <menozero> we were just chatting about Dylan and stuff, then I offended you not knowing tunes... <water> you're about the 100th former tunes lurker i've met <water> yeah i find it incredibly offensive <menozero> 1. I don't want to be rude. 2. you can't be rude just because of that. <water> i don't have the right, eh? <menozero> i'm sorry, but i never said i knew tunes. <menozero> i've read a lot, but not studied it. <water> i know, but you said you knew about tunes in '94 and that's a long time for an idea to (not) sink in <menozero> in 1994 for some months i've been on the ml. <menozero> then nothing. <menozero> some months ago I've tried tunes.org. <menozero> that's it. <menozero> the problem (which is not a problem) with a new projects, a really new projects, is that people don't understand it. 05:20pm <menozero> I have the same problem with a simple XML editor. <water> well tunes is 6 years old now <menozero> people need time. they need to come to it their way. <water> lol in 6 years?? <menozero> when they lose some Word 6 document and can't import properly the begin to understand there's a problem and start looking for a solution. <menozero> what's the problem behind tunes? why does it exist? <menozero> that's quite hard to figure from the home page. <water> or from anywhere else <menozero> yes. <water> and i totally disclaim any authorship of tunes docs other than slate <menozero> I loose interest in tunes years ago just because of that. <menozero> (i wish I could say I'm sorry, but I understand things quite easily, normally. so I thought that it was very hard or very not-understandable.) <water> well do you see any fault in dylan? <menozero> Not yet. <water> there's your problem <water> maybe that's the kernel of the problem that tunes tries to address <menozero> I'd like some pliant-ism in Dylan, but that would not be part of the Dylan style. <water> ew <water> have you tried maude? <menozero> no. <water> abi: maude? <abi> maude is a reflective rewriting logic language at http://maude.csl.sri.com/ or http://www.csl.sri.com/~duran/ <water> not a "practical" language, but very interesting because of what it can extend itself to handle <menozero> thanks for the urls. I usually know something. :) <water> it's the only language paradigm i know of that can actually solve equations in a general and configurable way 05:30pm <water> almost all of its ideas will be implementable in slate objects alone <menozero> nice. <menozero> as i said, i was interested in a "practical" pl. do you think it's wrong? <water> well in tunes terms it is <menozero> ok. <menozero> so, i'll even stop lurking... :) I really don't want you to be "pissed off". honest. <menozero> I think A and you think B and that's it. I have my resons and you yours. <water> whatever <water> no need for platitudes <menozero> (define "platitude", 'couse my italo-english can't get it, if you want me to understand.) <water> hm the meaning is subtly different from "over-used saying" <menozero> thanks. <menozero> banalita'. -:- lar1 [larman@adsl-63-204-134-241.dsl.snfc21.pacbell.net] has joined #tunes 05:40pm <water> yes that's about the right word <menozero> I'm looking at Maude pages. -:- eihrul [lee@usr5-ppp16.lvdi.net] has joined #tunes <water> re <eihrul> re <menozero> *Hi* I <menozero> I'm just a lurker who don't know tunes so just ignore me. <eihrul> ? <water> heh <water> eih: i've been giving him the treatment <water> eih: i still don't find my bmo implementation ideas satisfying <eihrul> what in particular are you musing over still on bmos? <water> providing a balance between protection and power <eihrul> protection from? <water> i guess the bmo's could just be really dangerous in the beginning and we could make them easier to deal with later <water> hm <water> well the bmo's can inspect the objects they handle <water> they basically can do anything to them <eihrul> well, shouldn't they be able to? <water> well we don't need that much power in order to get the features we want <water> like screening message-invocations and cloning 05:50pm <eihrul> well, how would you propose to limit their power and to what end is that actually useful? <menozero> Anyway, it's been a pleasure, nice "treatment". There should be a chapter in Software-Release-HOWTO on how to send away people almost interested in your project. <water> well it's useful in protecting novice users from screwing up the system really badly <water> meno: we have far too many people already <water> one more person with concept fusion wouldn't help us at all <eihrul> well, couldn't this be handled in the bmos that affect the particular bmo's meta-behavior? <water> perhaps but what protects the bmo's bmo? <water> :) <menozero> I don't have CVS permission. I just say stupid (because different) things. If you are not interested, just say so. You started talking about Dylan. I really don't want to disturb you. I'll see tunes.org in 5 years. <menozero> Bye. <eihrul> well, i suppose it'd infer some common bmo property available in all bmos that the objects are created from... <eihrul> so ensure that all critical objects know how to protect themselves and that other objects cannot damage critical objects <water> i think i'm making too big a deal out of it <eihrul> perhaps, for the first slate, i don't see security to be a big concern -:- SignOff menozero: #TUNES (using sirc version 2.211+4KSIRC/981227-pre0.9) <water> yeah we'd mostly be writing environment stuff anyway <eihrul> because we still don't have a working system (i know) to decide on how security concerns affect the objects at this point and where it is really needed <water> here's an interesting thought: how do i actually use meta-behavior to clone an object on message-reception? <water> i thought about this one a bit and it's sort of interesting <eihrul> you mean if the message, say, would modify the target object? <water> yes <water> (which is supposed to be the default, as far as we've discussed) <eihrul> i.e. (slot (set : foo)) ? * water/#tunes nods <water> the idea is that the bmo forwards the message to (object clone) 06:00pm <water> or some variant of that <eihrul> well, the cloning would be done in the 'apply' part of the protocol <water> yes <eihrul> when 'set' bound to 'foo' for input is applied to 'slot' <eihrul> so the 'set' message's result is a new slot object, whose value has been set to 'foo' <water> yep <eihrul> but, actually... the above example is a bit weird <eihrul> since the slot itself is just a message <eihrul> rmm, hmm, maybe not <eihrul> set would change the result expression of the slot message -:- ult [ult@user-37kbah1.dialup.mindspring.com] has joined #Tunes <eihrul> (this change encompassed in the new clone it returns) <water> i sometimes find it dubious to have every object return result implicitly <water> should we just have all results be explicit? <water> of course that's cumbersome <eihrul> well, the result is returned in this particular instance by application of the slot to its container -:- SignOff thomas: #TUNES (Ping timeout for thomas[193.217.63.152]) <eihrul> i.e. (object slot) -:- thomas [thomas@193.217.63.152] has joined #tunes <water> yes in this case it is <water> hm <eihrul> now... that expression itself is lazy <eihrul> and will not actually give a result unless their is some source of demand <water> however, when you assign, you don't want that result, do you? <water> yeah well that's basically assuming the outer object needs an explicit demand <eihrul> well, 'set' could alternatively return a new method <eihrul> that could be correctly applied to the container without demanding the result of the slot <water> hm <water> i can see ':' being a type of monad, though <water> since it returns a bottom-level state change <water> i mean, what would the code of a state slot contain? 06:10pm <water> if i say x : 4, what's the object x have as slots? * eihrul/#tunes hmms. <water> obviously it has a return slot which provides 4, but what's the code in the return slot <water> ? <water> oh wait <eihrul> (answer /container_of_4/4) <eihrul> or what-not <water> in order for : to work, we already know that namespace access has to be messed with <water> so if we simply have ":"'s meta-behavior to access another namespace, then the only slot of return would be 4(?) <water> that doesn't seem right <water> any ideas? <eihrul> well, x would have to contain a reference object to the real 4 <eihrul> (since it bridges namespaces) <eihrul> though it doesn't seem strictly essential <water> but reference objects could be just clones * eihrul/#tunes nods. <eihrul> with according bmo <water> yeah not original meta-behavior <water> er... not the mb of the original object <water> hm <water> ok obviously namespace access in the forward direction is not trivially the same thing as slot access as it is in self <water> for one thing, we've got result slots to think about 06:20pm <water> (when we access slots) <water> hm -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us700.javanet.com]) <water> so obviously if we want the object and not its results, we use '/' and not regular slot-invocation <water> this means '/' is more important than i originaly thought <eihrul> actually, no <water> why not? <eihrul> because application will still demand a result <eihrul> all that does is change lookup <water> yeah but the result of '/' is different <eihrul> and perhaps that is another question... <eihrul> how does '/' change the affect of lookup and or apply? <water> it has its own behavior <eihrul> eek <water> heh <eihrul> that's the problem with "special" constructs :) <water> yeah <water> well it's a reflective one, if we make it more than just a kludge <eihrul> because it requires intercepting the lookup before it gets to the bmo <water> it's one thing to say (object foo) .... it's wholly another thing to say "(/ object foo)" (syntax check, please?) <water> hm <eihrul> well, object/foo was originally intended to mean (object foo) itself <eihrul> not so sure about that now... <water> yeah and i see that it may have been a mistake 06:30pm <water> well right now i'm considering whether or not to reify symbols as is done in self/smalltalk <water> as in #foo vs foo <water> (quoting) <water> otoh they seem to need it in place of namespaces <eihrul> (object (/ : #foo))? <water> possibly <eihrul> though this also suggests a way /absolute might work... <water> how so? <eihrul> actually, i was wrong, it doesn't really... * water/#tunes returns to thinking about slots <water> it'll always have a ":" slot of course, within it <water> and the result slot *returns* what we think of as the slot's state * eihrul/#tunes nods. 06:40pm <water> making slots like those un-inspectable seems tempting <water> of course, my idea for the literal systems involves having *no* state whatsoever other than observable algebraic behavior (which is inherited) among numbers letters etc <water> the value would be determined by equations which referenced the objects <water> and the equations would just be other objects (a network of expressions connected by symbol-flow) within the same namespace, since they obviously return values if evaluated <water> that actually makes a lot of sense, whereas our basic slot idea would look suspicious to the user who started using the inspector to get around <water> maybe we should wave a meta-behavior wand, and be done with it ;) <water> but maybe there's something to that <water> eih: you still there, man? <water> bah <eihrul> yes <eihrul> eating <water> ok let me know when you get back 06:50pm >>> water [water@tnt-10-124.tscnet.net] requested TIME from TUNES -:- XeF4 [xef4@supernova.sevenheavens.com] has joined #tunes <water> hey xef4 07:00pm -:- SignOff XeF4: #TUNES (Read error to XeF4[supernova.sevenheavens.com]: EOF from client) -:- XeF4 [xef4@supernova.sevenheavens.com] has joined #tunes -:- nate37 [nate@ppp-207-151-70-37.dialup.pcmagic.net] has joined #tunes <eihrul> hey <water> am i bothering you? <eihrul> i suppose not <water> hm have you caught up on what i was talking about? 07:30pm <eihrul> yeah <water> well i was considering that the bmo could replace the apply phase with a direct forcing of the return value -:- SignOff XeF4: #TUNES (XeF4 has no reason) <water> iow the bmo has the state, although this screams meta-circular tower <water> what do you think? the bmo might have a table of lookup values or something like that which are only symbols <water> er s/symbols/objects like exemplified in the literal system/ <eihrul> yeah <eihrul> though if the symbols are not something the bmo strictly cares about, but need to be handled, this also screams composition <water> hm how so? <eihrul> well, it just depends on what particular symbols these are and whether they're expected to behave the same across all objects or not <water> i don't think i follow <eihrul> rmm, well, <- tired <water> you're always tired :) <eihrul> well, yeah, but right now i am especially so to the point where my level of coherency is affected 07:40pm <water> when can i expect you back on? <water> tomorrow? <eihrul> later tonight probably if i go to sleep now <water> ok beep me to get my attn if it's late <water> cya later, then 07:50pm -:- SignOff eihrul: #TUNES ([x]chat) * water/#tunes realizes some dual significance to the name "NeoCortex" of the software company in the movie "the matrix" -:- SignOff nate37: #TUNES (nate37 has no reason) -:- SignOff ult: #TUNES (Ping timeout for ult[user-37kbah1.dialup.mindspring.com]) -:- ult [ult@user-38lc63r.dialup.mindspring.com] has joined #Tunes -:- SignOff lar1: #TUNES (Read error to lar1[adsl-63-204-134-241.dsl.snfc21.pacbell.net]: Connection reset by peer) -:- lar1 [larman@adsl-63-204-134-241.dsl.snfc21.pacbell.net] has joined #tunes -:- nate37 [nate@ppp-206-117-3-246.dialup.pcmagic.net] has joined #tunes -:- SignOff ult: #TUNES (nite all) -:- SignOff lar1: #TUNES (See you next week in a brand new show!) -:- SignOff water: #TUNES (Ping timeout for water[tnt-10-124.tscnet.net]) -:- SignOff nate37: #TUNES (nate37 has no reason) -:- water [water@tnt-10-122.tscnet.net] has joined #tunes -:- nate37 [nate37@ppp-207-151-70-47.dialup.pcmagic.net] has joined #tunes -:- Eeeep [chico@usr31-dialup4.mix1.Sacramento.cw.net] has joined #tunes <water> eeep! <water> long time no see, eep <water> sorry, afk <Eeeep> hello =) 11:50pm [msg(TUNES)] newlog 2000.0406 IRC log ended Thu Apr 6 00:00:01 2000