IRC log started Thu Jan 20 00:00:02 2000 [msg(TUNES)] permlog 2000.0120 well, i don't have anything to discuss tonight, apparently good night, all -:- water [water@tnt-10-74.tscnet.net] has left #tunes [] 12:20am -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us845.javanet.com]) -:- witten [witten@adsl-gte-la-216-86-196-179.mminternet.com] has joined #tunes -:- witten [witten@adsl-gte-la-216-86-196-179.mminternet.com] has left #tunes [] -:- smkl [sami@glubimox.yok.utu.fi] has left #tunes [] -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- SignOff smkl: #TUNES (changing servers) -:- AlonzoTG [Alonzo@216-164-129-70.s324.tnt1.lnhva.md.dialup.rcn.com] has joined #tunes good morning om 05:30am -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes -:- rares [rares@wtrb-sh2-port221.snet.net] has joined #tunes Hum. Gakuk hum hei 06:30am -:- SignOff rares: #TUNES ([x]chat) -:- Kaufmann` [newbie@200.255.108.93] has joined #tunes -:- binEng [Anders@130.236.227.141] has joined #tunes There is no such thing as superior softwate. Software is either adequate or inadequate. - ArkenWork can you believe this? hardly wots softwate? it pisses me off when people try to invoke technological postmodernism as an excuse for crappy software What would "technological postmodernism" be? it's a vulgar example of utilarism Well I don't know what it is.. software should not be measured in whether it does the job or not, or by what end-users think of it technological postmodernism is the doctrine that it's impossible to make value judgements on technologies. which is ridiculous. does sound doubtable software should be measured by software critics just like art and literature and music ...or let me judge ;) that would be the day ;) 10:10am urgh but seriously the industry is the largest enemy of great software I'm afraid you're right yeah, but this guy is trying to argue that, since Windows is "adequate software", there is no reason for him (or anyone else, apparently), to adopt better software... according to him, there is no "better software" too many casual programmers and ununderstanding management, perhaps omg kaufie: blessed are meek :-) om... but still how will I get the $3 million dollars it will cost to do my OS? AlonzoTG: Prostitution. * AlonzoTG/#tunes expresses sincere gratitude to binEng * AlonzoTG/#tunes works binEng over with a 2x4 heh rob a bank preferrably a net bank how would you run away with electronic money? =\ uh,, hack their server and just give yourself an account with the money you want. bE: transfer the money to your bank on the Cayman Island and then go live there cayman island? I don't even know where that is do they have cable? It's the Cayman IslandS 10:20am but back to software.. it's nice when it is completed though an intent to build the greatest thing ever looks rather shallow to something which was actually built hey, this is cool... a Java X server :) slow or what? apparently not and it runs on MacOS, Windoze and a shitload of Unixen or unices whatever :) http://www.jcraft.com/weirdx/ 10:30am I'd like a type/object system that allowed the programmer to do all the argument validity checking directly in the parameter list, by way of boolean guards as in factorial(1) = 1; factorial(n | n :: Integer, n > 1) = n * factorial(n - 1) I suppose something like this could also be arranged in a message-passing-based system it's easy to add to lisp at least but it'll be runtime of course yeah 10:50am Something more complex I am pondering on adding such contraints a week I find time.. the problem is how to make a total order description(obj | accessed-at-creation-time? obj DocDB, exists? doc(obj)) = [:name name(obj), :type type(obj), :docstring doc(obj), :related relatives(obj)] wherein relatives(obj) would search the database for objects that have any relation to obj if you're serious and we find time this spring, it would be cool to make something like that yeah but we will need to find a way to order the various constraints I've been thinking for a while about adding a better type/object system into Scheme 11:00am another example, perhaps for an IDE: colour(sym | sym :: TypedSymbol, eval(sym, *curr-env*) :: #t (* every object has an "is-a" relation to #t *)) = 'black; colour(sym | inherit-from(colour(sym :: TypedSymbol)), eval(sym, *curr-env*) :: function) = 'red I find it easier to express this kind of idea in pseudo-code brb afk 11:10am back no, actually, I'm off for good -:- SignOff Kaufmann`: #TUNES (kathyanne gets kick'd / as leaves fall on still waters / she is here no more.) * binEng/#tunes is away 11:20am -:- hcf [nef@me-portland-us126.javanet.com] has joined #tunes -:- Downix [down@d-gnaps-103.ici.net] has joined #tunes -:- SignOff Downix: #TUNES (BitchX: a modern client for a old world) -:- dalvarez [ircusr@212.68.72.98] has joined #tunes -:- SignOff binEng: #TUNES (Ping timeout for binEng[130.236.227.141]) -:- bineng [Anders@130.236.227.141] has joined #tunes -:- SignOff hcf: #TUNES (Leaving) -:- SignOff dalvarez: #TUNES (Leaving) -:- Nicky [user9379@modem-86.conjuror.dialup.pol.co.uk] has joined #tunes -:- SignOff Nicky: #TUNES (Leaving) -:- trouble [user6033@209.80.14.125] has joined #tunes er -:- trouble [user6033@209.80.14.125] has left #tunes [] -:- zarq [zarq@10dyn182.delft.casema.net] has joined #tunes hey 01:10pm hi 01:20pm * AlonzoTG/#tunes left clicks on zarq * zarq/#tunes unmounts alonzo -:- jdl [jiml@ultra1.inconnect.com] has joined #tunes hey, bin, check your mail :) ah hi, yes I just saw I'll hang around for a bit if you want to discuss specific entries. ok, I'll look at it now then bin: msg me to talk -- I'll be on a different desktop. I don't know if abstract aspects often are the essential ones. 01:40pm And again, in 'abstraction', "essential" might be the wrong word No need to msg me with the comments... getting a msg just activates a beep so I know you're saying something. :) alright then .. /msg hey you is enough :) anyway. 'concretisation' might have been a word I just came up with, but I did mean something different than 'reification' (and read what I wrote earlier= ) anyway, regarding your comment, I agree... abstraction is a tough one well, the dictionary definition of reification I have IS the opposite of abstraction and is also the same as 'concretisation'. hmm.. yes I guess reification is to make something abstract concrete abi: foldoc for reification? reification may be sought in foldoc at http://wombat.doc.ic.ac.uk/foldoc/foldoc.cgi?query=reification http://wombat.doc.ic.ac.uk/foldoc/foldoc.cgi?query=reify : "To regard (something abstract) as a material thing." yep, that's what my dictionary says, too. Brian Rice uses it differently, but he has a history of redefining words for his own purpose (not that I don't) he does? I've understood his meaning of it as consistent with the rest of the world (on this one, that is) oh, maybe I'm wrong then. 01:50pm * jdl/#tunes is away again Should I erase everything on metamodels and codes, as they apparently differ from your (and probably also the right) definition? That's your call. Obviously, I'd prefer you used my definition, but it's your glossary. I have the feeling the semiotics tutorial uses it your way, so I'll let it go that way. 02:00pm hum hi Fufie jdl: What do you consider are Prism's semantic atoms if any? _Bit and _List you sure? They don't seem very semantic to me.. hi bin Okay, you caught me. :) Let me see if I can rationalize this one.... ;) I've heard the word 'epistimic' or perhaps 'emistemologic' for it. Brian uses it. I'll check bin: Okay, _Bit and _List are the Prism metacode's *syntactic* atoms. The metacode itself doesn't have semantic atoms, by design. -:- hcf [nef@me-portland-us905.javanet.com] has joined #tunes yes, I think Brian calls those epistemic building blocks brb sheesh... "epistemic building blocks"? What a mouthful ;) jdl: better than a mouth full of metas dictionary: "epistimic: of or relating to knowledge or the act of knowing". hcf: mea culpa... but I still prefer "atom" or "syntactic atom" to "epistemic building block" 02:10pm -:- pdm [pdm@host235-89.numega.com] has joined #tunes * jdl/#tunes is away again (msg me to get my attention) jdl: yep, epistemology is related to cybernetics * bineng/#tunes is back back bin: I think trying to come up with a unified philosophy is great, but I don't think you've got all the kinks ironed out yet. Hence my frequent "philosophy" brush-offs. I don't know how much that can be judged from the glossary, though.. even if I did have that in mind when writing 02:20pm ah yes, I agree I should move that to somewhere else, I'm just reflecting over that you said they didn't seem solid bin: they may have seemed shaky just because I didn't see the whole picture, just bits and pieces in definitions. yeah.. whatever. I don't want to argue that I have something I'm not ready to show up. I'll see if I can continue the stuff I sent to the mlist.. the "on modeling" bin: what I did see looked promising, fwiw fwiw? abi: fwiw is "for what it's worth" wish i knew, jdl argh abi: fwiw is for what it's worth wish i knew, jdl jdl: stop okay, I'm obviously clueless hcf: how do you add a factoid to abi? I thought it was " is " jdl: what needs a \ before it abi: fwiw is For \What It's Worth not that the cleartext makes it any clearer to me hcf: Ah, a lambda? ;) ah... I see. abi thought I was asking what "it's worth" was :) bineng: yeah ;) a conspiration! they're everywhere! (even on my computer's desktop, btw) bineng: \ as escape iz more common :p -:- air [brand@p0wer.qzx.com] has joined #tunes Didn't you know the longer strips on barcodes stand for 666? Yes it reveals how satan is in conspiration with the government. bineng, jdl: back to topic there is no topic -:- hcf has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System || slate || jdl's meta-isms meta-isms? :) ok 02:30pm bin: when you talk about isms -- you know "Is 'm hot? Is 'm tired? etc." ;) -:- Downix [down@d-gnaps-253.ici.net] has joined #tunes Hey-lo all hey downix hullo Downix hey-lo jdl: Do you think an entry for 'program' is needed/would be interesting? you talk about that in TW&TW, don't you? bin: I think I got one in there, actually ah ok -:- hcf has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System || slate || sphere/prism || jdl's meta-meta-modeling -:- jdl has changed the topic on channel #tunes to: TUNES, Free Reflective Computing System || slate || binEng's glossary the topic lines are long these days lol bineng: usually means more stuff is happening put the sphere one up, it'll PO Alonzo heh * jdl/#tunes grins Downix: that was my point hcf: Let's hope so :) hehehe jdl: You think 'concept' is not needed? Downix: why will it piss him off? air: he claims that he came up with Sphere although he didn't, ofc ahh bin: My reasoning there was that we're using concept in the traditional sense, and that everyone already knows what 'concept' is. jdl: Otherwise I thought I had a point in the limitation of concepts. But if you don't think it's needed.. bin: Let me look at your def again. I was going pretty quick, may have made some bad calls about what was needed/not needed. Webster: "The words conception, concept, notion, should be limited to the thought of what can not be represented in the imagination; as, the thought suggested by a general term." 02:40pm To tell you the truth, I don't really understand that definition. "Should be limited to the thought of what can not be represented in the imagination"??? sounds like the opposite of what it should be (and what I've read it as) what if the 'not' is removed? well, I still think everyone understands what a "concept" is. alright jdl: as for 'concrete machine', this was used by you, I think. or perhaps not bineng: aware of http://www.onelook.com ? hcf: no I wasn't bin: yeah, it was, but just to illustrate a point. We can include it but I don't think anybody actually uses that term. bin: also, with general definitions of "abstract" and "concrete," perhaps specific defintiions of "abstract Xxx" and "concrete Xxx" are unecessary. sure but what about 'virtual machine'? I'm not so clear over the definition myself. -:- eihrul [lee@usr5-ppp55.lvdi.net] has joined #tunes I think "virtual" and "abstract" may be closely related terms. I'm a little fuzzy on that one too. oh well, I didn't have any definition on that one myself dictionary: "virtual: existing or resulting in essence or effect though not in actual fact, form, or name" fuzzy alright So a "virtual machine" is a machine that exists and works but has no physical form... works for me I wonder if we should use "virtual" instead of "abstract" for some of our terms...? perhaps a machine/environment that was modeled after a physical machine? s/was/was not/ :P perhaps.. * jdl/#tunes shrugs what? Doesn't look like we have that many 'abstract ...' words ast is one word No, not really. I think the way it is now is fine. Let's leave virtual out for now. 02:50pm Fufie: what's that? it's virtually impossible to define virtual do we need a definition of 'constant' with your simplified meaning of it? s/no, not really/I agree, we don't have that many/ bin: what an ast is? an ast is, like, Andrew S. Tenebaum, hp at http://www.cs.vu.nl/~ast/ or Abstract Syntax Tree bin: your call... my original instinct was to say 'not needed' ah jdl: I only considered 'constant' interesting under my ""philosophy"" bin: okay, that makes sense. Take it out then? sure Fufie: That link is to tannenbaum :) * jdl/#tunes thought it was amusing Tannenbaum came before "abstract syntax tree" hehe, you're right :) tanenbaum is a nifty guy anyone got a good ast (abs.syn.tree) related link? jdl: How about 'cybernetics'? perhaps not interesting for Prism? hcf: not really, it's sortof implicit in compiler jargon Not for Prism, no, but I think it's a useful entry. fine then (Especially if we merge with Tunes glossary, since Brian (and others?) is so into cybernetics) hcf: though ASTs are also usually used for semantic (or semantically analysed) trees as well jdl: Wouldn't you want Far�'s comments in the glossary (re)moved for that? fufie: I saw a cool video where this guy created a electrical circuit code based on AST's. Then he wrote a genetic algorithm to manipulate the trees, and the results were in some cases equivalent and/or superior to patented circuits. jdl: Even if neither you nor me likes the definition of 'domain', do you think it appropriate to mention? Could be improved later. Cool stuff, but a little scary :) fufe: re Fare's comments -- I didn't realize he had some s/fufe/bin/ jdl: That's the politics bin: yes, I think domain is an important term (I use it all the time myself) but I can't define it :p jim: circuits are kindof ok to be evolved.. it's more difficult for ASTs for normal programming languages abi: dads? well, dads is at http://hissa.nist.gov/dads/terms.html jim: Koza's book has a chapter on circuits alone I seem to recall bin|jdl: check dads out that is.. Koza's first book fufie: only (IMHO) because languages have all kinds of syntactic cruft. With a customized Prism code it would be easier, I think (since Prism cuts out the syntax) 03:00pm hcf: whee, lotsa info * jdl/#tunes is looking at dads I'll have to look at it some other time jim: a trimmed AST is the tree of the actual code.. jim: I can't see why an AST ready for the backend will be any worse off than Prism jdl: Do you think 'programming language' and 'language' should be the same entry? bin: take a look at dads, it might be useful (it's a dictionary) ok fufie: no, no difference really, but an AST based directly on a language grammar has lots of cruft in it fufie: once I add "floaters" / "annotations" (haven't decided on the term yet) to the Prism metacode, it will be a little more unique. only beginners define an AST directly on the BNF grammar an AST's main point is to represent the code in a useful fashion for three parts: the semantic checking, the optimization and the code generation bin: Programming language vs. language: I think they're quite different... programming languages are a subset of all languages (and languages are a subset of all codes) sure. (unless you want to change terminology code->language) jdl: Haven't you read GEB?? fufie: yes, I know. AST's and the Prism metacode are very similar in concept fufie: different in execution (but not tons different) bin: juries still out on code->language -- it might be more clear, but I also think the distinction between lanuage (as a concrete code) and code is useful. bin: About the first third :) Great book. I see Prism (when I go much of the terminology nonsense) as a way to represent the ASTs in a simplified format basically.. ok. otoh, the word 'code' itself is confusing basically as the serialising format jdl: I ask because of 'holism'.. please don't use the word 'holism' Fufie: why? quaks use that word quacks I mean eh? quake? fufie: That will change once I add annotations. AST's don't have any sort of support for semantically-null annotations to the tree; the Prism metacode will. no.. quaks.. people who sell you snake oil quacks 03:10pm bin: Don't remember the word 'holism' from GEB... it's been a while, honestly. Fufie: I still don't get it completely.. but n/m, I don't intend to use 'holism' too heavily jdl: It is perhaps not mentioned until the middle or something don't use it bin: con artists selling fake/useless medicine tend to use the word "holistic" a lot, I think that's what fufie means. bin: Same for other new-age whackos ok the word 'holistic' is used by those who're unable to use scientific terms ...plus hofstadter ;) jdl: isnt prism a holistic panacea? ;) fufie: the annotations will allow Prism codes to support comments, symbols, etc., without adding complexity to the mainline semantics. hcf: smile when you say that, buddy ;) (oh, wait, you did) jim: I look forward to any decent extensions to normal AST thinking jdl: Don't you think 'interoperability' might need some explanation? (or not) bin: you're a CS student, no? Fufie: yes * bineng/#tunes refers to his hostname bin: in CS you can be roughly be 100% certain that someone who use the word 'holistic' when talking of his research is a wanker ;) fufie: I haven't worked out the details, but basically I'm thinking any Prism atom can have 0+ "annotation" atoms attached to it. Those atoms will have a type GUID that indicates whether they are symbols, comments, etc. And codes can recommend st standard annotations. I'll be sure to keep that word for myself then :) lol fufie: but annotations can be removed/ignored without changing the meaning of the model. jdl: Don't you think 'interoperability' might need some explanation? (or not) bin: holistic people are those who make a giant preprocessor macro shceme in C and call it object-oriented or that it will solve problems, instead of writing or using a better language.. (close to Bjarne this? ;) bin: oh, sorry.... um, I don't know. It seems like an obvious everyday word to me, but perhaps not for others hm don't ask me, I'm not native English bin: perhaps we should add "GUID" as well I'll let you stand for that part interoperable is a marketing term in my book -:- Kaufmann` [newbie@dial338.infolink.com.br] has joined #tunes Boy, DALNet sucks! 03:20pm then why do u hang out there everyday? Kaufmann`: What luck you're here instead then, isn't it? hey Kauf air, wish I knew :) bin, I was just there jdl: Were you unhappy about my 'isomorphism' entry? I was logged in for a minute, enough to witness three splits. lol : GUID / "Globally Unique Identifier." An identifier that is presumed to be unique anywhere on (or off) the planet. GUID's only apply within particular areas... for example, an IP address is guaranteed to be unique in the domain of IP got it (cont) but not necessarily unique in the domain of Ethernet addresses. In other words, there may be an IP address and an Ethernet address that share the same ID, but they are still considered to be unique within their domains. bin: I think isomorphism exclude any deviating models, arrows, objects in the two categories you compare Downix, re. company names... Ra Processing? bin: that should be clearer.. everything in one category should be mappable to the other and vice versa taken Kauf bin: let me doublecheck 'isomorphism' Kaufmann`: mentioning anything to do with the sun and processing makes me immediately think of cooling devices :) Fufie: right well, Th�rshammer? Sounds like a thrash metal band... bin: my problem with your definition was that it was specific to metamodels... not that I particularly like my definition aha better if I change it to 'codes'? bin: I'd still prefer a generalized def Okay, I've got it: _|_ (read "Bottom") Computing what do you say, Down? Kaufmann`: We Diverged From All The Other Companies... jdl: sure uh.. no heh jdl: What's your opinion on "Machine-readable metamodel"? okie bin: it's a metamodel that's readable by machines... seems straight-forward to me (so does it need a def?) True Software: Because all the programs we write always return #t! :) 03:30pm I'll remove it then kaufie :) jdl: What do you say, do you believe in the great metametametamodel? :) Kaufmann`: gensym oooh * zarq/#tunes is away: z gensym is a good company name lol Don't go insulting my religion! lol who needs a company name and what should they do? Fufie, Downix bin: if Prism is a metacode, then English is its metametacode, which means all the English phrase books in the world are metametametamodels. (So there!) ;) LOL * Fufie/#tunes thinks Hard On might be a good name for a hardware company otoh, english is sufficient for all higher meta-levels bin: it is not.. a lot of norwegian can't be expressed in english LOL Fufie Fufie: uhh. whatever :) bin: and english has terms not easily expressible in norwegian :) oooh, Macrohard :) Hard On Cafe :) * jdl/#tunes thinks Macrohard shouldn't be easily expressible in any language nor should the opposite.. * jdl/#tunes cringes at the imagry REPL Computing CrippleWare bin: okay, back to the glossary... where were we? jdl: A note, the syntax requires pluralis forms to be defined as aliases in case they are used.. Oh wait - Katmandu Inc. jdl: I'm working on it. What's a 'metasystem', anyone? bin: I realized that after I had gotten about half way through... at that point I decided to singularize them all and let you sort em out :) a metasystem can be an operating system a system of systems? jdl: Sure, just making sure you know gakuk fufie: Actually, I disagree... I don't think an OS would be a metasystem... I can't think of any examples of metasystems. in general, a system that can contain other systems perhaps? hey fare literally, a 'system of systems'. That does sound like a system capable of hosting other systems. bin: I don't think "containing" is enough to qualify for meta-. jdl: what more then? u guys need to do a little less talking and more coding :) 03:40pm air: welcome to tunes ;) :) air: I am their coding alibi ;) Downix, Hindu mythology allowed? air: rmm, and what do you have to show for it? Fare: What's a metasystem according to you? Kaufmann`: of course