03:11:29 billh joined #tunes
04:43:10 Fare joined #tunes
05:38:48 [QUIT] smoke quit: Ping timeout for smoke[15dyn156.delft.casema.net]
05:57:50 smoke joined #tunes
07:53:10 ult joined #tunes
08:01:09 [NICK] ink changed nick to: INK|sleep
08:47:20 [QUIT] gREMLiNs quit: Disconnecting
09:13:17 Kyle_L joined #tunes
09:33:51 tolillon joined #tunes
09:34:35 tolillon left #tunes
09:48:18 smklsmkl joined #tunes
09:50:19 <ult> heya
09:52:34 <smklsmkl> hello ult
10:14:22 [QUIT] INK|sleep quit: Read error to INK|sleep[ppp-206-170-2-31.sntc01.pacbell.net]: Connection reset by peer
10:33:08 [QUIT] lar1 quit: fastlane.openprojects.net forward.openprojects.net
11:10:56 eihrul joined #tunes
11:33:29 [QUIT] eihrul quit: [x]chat
13:26:07 Kyle_La joined #tunes
13:28:32 <Kyle_L> Who is over there?
13:33:57 smklsmkl is.
14:26:53 AlonzoTG joined #tunes
14:43:47 XeF4 joined #tunes
14:46:35 ame joined #tunes
14:46:44 <ame> hi all... I have question about linux semaphores...
14:47:22 <XeF4> you should ask #kernelnewbies about that
14:48:25 <ame> thanks...
14:48:29 ame left #tunes
15:43:38 [QUIT] AlonzoTG quit: Have Nice Day :)
16:21:16 [QUIT] smklsmkl quit: back tomorrow
19:16:10 Brianna joined #tunes
19:16:36 <water> hm
19:17:37 <water> have i lost you perhaps?
19:17:48 <bourbaki> was is internic?
19:18:01 <water> what?
19:20:43 <bourbaki> most of all the peoples in the channel have there country set to internic
19:20:55 <bourbaki> and not me?
19:20:59 <bourbaki> strange
19:21:00 water shrugs
19:21:17 <water> what client are you using?
19:21:36 <bourbaki> x-chat
19:21:50 <water> who knows
19:22:08 <water> i try to ignore lame software as much as possible :)
19:22:25 <water> even though i use it :/
19:25:20 <bourbaki> it is is often crashing
19:26:05 <water> how is fare doing, btw?
19:26:23 <bourbaki> i'd like to know too
19:26:46 <bourbaki> i'm living in canada and him in France so..
19:27:09 <water> i see
19:27:23 <water> where in particular do you live?
19:27:29 <bourbaki> montreal
19:27:33 <water> ok
19:28:12 <water> i live in seattle, so i had wondered if you lived near victoria
19:29:39 <bourbaki> ok
19:29:58 <water> icuc we do a lot of research here on programming languages and other ideas, so if you're looking for web pointers to learn from, just ask
19:30:14 <bourbaki> thanks
19:32:04 <bourbaki> icuc is the name of your compagny?
19:32:18 <water> heh
19:32:20 <lar1> heh
19:32:28 <water> nope, icuc means "in case you care" :)
19:32:37 <bourbaki> oo
19:32:57 <water> sorry, we use a lot of acronyms here to alleviate the pains of communication via keyboard :)
19:34:00 <bourbaki> i understand
19:34:33 <bourbaki> me, i'm makink a master degree in math
19:34:41 <water> cool
19:34:55 water has a library of math books
19:35:11 <water> to tide me over until i get back to college
19:35:21 <water> and to research
19:35:34 <bourbaki> wow, how many book?
19:36:00 <water> geez
19:36:04 water goes to count them
19:36:09 <bourbaki> i'm still at school to profit from libraries 
19:37:18 <water> about 30, not counting CS books
19:37:29 <bourbaki> ok
19:37:57 <water> CS makes for another 15, language theory another 7
19:38:05 <bourbaki> if i count my bourbaki's: 16 books
19:39:23 <bourbaki> so you see why i likes bourbaki
19:39:52 <bourbaki> i have to leave, see you again, bye
19:39:53 <water> actually i see that you *do*, but i don't see *why* :)
19:40:08 bourbaki left #tunes
19:40:39 <water> hm?
19:40:48 <water> did i offend?
19:40:56 <water> oh oops
19:41:07 <water> i didn't notice his send-off
20:10:18 <water> to repeat for the logs:
20:10:20 <water> * water has been thinking on arrow as well as slate
20:10:20 <water> <water> i just got a nifty book as well :)
20:10:20 <water> <water> "a functional pattern system for object-oriented design"
20:10:49 <lar1> Is tril no longer loging #tunes?
20:10:58 <water> apparently not
20:11:23 <water> it'd be nice to have an admin who actually knew enough and had enough time to keep things working properly
20:11:24 <lar1> Hmm, ok
20:11:36 <lar1> Indeed
20:11:56 <water> anyway
20:12:04 water goes back to reading and thinking
20:12:24 <water> most likely i'll write some more code for the squeak arrows prototype tonight
20:12:36 <lar1> Yay!
20:12:53 <water> i already have extended the previously published architecture a little, but it's not functional yet
20:13:33 <water> yeah i finally figured out how to implement it (as an oo architecture)
20:13:51 <water> of course doing it in squeak vs slate makes it a real bitch :)
20:14:07 <water> but even doing it in lisp would still be painful
21:07:27 lmaxson joined #tunes
21:07:33 <water> re
21:08:01 water waves to lmaxson
21:08:46 <lmaxson> lmaxson waves back.
21:09:04 <water> if i weren't working 80 hours a week at a job where i get no net access, i'd be more interactive on the mailing list discussion
21:09:16 <water> as it is, i can't help you out
21:09:51 <water> since i have a pile of books to read, more smalltalk code to write, and hopefully i can formalize a working outline for new documentation for slate/arrow
21:10:44 <water> did you ever notice the "epsilon" construct idea mentioned on the hll pages on tunes.org?
21:11:19 <lmaxson> Are you directing this to me?
21:12:00 <water> yes
21:12:20 <water> since you are the only person currently capable of a decent response at the moment
21:12:29 <lmaxson> Well, no, I don't remember the epsilon construct.
21:12:43 <water> and besides, i think it relates to your 4gl arguments
21:13:20 <lmaxson> I can go to the tunes hll documentation.
21:13:32 <water> maybe i should point it out
21:13:34 water looks
21:14:46 <water> for one, there is the "choice" construct point mentioned on http://www.tunes.org/HLL/semantics.html at 50% down from the top of the page
21:16:22 <lmaxson> I've just finished reading the Choice section.
21:16:49 <water> there's another paragraph iirc which covers it in a better light i think
21:17:08 <water> comments?
21:17:55 <lmaxson> I have no problem with the system filling in as much information on its own based on what is input as source.
21:18:43 <lmaxson> I still don't make a connection to your epsilon reference.
21:19:05 <water> a choice function is very similar to the epsilon construct of logic
21:20:21 <lmaxson> I guess I would have to plead ignorance of the epsilon construct of logic.
21:20:36 <water> which is an optimistic request for an object satisfying a *specification*
21:20:49 <water> epsilon... "there exists"
21:20:57 <water> it's Fare's term, not mine
21:21:08 <water> existential quantification
21:21:28 <water> as i said, an optimistic request for an object fulfulling a specification
21:22:03 <lmaxson> Having some experience with APL, REXX, and PL/I, all of which support implied attributes based on contextual use I don't see a conflict.
21:22:14 <water> a conflict?
21:22:20 <water> what are you talking about?
21:22:40 <lmaxson> A variable name and its attributes.
21:22:45 <water> sometimes i don't understand what you mean
21:23:01 <water> again, what does that have to do with what i was discussing?
21:23:10 <water> i mean a much more general idea
21:23:42 <lmaxson> When I read the material on choice it brings to mind implicit declarations.
21:23:44 <water> like taking a Z specification and creating objects optimistically that satisfy it
21:24:02 <water> that's not at all what tunes is suggesting
21:24:37 <water> (implicit declarations per se)
21:25:29 <lmaxson> I'm simply saying that's the connection I make on reading the Choice material.
21:25:42 <water> i see
21:26:27 <lmaxson> I'm not aware that you have other options for attributes other than explicit or implicit (context-based) attributes.
21:26:49 <water> yes there are many schemes for it
21:27:13 <water> most involve partial-evaluation ideas, relating to reflection upon context algebras
21:27:29 [QUIT] td quit: Ping timeout for td[ilm25-51-108.ec.rr.com]
21:27:37 td joined #tunes
21:28:14 [QUIT] td quit: [BX] Time wasted: 1d 17m 46s
21:28:15 <lmaxson> Do you have a choice, regardless of the dynamics involved, of something other than explicit or implicit attributes?
21:28:21 <water> have you any language documents yet?
21:28:34 <lmaxson> On SL/I?
21:28:38 <water> it depends on what you mean
21:28:45 <water> yes, on sl/i
21:29:10 <lmaxson> I mean if the writer "declares" the attribute, it is explicit, otherwise implicit.
21:29:19 <water> if the implementation is open and written in the hosted language, you have many options
21:29:56 <lmaxson> In both REXX and APL the attributes are implicitly related to the last contextual use.
21:29:56 <water> so, you just mean currying declarations as functions on symbols?
21:30:29 <water> but do REXX and APL allow modification of the mechanism from the source language level?
21:30:37 <lmaxson> If I declare a variable as fixed bin (31,7), that sets it.
21:31:16 <water> can you change the implicitness of declarations using source code in the program without resorting to non-uniform syntax?
21:31:33 <lmaxson> If you write a = '007', a is a string variable.  If you write a = 007, a is an arithmetic variable.
21:31:52 <water> yes i am familiar with implicit type declaration of that sort
21:32:02 <lmaxson> What it is, its current state, depends upon its last use.
21:32:09 <water> every programmer learns that in high school at the latest
21:32:33 <water> yes but that mechanism for deciding what type a variable will be... can you change it?
21:32:39 <lmaxson> COBOL, for example, does not allow implicit declarations.
21:32:46 <water> yes i know
21:32:51 <lmaxson> Change the mechanism.
21:32:54 <water> that's not what i'm asking
21:32:56 <lmaxson> ?
21:33:01 <water> heh
21:33:13 <water> i think i have got your attention
21:33:29 <water> the compiler does the type coercion implicitly, yes?
21:34:10 <lmaxson> If you switch compiler to interpreter and account for the dynamics of the order of execution, the answer is yes.
21:34:12 <water> but the compiler is not written in the hosted language, so you cannot use the hosted language to change the implementation, i submit
21:34:48 <lmaxson> We agree on that point.  That's why SL/I is written entirely in SL/I.
21:34:51 <water> tunes is about this process... doing so dynamically
21:35:24 <water> but does SL/I have a uniform syntax, similar to lisp or forth?
21:35:27 <lmaxson> Even the syntax is written in SL/I.
21:36:28 <lmaxson> The rules under which the language operates is written in the language.
21:36:34 <water> if it doesn't, i submit that SL/I will have complexity limitations that tunes will not
21:36:41 <water> that's not what i mean
21:36:58 <water> for example, Maude is a specification language that specifies itself
21:37:46 <water> but maude is only based on equational rewrite logic techniques
21:38:02 <water> how can SL/I possibly do better?
21:38:24 <lmaxson> SL/I includes the various target machine architectures as specifications.
21:38:28 <water> i suggest you look at maude
21:38:35 <water> so what?
21:38:45 <water> that's totally irrelevant to my question
21:38:55 <water> we are discussing semantics
21:39:00 <water> and syntax
21:39:04 <lmaxson> You were looking for a difference between SL/I and Maude.
21:39:09 <water> not supported hardware types
21:39:22 <water> that's another KIND of difference
21:39:34 <water> maude can and has been used for hardware specification
21:39:47 <water> so has lisp
21:39:54 <water> so it is not an issue
21:39:58 <lmaxson> The point is that the syntax, the semantics, the target achitectures are all defined in SL/I.
21:40:03 <water> so?
21:40:13 <lmaxson> Nothing is missing.
21:40:13 <water> i could do the same in lisp
21:40:23 <water> in fact, it HAS been done in lisp
21:40:47 <water> and yet those lisp systems were not what tunes hopes to achieve
21:40:55 <lmaxson> I am not arguing that it cannot be done elsewhere.
21:40:57 <water> so you have proved nothing
21:42:01 <lmaxson> Actually I wonder if there is an interest with respect to the Tunes HLL in developing a metric, something quantitative (and comparative) as opposed to qualitative only.
21:42:16 <water> there always has been
21:42:34 <water> however current CS research doesn't have that within its current scope
21:42:47 <water> it's not considered viable
21:43:02 <lmaxson> Is there a reason why?
21:43:22 <water> of course, most of the best systems aren't as good as the promises of tunes
21:43:55 <water> or as far-reaching or as well-integrated or what-have-you
21:44:41 <water> there are of course a few projects using languages like ML to do something similar to tunes, but they aren't very far along in progress
21:45:29 <lmaxson> Well, if CS is not up to it and so far has not managed to pull off a tunes, then why should we feel restricted to its consideration?
21:45:55 <water> what do you mean, "restricted to its consideration"?
21:46:30 <lmaxson> Why should tunes not do what CS has failed to do: provide a language metric?
21:47:01 <lmaxson> I have long been disappointed by the lack of science in CS.
21:47:04 <water> btw, Fare and i do our own research... that is what has produced Fare's several papers and eihrul and i's work on slate and arrow
21:47:25 <water> because we barely have the time to do what we do
21:47:41 <water> as i said, i have an 80-hour a week job right now
21:47:59 <water> and i know that the other tunes members are only part-time
21:48:09 <lmaxson> Well, it may be something that the rest of us could ponder.
21:48:20 <water> besides, it's an incredibly complex problem to develop a metric for
21:48:47 <lmaxson> We could start with operational definitions for the various tunes hll requirements.
21:48:47 <water> people have a difficult enough time profiling existing programs or benchmarking hardware meaningfully
21:48:59 <water> heh
21:49:05 <water> you know, i've asked that of Fare
21:49:26 <water> he has claimed that the requirements are clear enough as is (to him)
21:50:07 <water> but then he knows of many research developments i'll wager that you're not familiar with
21:50:24 <lmaxson> All we are talking about is making them "scientifically" clear, capable of independent evaluations.
21:50:36 <water> heh
21:50:44 <lmaxson> Water, there's scads I'm not familiar with.
21:51:01 <water> and i say that such a prospect is incredibly daunting, intellectually
21:51:25 <water> Fare has done well enough with his papers
21:51:35 <water> i suggest you read his LambdaND
21:51:50 <water> look in http://www.tunes.org/papers/
21:52:08 <water> darn, it's not there
21:52:15 <lmaxson> Normally for my clients I provide operational definitions that they can use in their own evaluations to allow them to achieve greater self-sufficiency.
21:52:37 <water> yes but your specifications are not potentially *self-modifying* :)
21:52:51 <water> in the case of usual software
21:53:51 <water> here are the papers: http://fare.tunes.org/tmp/
21:54:11 <water> all of those are tunes-related and most are formal papers
21:54:13 [QUIT] MysticOne quit: I am enough of an artist to draw freely upon my imagination.
21:54:28 <water> although Marvel and Pliant are unrelated
21:54:34 <water> (mostly)
21:54:45 <lmaxson> Where would you suggest I begin?
21:54:58 <water> it depends on your background
21:55:28 <lmaxson> brb
21:55:29 <water> i suggest ./rndlc/
21:55:33 <water> ok
21:55:39 water returns to reading
21:57:15 td joined #tunes
21:57:52 <water> re td
21:57:57 <td> lo water
21:58:12 <td> whats up?
21:58:39 <water> reading, thinking, trying to put together arrow code and slate and arrow docs
21:58:54 <lmaxson> I see four references on the rndlc page.  Are they different?
21:59:00 <water> for the life of me, i can't seem to figure a proper outline to explain the ideas
21:59:24 <water> they're the same paper in different formats
21:59:39 <water> i trust you have ghostview or another postscript viewer?
21:59:47 <lmaxson> Yes.
22:00:08 <lmaxson> I can be trusted on some things.<g>
22:03:20 <lmaxson> I've downloaded the rndlc.ps document and am looking at it now.
22:03:43 <water> i think i still have a copy somewhere
22:04:09 <lmaxson> This one is 22 pages in length.
22:04:43 <lmaxson> I do have a question on reflection.
22:05:53 water found the local copy
22:06:02 <water> what's the question?
22:06:34 <lmaxson> Am I reading it incorrectly that it is a limited form of reflection under discussion?
22:06:44 <water> heh
22:06:53 eihrul joined #tunes
22:07:13 <water> you expect a fully general paper about tunes that is also very technical? :)
22:07:17 <water> hey eihrul
22:07:21 <lmaxson> Software normally has a purpose.  I assume that truly reflective software would determine its own purpose within its own dynamics.
22:07:36 <water> huh?
22:08:03 <lmaxson> I've seen nothing that would allow the software to determine what it was going to do.
22:08:04 <water> sometimes the word reflection means just "the capability to reflect"
22:08:21 <water> are you serious?
22:08:45 <water> of course reflection makes that possible
22:08:47 <lmaxson> I'm serious as to what limitations to assume.
22:09:10 <water> only within the limits of the meta-programs written, of course :P
22:09:49 <lmaxson> Remember that when I was active in studying cybernetics in the 50's my personal hero was D. Ross Ashby of "Design for a Brain" fame.
22:10:24 <water> oh great
22:10:43 <lmaxson> The point is I can go either way without trouble.
22:10:56 <water> that's almost as bad as douglas hofstadter :)
22:11:07 <lmaxson> Almost.
22:12:19 <lmaxson> The point is that the "purpose" per se is set external to the software.  It is not a candidate for reflection.
22:12:28 <water> um no
22:13:16 <water> if the hosted language implements itself, you can write software that "sets its purpose" as you put it
22:13:19 <lmaxson> I just wanted to make sure that I wasn't setting limits not present in the documentation.  Afterall cybernetics would go to the next level.
22:13:46 <water> cybernetics needs a drastic overhaul
22:14:15 <lmaxson> You probably don't like neural nets either.:-)
22:14:16 <water> no one's seriously updated the subject in decades
22:14:48 <lmaxson> It's like a number of other good ideas.
22:14:48 <water> i do like neural nets, but they are very outmoded.... so much better could be made
22:15:04 <water> no offense, but i really need to read
22:15:24 <lmaxson> Then I will go to my reading as well.  Until the next time.
22:15:28 lmaxson left #tunes
22:15:28 <water> and this conversation is taking too much of my all-too-sparse free time
22:15:42 <water> yo eih!
22:15:53 <water> what's up with that guy?
22:16:05 <water> (eihrul, not lmaxson)
22:43:49 [QUIT] rares quit: Ping timeout for rares[nwhn-sh8-port111.snet.net]
23:04:48 <water> eih: in case you haven't noticed, i got the copy of fpsood today
23:05:32 <water> although most of my reading today has been on logic and set theory and related developments
23:30:31 <water> this book is really good
23:31:42 <water> except i don't think it touches on linearity vs non-linearity :/
23:40:39 <water> oh here it is... called the "unsharable reference" pattern
23:41:56 <water> although there's no mention of how to effectively duplicate references under those constraints
23:42:28 <water> which i suppose isn't easy for ordinary oo languages (perhaps)
23:45:21 I440r joined #tunes
00:18:36 <water> hm