IRC log started Sun Jan 30 00:00:00 2000 well, any sexp syntax system will only handle (produce and consume) lists serial processors computers don't access data an eyefull at a time :( [msg(TUNES)] permlog 2000.0130 so either i support a standard ordering of slots in a set or i record the order in which the slots were entered or i throw out sexp for object-based syntax elements i like the latter though it makes the system do a little more work which latter? latter = record ordering hm not necessarily more work more booklkeeping yeah, but all you have to do is record the object as list and just treat it as a set which is how smalltalk and self handle sets anyway if i *do* use option 3 (generalize sexp), then what i'm really doing is enhancing the language with more syntax operators than 'car' and 'cdr' but that might be complicated for code-transformations (macro-style) -:- witten [witten@un.torsion.org] has joined #tunes back in 10 min heh. alt.binaries.e-book has quickly become a "text warez" list heh people type in / ocr books, or what/ ? ocr -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us1028.javanet.com]) ah that's actually encouraging like for instance the ad&d manuals have been hacked people are stealing important and valuable items not just trhe usual rot as well as a bunch of famous sci-fi hacked? ocr'd oh 12:10am intellectual property is going to face lots of challenges there are some crappy computer software manuals though :( kinda hard to "ctrack" where there be no barriers" yeah intellectual property sounds so much like sanitation engineer hehe yeah "here's lots of technology to make you more powerful, but please don't violate companies' arbitrary copyrights" :) "please"? they break down your door and haul you off to jail if you do heh yeah i forgot about that people need to evolve out of the commie need based system this is a capitalist country for God's sake although it'll get increasingly tough to break down the doors and haul off to jail about 90% of the population since when was technology about giving power to people? ;-) rares: eh? i though it was about concentrating it in the hands of the elite :) ah well. everything tends in that direction most of the time anyway no, it's all about arrow :) communism says give to those who need most and everywhere you go it's paranoia, leghal terrorism, and NDAs heh. communism doesn't work, either exactly we're supposed to be a capitalist (value based) society not a commie (*survival based") society rares: NDAs and communism? yes i'm not following the connection Control this, control that, every body freak out don't speak a word or someone will take all you own and? soon, please as opposed to value based where you actually are satisfied and HAPPY at some point thx wow. those ad&d manuals are pretty big... maybe they indluded pictures they'd be useless without the tables well ad&d pics are usually pirated for their own sake 12:20am -:- SignOff rares: #TUNES ([x]chat) too bad the conversation turned to politics 12:30am water: heh. that was the most minor politics discussion i've ever seen on IRC yeah, it could have been a lot worse 12:40am -:- rares [rares@wtrb-sh4-port223.snet.net] has joined #tunes damn, i just died uh so you have this script on your client that informs us? how user-friendly :) heh, no :) 12:50am darn it, a wysiwyg html editor for ce shouldn't cost so much s/ce/wince how much? wysiwyg html editor? evil. =) $20 it's not like it's that hard to write, i just don't have the time would it take $20 worth of your time? much more than that well then there you go but i can web-serve for free ironix ironic web serve from a wince machine? yep * witten/#tunes winces 56k modem for mine hell, i could just use squeak hmm, my damn hdd is thrashing like crazy hmm, close netscape and it stops heh also freed up 50meg of ram and swap that's good ol' netscape for ya 01:00am i'm running xchat what video card? millenium II why? hmm i want to know how good the riva tnt support is in xf i don't know -:- SignOff washort: #TUNES (Ping timeout for washort[d102.narrowgate.net]) i'm going to get a g400 though. i hope the xfree support is decent whoa! emacs for wince! oh it is i have a g200 microsoft uses emacs internally for development which is horribly ironic but im using an 8meg frame buffer so its kinda slow heh and my g200 only has 8meg ram :) my riva has 16meg so it should run faster, but i know that matrox has superior support in linux and dont want to have xf not utilize all the accelleration on the riva umm 2.3.41 has DRI AGP USB support supposedly quite solid dri? Direct Rendering Interface 01:10am -:- SignOff water: #TUNES (Ping timeout for water[tnt-10-216.tscnet.net]) -:- SignOff rares: #TUNES ([x]chat) -:- smoke [smoke@16dyn132.delft.casema.net] has joined #tunes -:- water [water@tnt-10-3.tscnet.net] has joined #tunes bleh. emacs for wince really sucks why? even the elisp support is terrible, and they went out of their way to make it text-based on a system where the api's are gui-based -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-12.vpn.uib.no]) i can't even use my syntax-highlighting .el files hm well the base for emacs is text oriented, so it was probably easier to port it that way syntax highlighting isn't supported in the text console in unix either gnu emacs makes very well for a complete rewrite 01:40am i don't see how, really... the wince api's would just mean they'd have to write another layer they probably have done that :) just write a simple widget that does fixedwidth textmode emulation, and plug that in where the textconsole .c was darn, a nice free rpn-style calc that won't run on my particular system yeah i guess supporting variable width fonts and losing freedom over key-controls would've been more troublesome i think hm they wouldn't have had to do that, though how does one call the phenomenon in which a program is very portable, but in fact isn't (like emacs) :) heh they should've abstracted the design more from the text console? i dunno, but the concept of an elisp-extensible text editor shouldn't be too hard to port yes imo but at the time they designed it, there were only very experimental gui's wouldn't it be hard to abstract then? i don't see it as very hard to write an elisp-extensible wince app that only uses fixed-width fonts, supports multiple buffers, etc -:- witten [witten@un.torsion.org] has left #tunes [] wouldn't it be hard to support all the keypresses with a standard widget? the control characters hardly even apply on wince note that i don't know winCE well, i suppose you could map one of the hw buttons to CTRL but otherwise all input goes from the pen to the system's virtual kb or jot pane and then to the app in focus 01:50am aha the wince ui approach is to put the most useful tools at the top button-bar (which fits at most 10) well, that's for the palm-size, the handheld size has twice the screen width i thought wince looked like win9x it does well, the widgets do i think it's a terrible gui so do i but if you want the hw... well win9x then. i don't know winCE except from glossy folders and linux ce isn't ready for prime-time yet hm. a free forth cross-compiler for my machine 02:00am yes! i found a free html editor w3 ? w3 is probably nice, but will stop my whole xemacs as well abi forget w3 water: I forgot w3 i'm about to try it out jfc "the beta period has timed out" on freeware, and the company doesn't even sell the software 02:10am it's funny how `natural' opensource software gets to feel after a while.. m i mean that it's hard to understand why people want to actually sell the software except for their own good i just can't wait until squeak gets really mature (maybe a year or so) (proprietrary that is) i've checked out squeak this week, and i was both impressed and disappointed with it well, the jit support hasn't been added yet and the interface still isn't nearly completed keep in mind also that it's very difficult to notice all of the features that are there since the docs are all >8 months old, except for code comments -:- SignOff air: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep) trust me when i say that once the 2.x series gets closed out, a lot of that hidden code will become very useful and accessible 02:30am now, for example i don't mind paying 10$ for a good wince app but honestly anything over 40$ can't possibly be worth it unless its for a job (=> the company pays for it) and even 20$ for a simple app is overkill 02:40am heh 02:50am -:- Fufie [stig@tunnel-44-14.vpn.uib.no] has joined #tunes hey fuf morning water * Fufie/#tunes goes downstairs to grab some breakfast 03:30am well, i thought about lisp macros and realized that a primitive 'apply' within slate would handle that rather nicely within sexp the problem is that not quite everythin fits as an apply just yet I saw the mail to slate, I'll try to find time today to have a look ok if it's any comfort, i can guarantee that message-send composition always allows macros the real trick is to unify objects properly with messages as functions 03:50am is that much harder than: (message object arg1 arg2 .. argN) ? :-) (a very lisppy answer I guess :) -p that doesn't look right the object defines its protocol that's an important diff between oo and fp ([obj msg] arg1 arg2 .. argN) hm that suggests applying the object to the selector, creating a new object-function which is applied to arguments where [obj msg] is an expr that returns the function yes but that makes partial-eval weird (which was a feature i had wanted) how does it make it weird? well, partial-eval would involve closing a selector with one argument binding to make another selector you mean currying? that way, the selector changes before getting passed to the object sort of ([obj (curry msg arg1)] _arg1 _arg2 .. _argN) i hate that word no that doesn't work how do you designate binding arg2 curry once more or give a second argument bs 04:00am what if i want to bind the 2nd argument but not the 1st ([obj (curry msg arg1 arg2)] _arg1 _arg2 .. _argN) that's useless i can't distinguish which inputs i'm binding -:- hcf [nef@me-portland-us924.javanet.com] has joined #tunes ([obj (curry msg :arg (2 arg2))] _arg1 _arg2 .. _argN) lol i thought you're ideas about syntax were all simple? :) I just added keywords :-) s/you're/your you fudged it... that's too weird to grok the point of prototype oo is intuitiveness (let ((msg (curry 'foo :arg (2 arg2)))) ([obj msg] _arg1 _arg2 .. _argN)) pah it's obvious i need to keep working with original oo syntax lol that's crap it looks nicer in emacs than on irc :-) no oo prototype language will ever use "let" how will you bind values to names then? what are you sending the let message to? LET is a special operator oooooh.... "special" no thanks how will you bind values to names in slate then? that's just what i need... some way to break the oo mmodel just to get the system to work with messages to namespaces similar to that done in self I never like that counter-intuitive lol yeah sure like(liked agh admit it... you don't really like objects :) I find message-passing restrictive compared to generic functions you just want functions, but they're too broken, so you need to emulate objects :) no offense, but i have better ideas for slate object protocols than usual oo langs but a generic function cannot belong to an object not to a protocol, no but it can belong to an object something like an interface or a sort of meta-object how can a generic function belong to an object? (a generic function is an object with its own meta-object, but it doesn't belong to any objects.. it's just specialised for some classes) well, what makes a function generic? 04:10am polymorphism? it dispatches on all of its arguments yes, that's polymorphism boy are you stuck on lisp terms why can't one regard a generic function as an object? a generic function is an object but it doesn't belong to an object :-) well, in slate, all objects are functions and vice versa i had the greatest trouble implementing functions sharing two objects correctly in c++ because there was no way to represent it as an object you're really stuck in the lisp-mentality now or c++ even worse :) I am not -:- jeroen_ [jeroen@ap01-d03-168.dial.freesurf.nl] has joined #tunes where do generic functions belong to then? a global function-share? in oo, meta-objects can define shared polymorphic protocols or do they belong to pairs of objects? this is pointless it's like trying to explain oo to people who keep saying "where are the functions?" * Fufie/#tunes is smiling hello jeroen oop ain't fp * smoke/#tunes thinks he knows OO better than FP it doesn't translate at all oop is a subset of lisp hi lol ok now you're just being boorish fufie: that seems untrue, or badly formulated at the least :) not really oop may be equivalent in a sense to a subset of lisp smoke: it isn't untrue.. it is a revelation one gets after programming lisp for some time well, you reject any self ideas, which is one of the bases for slate water: where can i read about self? lisp can't handle non-determinism properly among other things ok self.sunlabs.edu .. sorry it cannot? *smiles* not well enough ehm that site is down :( abi: self? it has been said that self is a prototype-based object system at http://www.sun.com/research/self/ or old mirror at http://www.cs.ucsb.edu/oocsb/self/ * Fufie/#tunes thinks water needs to actually hack lisp code i have it's much easier to handle streams in squeak than in lisp * Fufie/#tunes smiles again wow. what a jackass haha this is NOT about which lang is better no it isn't it's about understanding what one speaks about you clearly do not understand lisp i'm saying that oo langs (as opposed to oo from within fp) is more intuitive that's right! because lisp is non-intuitive but what if your intuition has changed ? either by learning OO or non-OO ? oh god a lot of people find mathematics intuitive, and even more find it non-intuitive "direct-manipulation" user-interface does that mean jack to you too? * Fufie/#tunes thinks water talks like a smalltalker * water/#tunes thinks that Fufie talks like a lisper and smalltalk is a nice language but it is still just ya programming language so is lisp it isn't a programmable programming language yes it is the key thing about understanding lisp is the programmable part i know 04:20am so if one wants oo one creates oo if one wants non-determinism shut up with the lecture, i've already gone over it one creates non-deterinism no kidding * Fufie/#tunes thinks he should head outside when it's not raining.. a few errands sounds like a good idea have fun afk geez that guy claims he likes the slate idea, then berates everything about it that ain't lisp :) of course, no one really cares if only i could get someone to help with the issues without turning slate into their own idea hm maybe smalltalk isn't a programmable programming language, maybe it's an informational progrmmaing language or something nobody did that w/ arrow, did they? heh yeah they did much more of that there's a paper i have where they go over why oo isn't just a programming language idea they actually prove that it's more expressive the problem of course is that you can't program much of it, so programmers feel less powerful does self or beta have any irc presence at all? because it actually gives the end-user more power nope -:- jeroen_ [jeroen@ap01-d03-168.dial.freesurf.nl] has left #tunes [] hm could it be that visualizing objects as done in self and squeak decreases your power to alter, create and build new ideas? *decreases*? it's like the lagrange/laplace discussion in mathematics where you are either allowed or disallowed to have a graph for a function or an image for a given thought, as to increase your skills yes decrease. i always read about how it's intuitive and helpful, but it may as well be irritating, disrupting, neglecting (other things) and limiting i don't quite follow 04:30am ehm. if i have a problem and want to program a solution for it, i usually do that by very abstract means in say C very abstract and C? i'd imagine (though do not know for sure) that in Squeak, one would think about the problem as consisting of objects living in a preferabbly 2d space is that a rude misconception of self/squeak/smalltalk? yes (i assume i'm wrong about the 2d space) but then i don't like the self/smalltalk representation idea as is aha allow me to work out my thought a little more.. slate won't follow them in ui, it'll succeed them humans always think in objects intuitively because that's what we're dealing with all day but what if our brains and/or computers are `better' programmed in a non-materialistic envisioned way? no, objects in langs are supposed to be representations (which we do not yet know of) oh geez proglang objects are not supposed to act like material objects i know they map to the domain as representations, not as factual objects but that's the basic thought for it.. no it's the wrong thought starting from material objects the thought extended to non-material objects no no? that's what i always thought it was. proglang objects started from mathematical positivistic theories not from materialism at all ehm the representations they are might derive from your *observations* about material objects, but they do not ever represent the things in themselves oh oh okay, then i wasn't that much off it seems i'll have to write a very controversial user's manula for slate i was scared you were going to say i should read `object' as `goal' intead of `physical entity' -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us924.javanet.com]) wtf is a goal? a target, a goal.. a what? something your heading for oh nevermind let's not even go there lol still i'd like to know more about what can be said about my problem with objects being a concept that stems from human nature where could i read about the basis for OOP ? hm i'd really like to know more about it before saying any more stupid things :/ i have one which addresses the issue from info theory standpoint aha, could i read that? 04:40am i have only mathematical backgrounds and amateuristic informatics theory backgrounds http://www.dmoz.org/Computers/Software/ oops sorry wrong url altogether -:- hcf [nef@me-portland-us129.javanet.com] has joined #tunes ack! the site is gone what site? the one with "Towards a Unified Concept of Information" it was at splendor.unibe.ch the particular path is gone maybe i have a local copy still hm 04:50am water: i'll bbl - could you send the url to memoserv please?