00:00:00 --- log: started forth/03.07.27
02:41:21 --- quit: kc5tja ("[x]chat")
03:24:51 --- join: mur (murr@baana-62-165-186-178.phnet.fi) joined #forth
03:32:45 --- quit: mur (Read error: 104 (Connection reset by peer))
03:33:23 --- join: mur (murr@baana-62-165-186-178.phnet.fi) joined #forth
06:41:49 --- join: supergamerno1 (ybbot@pD9E79258.dip.t-dialin.net) joined #forth
06:51:39 --- quit: supergamerno1 (Read error: 60 (Operation timed out))
06:52:10 --- join: rO|_ (~rO|@pD9EE192C.dip.t-dialin.net) joined #forth
06:52:20 --- join: w1k1_ (~w1k1@pD9EE192C.dip.t-dialin.net) joined #forth
07:00:11 --- quit: rO| (Read error: 60 (Operation timed out))
07:00:11 --- quit: w1k1 (Read error: 60 (Operation timed out))
07:00:13 --- nick: w1k1_ -> w1k1
08:40:32 --- quit: rO|_ (Read error: 60 (Operation timed out))
08:58:23 --- quit: w1k1 (Read error: 110 (Connection timed out))
09:07:08 --- join: rO|_ (~rO|@pD9545995.dip.t-dialin.net) joined #forth
09:09:00 --- join: w1k1 (~w1k1@pD9545995.dip.t-dialin.net) joined #forth
10:21:56 <Herkamire> *#$&%^*@ I just found a bug in my LIT
10:22:01 <Herkamire> I've been using that for months
10:29:47 <Herkamire> numbers between 0x8000 and 0xffff were were coming out negative
10:31:57 <XeF4> you only litted the low 16bits on a ppc?
10:38:59 <Herkamire> on ppc the 16 bit load immediate is signed.
10:39:36 <Herkamire> I was using just that instruction for all (32 bit) values whose top bits were all ones or all zeros.
10:40:59 <Herkamire> the problem is in my code that decides if the number requires a 32 bit lit or a 16 bit lit.
10:41:29 * XeF4 has ESP.
10:42:39 * mur has lost all pointer
10:42:45 <mur> i've lost my goals >;)
10:44:55 <XeF4> eiku yliaistillisuus
10:45:02 <XeF4> =)
10:45:08 <mur> Extraspe..
10:45:09 <mur> nm
10:45:14 * mur was designing symlang
10:46:58 <mur> i got inspiration from cobol :)
10:47:07 <mur> MyName    $
10:47:13 <mur> i left out now the defination words
10:47:25 <mur> it's only special symbols and names
10:47:40 <mur> it was first MyName DEF BYTE *
10:47:42 <mur> or something
10:48:31 <mur> XeF4, Herkamire any ideas how to do jump and conditionals to maintain style  keyword Code Argument ?
10:48:51 <mur> label's code is ":"
10:49:08 <mur> rule 1. unknown keyword must be followed by known keyword
10:49:32 --- quit: w1k1 (Read error: 60 (Operation timed out))
10:52:22 <Herkamire> mur: I don't understand what sort of language you're developing
10:52:57 <Herkamire> for conditionals and loops you generally need some easy way to designate sections of code.
10:53:14 <mur> i will have () as segment
10:56:53 <XeF4> mur: I don't see what you're trying to do, actually
10:57:14 <XeF4> mur: what is code and argument in this case?
10:57:24 <mur> i will want a neat representation for testing value and call function. 
10:57:30 <mur> code is function or defined keyword
10:57:33 <mur> no
10:57:38 <mur> not defined but existing template
10:57:46 <mur> and argument is argument list to be passed
10:58:02 <XeF4> sounds like a kauhea sotku to me..
10:58:14 <mur> hehe
10:59:01 <XeF4> I mean .. I guess the argument could be a symbol or anonymous function, but that syntax would make it totally unintuitive
10:59:13 --- quit: rO|_ (Read error: 110 (Connection timed out))
11:00:40 <mur> ;)
11:00:52 <mur> owner, function, data
11:01:51 <mur> e.g. porche Car, porche paint red
11:03:22 <XeF4> which does what, exactly?
11:21:14 <mur> bah
11:21:15 <mur> nm
11:21:27 <mur> i'll design langauge which only has functions
11:21:30 <mur> :)
11:21:56 <mur> <function> [<arguments>] <fucntion> [<ags>]
11:22:04 <mur> i meant i'll do parser for such
11:22:13 <mur> that is ofcourse, basic :P
11:24:12 <XeF4> then what are the arguments?
11:24:32 <mur> well after function
11:24:40 <XeF4> right, but are the arguments functions too?
11:24:42 <mur> command arguments \n
11:24:51 <mur> not this time 
11:24:57 <mur> i make only simple parser 
11:25:13 <mur> dont bug me with your so good questions i'll lose motvation soon ;)
11:26:16 <mur> i'll code parser with multiple stacks, so i can create multitasking :)
11:28:31 <XeF4> I mean.. have you looked at s-exps?
11:31:42 <mur> no
11:31:58 <mur> or i coudl have brackets inside them
11:31:58 <mur> hmm
11:34:47 <XeF4> so look at s-exps before you go reinventing something badly =)
11:34:57 * XeF4 not a professional motivator btw
11:35:44 <mur> ;)
11:35:52 <mur> what are s-exps even? :)
11:37:01 <XeF4> notation for writing linked lists with syntax ( car cdr ) where car is the first node in the list and cdr is everything else
11:39:19 <XeF4> normally the lists are stored as pairs, so eg (foo 1 2 3) is really (foo . (1 . (2 . (3 . nil)))) where . marks the divide between car and cdr
11:39:51 <XeF4> so (function arguments) is just a list where function is the first node and arguments is everything else
11:40:02 <XeF4> this is how lisp works fwiw
11:40:17 <mur> for what i ?
11:41:17 <XeF4> i?
11:41:45 <mur> fwiw?
11:44:41 <XeF4> for what it's worth
11:48:19 <mur> is it overloading if you redefine == etc operators?
11:50:50 <XeF4> redefine for some situations or redefine permanently?
11:51:36 <mur> either
11:53:18 <XeF4> if you just redefine == so afterwards, it _always_ does the new behaviour, that is not overloading
11:54:04 <XeF4> but to make it compare numbers with some argument types and eg, play music with others, that is overloading
12:05:10 <mur> boah
12:05:17 <mur> i shoudl do very simple language
12:05:32 <mur> what is the minimal set of commands?
12:06:45 <mur> test, set, get, jump?
12:09:17 <XeF4> test,inc,dec,in,out,jump probably
12:09:26 <XeF4> brainfuck is one good example
12:09:32 <mur> yes
12:09:53 <mur> BF didn't have test or did it?
12:10:28 <mur> <vasc> you need at least 3 instructions: Zero, Succ, Jmp
12:10:43 <mur> do you think Zero is JZ?
12:11:20 <mur> successor  is memory[successor] ?
12:11:43 <XeF4> wouldn't know what he means
12:12:03 <XeF4> bf had test-for-zero
12:12:04 <mur> succ(0) = 1, Succ(succ(0)) = 2
12:12:46 <mur> in and out are put to memory and take from memory, right?
12:17:18 <XeF4> in=read from device to memory
12:17:25 <mur> hmm
12:17:28 <XeF4> out=write memory contents to device
12:17:30 <mur> did Bf has that?
12:17:33 <XeF4> yes
12:17:41 <mur> clever
12:17:48 <XeF4> it had putchar and getchar
13:03:37 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth
13:10:11 --- join: GnuVince (~vince@modemcable158.188-200-24.mtl.mc.videotron.ca) joined #forth
15:50:16 --- join: jstahuman (~justahuma@pcp053555pcs.brlngt01.nj.comcast.net) joined #forth
15:52:44 --- quit: mur (Read error: 110 (Connection timed out))
15:57:36 --- quit: Herkamire ("seeing Johny English")
16:44:48 --- join: njd (~junk@njd.paradise.net.nz) joined #forth
16:49:29 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth
16:49:30 --- mode: ChanServ set +o kc5tja
16:54:08 <Robert> Hi :)
16:54:19 <kc5tja> re
16:54:26 * kc5tja is catching up on Ozy & Millie comics.
17:01:54 <kc5tja> Doggone it, now the Ozy & Millie website isn't responding.  >:/
17:02:04 <kc5tja> And I'm only half-way through their archives.
17:34:25 --- quit: njd ("[BX] I theenk I need a beeger box!")
17:44:59 --- join: gilbertdeb (nyue@fl-nken-u2-c3b-45.miamfl.adelphia.net) joined #forth
17:47:18 <kc5tja> re
17:47:32 <gilbertdeb> hello
17:47:46 <gilbertdeb> I was just looking at withGlee
17:48:52 <kc5tja> What is that?
17:49:02 <Robert> Hi gilbertdeb
17:49:04 <gilbertdeb> YAAD.
17:49:06 <gilbertdeb> hi Robert 
17:49:12 <gilbertdeb> Yet Another APL Derivative
17:49:59 <gilbertdeb> whats news in #forth ?
17:50:41 <Robert> 03:03:52 < I440r> i HATE c
17:50:48 <Robert> I think I440r hates C.
17:50:58 <Robert> But that's not really news, is it?
17:51:05 <onetom> use squeak then ;p
17:51:10 * onetom hides away ;)
17:51:16 <gilbertdeb> :)
17:51:22 <gilbertdeb> hey onetom playing with squeak now?
17:51:23 <kc5tja> Nothing wrong with APL.  Klingons love it, for example.  :)
17:51:34 <kc5tja> Actually, I like the concept of an array-based language.
17:51:44 <onetom> im playing w squeak again in the last few days
17:51:53 * kc5tja is reading Ozy and Millie.
17:52:04 <gilbertdeb> whats Ozy and millie?
17:52:10 <onetom> i loooove it. now i know the basics of smalltalk, so i also understand the basics
17:52:31 <gilbertdeb> onetom, it is a pretty interesting language.
17:52:32 <onetom> but u know what i miss from it?..
17:52:36 <gilbertdeb> what?
17:52:39 <gilbertdeb> rpn?
17:52:44 <onetom> colorforth ;p~~~
17:52:57 <gilbertdeb> you miss it? why?
17:53:02 <ianni> cool
17:53:10 <onetom> coz smalltalk is pretty chatty
17:53:13 <ianni> ObjC is nice...
17:53:32 <onetom> for a lot of tasks i would prefer our beloved colortext representation
17:53:56 <onetom> probably we could even design a colored smalltalk..
17:54:23 <gilbertdeb> but if its just a matter of color, vim can take care of that no?
17:54:39 <onetom> anyway, im just dreamin now. i havent practiced smalltalk a lot yet
17:54:57 <gilbertdeb> I hate it quite a bit.
17:54:58 <onetom> nooo, coloredtext is not just about coloring
17:55:15 <onetom> its about consciseness
17:55:17 <gilbertdeb> I bought the yellow book for squeak, and it is the WORST book I've ever read.
17:55:27 <onetom> its about easy interpretation
17:56:01 <gilbertdeb> for who? 
17:56:11 <onetom> for the computer
17:56:27 <onetom> & for ppl, tho its debatable
17:56:29 <gilbertdeb> why make the computers job easier?
17:56:41 <onetom> for speed
17:56:44 <gilbertdeb> intel churns out gigabyte after gigabyte to make our work easier :D
17:56:46 <onetom> nothing else
17:57:24 <gilbertdeb> onetom, programs should really be written for people first, then computers second.
17:58:06 <gilbertdeb> I read an article from the guy who wrote THE in which he mentioned a forth project he was involved in.
17:58:28 <gilbertdeb> he made a lot of noise (pro) about the readability of the forth code they generated.
17:58:36 <onetom> THE? the editor?
17:58:44 <gilbertdeb> not the. but THE
17:58:49 <gilbertdeb> THE Humane Interface.
17:58:54 <onetom> ah
17:59:17 <onetom> okay, most ppl cant step over their thinking habits
17:59:17 <gilbertdeb> lemme fish it out incase you're not interested ;)
17:59:22 <onetom> thats not a huge problem
17:59:34 <onetom> but for those who can, its more readable
17:59:50 <onetom> im absolutely interested
17:59:52 <ianni> where can i find info about THE
17:59:56 <ianni> hard to google, for obvious reasons
18:00:29 <gilbertdeb> hehehe. ask kc5tja.
18:00:36 <onetom> yeah :)
18:00:37 <gilbertdeb> he's the THE expert around here.
18:00:42 <ianni> cool
18:03:51 <tathi> ianni: humane.sourceforge.net
18:04:14 <ianni> sweet!
18:04:29 <ianni> Ive read half the book now...
18:04:52 <tathi> yah, I read it recently.  Not bad...
18:05:12 <ianni> sweet, it's mac, too
18:05:19 <ianni> goes to figure
18:05:32 <ianni> yeah fairly insightful
18:05:44 <gilbertdeb> http://www.acmqueue.org/modules.php?name=Content&pa=showpage&pid=22
18:05:49 <gilbertdeb> that took long enough to fish out!
18:06:11 <kc5tja> What's this?
18:06:19 <gilbertdeb> the woes of IDEs.
18:06:27 <gilbertdeb> forth gets mad props :)
18:06:29 <ianni> heh my eyes went right to that
18:06:33 <ianni> oddly enough
18:06:37 <gilbertdeb> or rather, good forth coding gets mad props in the article.
18:06:44 <ianni> i love my java ide...
18:06:53 <ianni> refactoring is a cool timesaver
18:07:16 <kc5tja> What are "mad props?"
18:07:38 <ianni> "many compliments"
18:07:43 <ianni> lol
18:07:46 <kc5tja> hmmm
18:07:49 <gilbertdeb> some new fangled term i picked up on mtv(?) i think.
18:07:52 * kc5tja will ahve to read this.  :)
18:08:48 <ianni>  it's a street term.
18:09:03 <gilbertdeb> is it?
18:09:11 <ianni> mad is slang for 'a lot', props is like, respect
18:09:16 <ianni> yes
18:09:25 <gilbertdeb> a lot of respect!
18:09:28 <gilbertdeb> that makes sense.
18:09:30 <ianni> *nods*
18:09:42 <gilbertdeb> mad props it is.
18:09:42 <ianni> so does mad props :)
18:09:54 <ianni> if u talk the talk
18:10:13 <gilbertdeb> don't worry: I can walk the walk in my dreams.
18:11:04 <ianni> that's all that matters 
18:11:37 <kc5tja> This new-fangled Americese is all pants.
18:11:55 <gilbertdeb> heh?
18:12:06 <gilbertdeb> 'all pants'? now whats that?
18:12:27 <kc5tja> British derogatory slang.
18:12:43 <kc5tja> Used in this context, it means, "Full Of Shit."  :)
18:13:00 <gilbertdeb> heheheh
18:13:01 <Robert> So kc5tja is British now? Yay.
18:13:43 <kc5tja> Nope -- just a lone American who happens to take great pride in the language he's been taught, and who cringes every time I see the very foundations of our society and our ability to communicate eroded away by crass commercialism.
18:14:25 <Robert> Commie.
18:15:07 <ianni> kc5tja: hey, it's just a language :)
18:15:12 <ianni> it's not american
18:15:19 <ianni> its the streets
18:15:25 <ianni> i guess it starts here
18:15:27 <gilbertdeb> crass commercialism? but commericialism is GOOD for the economy!
18:15:51 <Robert> Of course. Consumtion is always good, right? Right?
18:15:54 <ianni> commercialism aint got nothin to do with it
18:15:57 <gilbertdeb> right!
18:16:00 <ianni> any more than...
18:16:05 <kc5tja> ianni: American English is defined by street talk, yo.
18:16:30 <kc5tja> It's like Cockney English in Britain -- not even their next-door neighbors understand half the things they say.  :)
18:16:31 <ianni> nahh, cos you got words that are 'chiefly british'
18:16:33 <Robert> Swedish is defined by 12 persons sitting in a room.
18:16:37 <ianni> and like spelling differences
18:16:40 <ianni> ha.
18:16:43 <ianni> how does that work Robert 
18:16:51 <kc5tja> Robert: French is like that too.
18:16:52 <ianni> yeah, it is rather
18:16:53 <Robert> ianni: They...define words.
18:16:56 <Robert> And guess what?
18:17:03 <Robert> They're working on a swedish dictionary.
18:17:06 <ianni> ah, yeah, the academie française
18:17:08 <Robert> They're on T now, I beleive.
18:17:12 <ianni> nice
18:17:15 <gilbertdeb> Robert, whats taking them so long?
18:17:23 <gilbertdeb> is Swedish going to be programmable?
18:17:26 <Robert> The nice thing - they started like.. 50 yerars ago or so.
18:17:28 <Robert> Maybe longer.
18:17:59 <gilbertdeb> so currently there is no Swedish dictionary?
18:18:16 <gilbertdeb> what if you want to look words up? you get a Finnish dictionary?
18:18:19 <Robert> There are several, but not any official, containing all swedish words.
18:18:31 <gilbertdeb> ALL? thats rather ambitious!
18:18:39 <gilbertdeb> guess why they've been at it for 50 years.
18:18:45 <Robert> Well, they have a LIST of all these words.
18:18:55 <Robert> But they haven't described them all, yet. :)
18:19:08 <gilbertdeb> tell them to hurry.
18:19:15 <gilbertdeb> what if the die? then the process has to start over?
18:19:17 --- join: hefner (hefner@pool-151-196-177-226.balt.east.verizon.net) joined #forth
18:19:18 <Robert> I don't think they will.
18:19:21 <Robert> gilbertdeb: :D
18:19:23 <Robert> Right.
18:19:46 <gilbertdeb> at this rate, you'll never have a dictionary containing ALL words.
18:20:06 <Robert> I'm not holding my breath waiting for it, really.
18:20:16 <gilbertdeb> apparently neither are the authors.
18:20:23 <gilbertdeb> thats why they're taking their sweet time ;)
18:20:41 <Robert> :)
18:23:46 <kc5tja> Even maintaining the English dictionary takes many many years, and that is with hundreds of thousands of people, each hacking away at modern-day computers.
18:23:53 <kc5tja> I'm talking the unabridged dictionary.
18:24:00 <kc5tja> And it still doesn't contain every word.
18:24:13 <kc5tja> Still, when you have a simple, orthogonal language, the task becomes much easier.
18:24:13 <gilbertdeb> the full ODE? 
18:24:18 <kc5tja> I'm not familiar with Swedish.
18:24:36 <gilbertdeb> they could ask all the Swedes of the world to not use words other than those they've currently defined.
18:24:44 <kc5tja> But if it's phoenetic, then the number of words to record will drop remarkably, because the number of like-sounding-but-wildly-different-sounding words will drop.
18:25:06 <gilbertdeb> you mean like 'wu li'?
18:25:13 <kc5tja> If the language is agglutinative, then the rules for accumulating affixes can be documented, and then applied to the overwhelming majority, if not the entirety, of the language's core vocabulary.
18:25:46 <gilbertdeb> I think swedish might be agglu@#DASFA@@adfaative.
18:26:16 <gilbertdeb> Robert, is swedish agglu@#@@#$!!~!!ative?
18:26:26 <kc5tja> If you combine perfect phoenetics with perfect agglutination (or very close to it), you get a language like Esperanto, which can be taught to 3-year olds and 30-year olds alike in as little as 6 months.
18:26:38 <gilbertdeb> with fluency?
18:26:50 <kc5tja> Fluency comes with practice, as with any language.
18:27:02 <kc5tja> But I challenge you to teach anyone all the core rules of English in only 6 months.
18:27:04 <gilbertdeb> people can be forced to learn it if ALL programming languages used esperanto as well as their docs.
18:27:22 <gilbertdeb> kc5tja, I challenge you to teach an english prof that.
18:27:43 <gilbertdeb> they'll go as bonkers as pirsig went over quality in zen and the art of motorcycle maintenance.
18:27:50 <kc5tja> gilbertdeb: English professors are in college no less than 16 years to get their professorhood.  That alone is proof of English's complexity.
18:28:00 <gilbertdeb> heheh.
18:28:17 <gilbertdeb> sixfreakenteenyears. how fulfilling a life?
18:28:21 <Robert> You mean English has core rules? 
18:28:39 <kc5tja> Robert: There are so many that there might as well not be any.
18:28:42 <Robert> btw, don't English professors do other things than just learning words? ;)
18:28:43 <gilbertdeb> Robert, Yes. You can't say 'mad props' to an english prof for instance :)
18:28:46 <kc5tja> I before E, except after C.
18:28:48 <gilbertdeb> he'd thing something else entirely.
18:28:51 <kc5tja> But only in certain cases.
18:28:56 <Robert> :D
18:29:09 <Robert> Exceptions to exceptions are nice.
18:29:15 <kc5tja> Aren't they?
18:29:17 <gilbertdeb> Robert, they learn vowels and dream about living the ultimate life: The character in My fair lady :)
18:29:32 <Robert> I and a friend counted the number of ways you could spell a "ch/sh"-like sound in Swedish.
18:29:37 <kc5tja> The letter 'a' has something like 10 or 12 different pronunciations in English, IIRC.
18:29:45 <Robert> I think we came up with about 10 different ways.
18:30:31 <gilbertdeb> kc5tja, the alphabets are a poor way of describing the english letters.
18:30:33 <kc5tja> The only real way to really learn the introcasies of English is by rote memorization.
18:30:34 <Robert> t, tj, stj, sj, sk, skj, ch, sch, j, rs, etc. etc.
18:30:49 <gilbertdeb> kc5tja, I agree.
18:30:56 <kc5tja> gilbertdeb: For English, this is true.  This is patently false for *most* of the human-spoken languages, though
18:31:19 <gilbertdeb> you mean like hawaiian?
18:31:33 <kc5tja> I don't know anything about Hawaiian except that it lacks consonants.  :)
18:31:34 <gilbertdeb> they dropped a bunch of their bowels somewhere in the pacific.
18:31:47 <gilbertdeb> eh. did i say bowels?
18:31:55 <tathi> rofl
18:31:57 <kc5tja> Yes,
18:32:14 <kc5tja> And yes, they did drop their bowels somewhere in the pacific, long before toilets showed up there.  :)
18:32:24 <gilbertdeb> lol.
18:32:36 <kc5tja> But Hawaiian is, with few exceptions, pretty much an all-vowel language.
18:32:43 <tathi> ah, English isn't that bad.  Most of the exceptions are obviously borrowed from other languages.
18:33:10 <gilbertdeb> true. the language had whorish tendencies.
18:33:18 <gilbertdeb> still has whorish tendencies.
18:34:12 --- join: TheBlueWizard (TheBlueWiz@207.111.96.155) joined #forth
18:34:12 --- mode: ChanServ set +o TheBlueWizard
18:34:18 <TheBlueWizard> hiya all
18:34:49 <gilbertdeb> hi tbw
18:34:52 <kc5tja> tathi: Not precisely true, for the core of English itself first came into being amongst illiterate tribes -- a history quite unlike the otherwise quite literate Romans and Greeks.  
18:35:29 <kc5tja> re TheBlueWizard
18:35:35 <TheBlueWizard> hiya gilbertdeb
18:35:39 <TheBlueWizard> hiya kc5tja
18:36:46 <Robert> Hi,
18:38:08 <TheBlueWizard> hiya Robert
18:48:21 * kc5tja is catching up on Ozy and Millie still.
18:48:35 <gilbertdeb> whats ozy and millie? 
18:48:41 <gilbertdeb> something from mills and boon?
18:49:08 <TheBlueWizard> kc5tja: so you're an Ozy and Millie fan? I am too :)
18:49:29 * TheBlueWizard points gilbertdeb towards http://www.ozyandmillie.org
18:50:12 --- join: I440r (~x@dialup-67.29.205.119.Dial1.Cincinnati1.Level3.net) joined #forth
18:51:07 <gilbertdeb> hehe. thats a funny one on the first page.
18:51:48 <TheBlueWizard> hiya I440r! :)
18:52:03 <I440r> hi :)
18:53:32 <TheBlueWizard> gilbertdeb: I consider O&M one of the better online cartoons around...it has some of the flavors taken from Calvin and Hobbes, and from one other cartoon I forgot at the moment...anyway, pretty good overall
18:54:35 <kc5tja> Yes, I agree.
18:54:56 <kc5tja> I do wish that the author would strike out on more intellectual and imaginative fantasy though.
18:55:20 <I440r> anyone here code 8051 ?
18:55:30 * kc5tja has always heard of it, at least since a couple of years.  But I've never gotten around to actually reading it until now.
18:56:07 <kc5tja> I440r: Hey :)  No.
18:59:08 <TheBlueWizard> I never have progged 8051....so, nope
18:59:39 <gilbertdeb> why 8051? for fun?
19:00:00 --- quit: I440r (Read error: 113 (No route to host))
19:00:13 <TheBlueWizard> kc5tja: I rather let the author of Ozy and Millie be as is...he's doing just dandy...just getting used to, that is all :)
19:01:59 <gilbertdeb> Robert is the great swedish dictionary going to contain this word beginning with H ;) ? http://www.ozyandmillie.org/2003/om20030722.html
19:07:45 --- quit: skylan (Read error: 104 (Connection reset by peer))
19:08:28 * kc5tja hasn't gotten there yet.
19:08:35 <kc5tja> I'm still in February of 2001.  :)
19:08:43 <TheBlueWizard> kc5tja: how far have you been in O&M archive?
19:08:43 <gilbertdeb> heheh/
19:08:56 <TheBlueWizard> oh...that answers my question :)
19:25:45 <kc5tja> Yeah, I started from the beginning
19:25:59 <kc5tja> I'm in March 2001 now.  :)
19:26:27 <TheBlueWizard> cool...I take it you're enjoying it, hmm?
19:27:15 <kc5tja> Yeah
19:27:25 <kc5tja> But I'm a slow reader, and the sluggishness of their servers don't help much.
19:27:41 <kc5tja> And I cannot fathom why they use progressive-scan images.  That is such a sore waste of network bandwidth.
19:27:52 <kc5tja> Then they have the nerve to say, "Modem users -- please be patient."
19:27:59 <kc5tja> (well, they don't, but some other comic sites do)
19:30:37 <TheBlueWizard> I haven't seen such blurb from a comic site, but, yeah, I feel your pain...I even once got a prompt saying "your computer is so fscking slow buy a new faster PC" ... grrr
19:33:04 * kc5tja nods
19:33:20 <kc5tja> Total contempt on the part of the site maintainer.
19:33:57 <kc5tja> You know, one day while talking about how the Amiga was in some ways better than Linux in the #hamradio channel, I was actually commanded to "continue to use your stupid Amiga, and live in the past."
19:34:01 <kc5tja> What morons.
19:34:17 * TheBlueWizard nods
19:34:35 <kc5tja> I simply retaliated with, "I will, but only if you promise you'll NEVER complain about configuration or usability problems EVER AGAIN, lest you be kick-banned from this channel."
19:34:38 <kc5tja> I've never seen him since.
19:34:41 <TheBlueWizard> elitism syndrome...hate that
19:34:52 <TheBlueWizard> heh
19:38:45 <jstahuman> kc5tja: is there a place on the web that has Amiga OS documentation to the nittty-gritty?
19:40:16 <TheBlueWizard> jstahuman: AFAIK no
19:41:07 <jstahuman> how many people wrote the thing anyway? and what languages were used?
19:41:14 --- join: Herkamire (~jason@h0030657bb518.ne.client2.attbi.com) joined #forth
19:42:50 <TheBlueWizard> hiya hefner
19:43:20 <TheBlueWizard> oops...I meant to say hiya Herkamire :P
19:44:11 <gilbertdeb> TheBlueWizard, oh the woes of tab completion!
19:44:24 <gilbertdeb> why don't people have easily tab completeable names _anyway_ ? ;)
19:45:22 <gilbertdeb> jstahuman I suspect it was written in M68K asm or C.
19:45:36 <gilbertdeb> what else might they have used?? Bliss?
19:46:06 <TheBlueWizard> jstahuman: in late 80s and early 90s, there were a fair number of people developing AmigaOS...then Commodore went bankrupt in '94, and passed through various hands and now ends up in Amiga Inc.'s hands, IP wise.  The present line (3.x) is developed by Haafe & Partners as a contract agreement with Amiga Inc.
19:46:50 <Herkamire> hi TheBlueWizard  :)
19:48:02 <TheBlueWizard> yes, the earlier ones are written in a mix of M68K and C (and I believe the AmigaDOS portion used to be written in BCPL, hence the weird pointer encoding requirements when calling AmigaDOS API)...nowadays, PPC and C
19:48:29 <TheBlueWizard> I believe what I stated above is reasonably accurate.
19:48:42 <TheBlueWizard> helpful so far?
19:48:51 <gilbertdeb> bcpl? 
19:49:17 <gilbertdeb> I shouldn't have asked 'what else could they have used' should I?
19:50:58 <TheBlueWizard> BCPL is an earlier language...ancestor of B language, and in turn the ancestor of that oh-so-famous C language
19:51:26 <gilbertdeb> yes I've heard of BCPL.
19:51:39 <gilbertdeb> the author is still alive and well in england I think.
19:53:56 <TheBlueWizard> really? interesting
19:55:49 <gilbertdeb> TBW: http://www.cl.cam.ac.uk/users/mr/
19:56:37 <kc5tja> jstahuman: Not to my knowledge.
19:57:27 <kc5tja> That's a pity.  We'll have to do something about that, now won't we?
19:57:48 <kc5tja> Why is it that language designers seemed to go out of their way to make things so ruthlessly complicated way back then?
19:58:09 <gilbertdeb> kc5tja, way back the? like in the time of forth :)
19:59:16 * Herkamire reads the #forth log for today
20:00:21 <Herkamire> with reguard to colorforth and gilbertbsd's question "why make the computers job easier?":
20:00:42 <Herkamire> when you make the computer's job easier, you make your software simpler.
20:00:48 <gilbertdeb> is that right?
20:00:49 <TheBlueWizard> I believe back then the designing of a language was, um, rather unscientifical (for lack of better wording), hence Fortran, COBOL, and such...ahhh, the dark age :)
20:01:07 <gilbertdeb> TBW but lisp came out rather quickly!
20:01:13 <Herkamire> simpler software is easier to maintain, smaller, faster, and generally more usefull.
20:01:24 <gilbertdeb> I wo'nt mention IPL because I hear it was never really compiled or interpreted on a real machine (just on paper)
20:01:28 <Herkamire> and less buggy of course.
20:02:43 <Herkamire> I've been experimenting with colorforth for a few months now, and I keep coming up with wonderful human interface features and optimizations that will be very easy to do because of my (simple) colorforth core.
20:03:11 <gilbertdeb> herkamire have you ever used Acme from plan9? 
20:03:20 <Herkamire> gilbertdeb: no
20:03:29 <Herkamire> haven't tried plan9
20:03:32 <Herkamire> read a bit about it.
20:03:50 <Herkamire> looked interesting if I had a network of x86 machines.
20:03:53 <Herkamire> but I don't have any
20:03:56 <gilbertdeb> aww, burn the iso and run the demo. give it a spin. the iso.bz2 is only 65meb.
20:04:19 <Herkamire> does the demo run without touching the HD?
20:04:46 <Herkamire> I don't have an x86 box that I can screw with, but if it's just popping in a CD I can try it on somebody elses box.
20:04:59 <Herkamire> what's acme?
20:05:13 <gilbertdeb> herkamire yes it doesn't mess with the harddisk.
20:05:28 <gilbertdeb> acme is the plan9 'official' text editor.
20:05:39 <gilbertdeb> there is a linux clone called wily ...
20:05:52 <gilbertdeb> I haven't heard too many raves of it but I rather like the way it behaves.
20:06:38 <Herkamire> can't I just try wily then? there's an ebuild (install/build script) for it and everything.
20:06:59 <gilbertdeb> yes. but the Edit part of it does not work.
20:07:26 <Herkamire> humph
20:07:30 <gilbertdeb> but I can tell you what Edit is... when you type it in, it lets you use ed (especially SAM) regexps and commands in the document.
20:08:00 --- quit: tathi ("leaving")
20:08:31 <Herkamire> you type the commands and stuff and it goes into your document? then you can somehow execute the current line or something?
20:08:48 <gilbertdeb> I was reading about THE earlier, and it reminds me most about acme ... just wanted to see what others thought.
20:08:52 <gilbertdeb> herkamire correct.
20:09:03 <Herkamire> gilbertdeb: so?
20:09:13 <Herkamire> emacs does that I believe.
20:09:57 <gilbertdeb> thats not the only thing ofcourse.
20:10:01 <kc5tja> re: Herkamire's comment -- when you make the computer's job easier, you make your own job easier.
20:10:02 <Herkamire> I intend to read THE soon. I'm borrowing a copy in a couple weeks.
20:10:15 <Herkamire> kc5tja: well put :)
20:11:01 <Herkamire> I consider it my job as a programmer to make the computer simpler and more useful
20:11:27 <gilbertdeb> Herkamire, make the computer simpler how?
20:11:36 <gilbertdeb> you mean the interaction with the computer simpler?
20:11:57 <TheBlueWizard> I must go...bye all!
20:12:01 <kc5tja> Herkamire: Thanks.  :)
20:12:08 <Herkamire> I find the interfaces to the OS to be too complex and restrictive to simplyfy without restricting it further, so I've decided to make my own OS
20:12:10 <kc5tja> TheBlueWizard: See you later.
20:12:26 <gilbertdeb> Herkamire, do check out plan9 to that end. 
20:12:26 <Herkamire> TheBlueWizard: bye :)
20:12:26 * kc5tja finds current OSes to be sorely missing the point.
20:12:32 <TheBlueWizard> bye kc5tja
20:12:43 --- part: TheBlueWizard left #forth
20:12:59 <kc5tja> All these complex APIs -- what purpose do they serve?  Advanced graphics, advanced sound, advanced media -- all to satisfy our ideal need for a Star Trek UI.
20:13:29 <kc5tja> I don't want a Star Trek UI.  I want to get real world work done.
20:13:34 <Herkamire> kc5tja: I agree. I could be wrong, as I haven't gotten into drivers yet, but I can't imagine that it's any harder to write an OS than to interface properly with linux, Windoze or mac.
20:13:45 <gilbertdeb> kc5tja, I've been thinking that with a simpler, quieter back end machine, things _might_ be simpler!
20:14:26 <Herkamire> I'd love to see simpler hardware, but for now I'm working with what I've got. hopefully I can make something simple and usefull out of the hardware I have
20:14:47 <gilbertdeb> Herkamire, what would you like to use your machine for?
20:14:54 <gilbertdeb> as an entertainment system?
20:15:08 <Herkamire> kc5tja: I don't see what those APIs do... game companies make their own graphics and physics engines, and their interactions with the OS is mostly just trying to get around it.
20:15:26 <Herkamire> gilbertdeb: work station.
20:15:37 <gilbertdeb> hmmm. plan9 again :)
20:15:49 <Herkamire> I want to create usefull software and fascinating software (fractals, graphics etc)
20:16:08 <gilbertdeb> herkamire now I"m foaming at the mouth... go get plan9 :)
20:16:14 <Herkamire> I want to make at least one good game too.
20:16:25 <Herkamire> gilbertdeb: I don't have an x86 box
20:16:29 <gilbertdeb> I do hope you find a machine it likes. it is as picky as picky can be.
20:19:16 <Herkamire> I liked some of the suff I read about plan9. it still hase a windowed environment which I view as a mistake.
20:20:33 <gilbertdeb> they have reasons for the gui.
20:21:53 <Herkamire> I have a gui too, but I don't like partially overlapping windows. I see no use for that. I say either a window is fully visible so you can use it, or it should be squirreled away some where so you can call it forth when you want it.
20:25:41 <gilbertdeb> plan9's window system is really quite different.
20:25:55 * kc5tja doesn't mind the GUI.
20:26:02 <gilbertdeb> it used to be 8.5 but the current one is rio.
20:26:03 <kc5tja> But I'm a fan of full-screen application spaces.
20:26:20 <Herkamire> kc5tja: me too
20:26:26 <kc5tja> My Linux window manager, for example, is Ion -- I haven't resized a window in about a year now.  :)
20:26:32 <Herkamire> I use ion as my window manager.
20:26:49 <kc5tja> WOW!  Someone else who's heard of ION!!!!
20:26:56 * kc5tja falls over in shock
20:27:05 <gilbertdeb> I've heard of ion and I think speuler uses it all the time.
20:27:06 <Herkamire> more than heard of it :) I use it at home and at work
20:27:16 <gilbertdeb> but I don't use ion. I prefer twm.
20:27:17 <gilbertdeb> or kde.
20:27:20 <Herkamire> tathi uses it too (exclusively as far as I know)
20:27:20 <gilbertdeb> either extreme.
20:27:44 <Herkamire> the only programs I use that don't work well with it are The Gimp and sodipodi
20:28:00 <gilbertdeb>  sodipodi
20:28:02 <gilbertdeb> ?
20:28:02 <kc5tja> I don't get how twm can be called the "tabbed" window manager -- all the screen shots of, and indeed my own experiments with, twm clearly show windows.
20:28:13 <gilbertdeb> I know better than to ask.
20:28:17 <gilbertdeb> I'll just google.
20:28:24 * kc5tja uses gimp with Ion quite well.
20:28:35 <kc5tja> It requires a split pane for any serious work though.
20:28:54 <kc5tja> It actually kind of reminds me of Deluxe Paint for Amiga...  :)
20:31:05 <Herkamire> kc5tja: yeah, it's not too bad for the gimp and sodipodi, just a little annoying.
20:31:34 <Herkamire> once I've got the windows in the right pane it's ok. it's no worse then moving the windows around in an ordinary window manager
20:32:14 <Herkamire> sodipodi is a very nice little vector editor (like Adobe Illustrator)
20:32:33 <Herkamire> it saves/edits SVG files, and will render at any resolution to PNG
20:35:59 * kc5tja nods
20:36:05 <kc5tja> hmmm....
20:36:09 * kc5tja never heard of it before.
20:36:12 <kc5tja> I'll have to check it out.
20:36:22 <kc5tja> Can it be used for things like simple desktop publishing?
20:36:35 * kc5tja would like something better than xfig for his advertising material.
20:37:14 <Herkamire> kc5tja: yeah. I hear you can convert svg to ps with some java program.
20:37:30 <Herkamire> I did my logo and business card in sodipodi.
20:38:06 <Herkamire> I haven't done more than a single word in text, so I don't know how good it is for blocks of text and the like. it's very nice for creating graphics though.
20:41:01 <kc5tja> hmmm
20:41:04 * kc5tja will have to play with it.
20:44:54 --- join: njd (~junk@njd.paradise.net.nz) joined #forth
20:48:33 <Herkamire> plan9 intro: "Unlike X windows, where a new application typically creates a new window to run in, an 8½ graphics application usually runs in the window where it starts."
20:48:48 <Herkamire> I like that. it's like the text based programs I like so much.
20:50:01 <Herkamire> I use terminal-based (usually curses) programs for everything I can. I think this is because they tend to have the sort of interfaces I like :)
20:52:18 <jstahuman> i like terminal based apps because virtual terminals make for a great window manager
20:52:56 <jstahuman> plus the fonts are about the right size
20:53:09 <jstahuman> hehe
20:53:23 * kc5tja nods
20:53:59 <kc5tja> Well, I don't know about the font size; I've often greatly desired 128-character wide lines when programming.  I *hate* having to line-break long lines.
20:54:14 <Herkamire> jstahuman: yeah :) I use xterm and ion, but very simularly to the terminal.
20:54:29 <jstahuman> when will XFree86 have good font config by default. I am so lazy to fix X once it's installed, so I opt for the defaults that make very annoyed
20:54:52 <kc5tja> What I find amusing is even when you have things configured correctly, it's still not guaranteed to work.
20:55:11 <kc5tja> I can display Arial and other true-type fonts perfectly, except Times New Roman.  For whatever reason, it has a propensity to segfault.
20:55:13 <Herkamire> yeah, I have some gtk2 programs, and the font is even smaler. like hard to read. (and I have good vision)
20:55:23 <kc5tja> (not all of X, but the application using it)
20:55:47 <jstahuman> that's one thing MS did very well
20:55:49 <jstahuman> fonts
20:55:59 <Herkamire> jstahuman: when?
20:56:06 <jstahuman> way back when
20:56:18 <jstahuman> font support i mean
20:56:54 <Herkamire> I presume you're not compairing ms with apple
20:57:15 <jstahuman> no, even 3.1 had nice font support
20:58:48 <jstahuman> i don't remember myself being worried about fonts not working properly 3.1 and up
20:59:30 * kc5tja nods
20:59:54 <kc5tja> Although, I find AmigaOS' font management even better, in that AmigaOS didn't have to run the whole damn registery of fonts every time it booted.
21:00:05 <kc5tja> That's the one thing that 3.1 did so wrong -- the more fonts you had, the longer it took to boot.
21:00:06 <jstahuman> and i don't know about 3.0, since I didn't have a windows computer at 3.0 time
21:00:19 <kc5tja> 3.0 and 3.1 were the same, except 3.1 had TrueType font support.
21:00:40 <jstahuman> and I never had an amiga
21:01:11 <gilbertdeb> kc5tja, the more fonts you have in *nix, the longer it takes to boot.
21:01:43 <gilbertdeb> I found that not very amusing after moving all the fonts I had from an old WP2000 cd to my fonts directory under linux.
21:01:46 <jstahuman> i went from dos to win 3.1 to struggle with 95 and 4 meg of ram, to linux
21:02:49 <jstahuman> and BSDs, since BSDs are easier to manage
21:03:00 <gilbertdeb> freebsd rocks!
21:03:05 <jstahuman> but I can't wait for the next big thing
21:03:14 <gilbertdeb> I should re-install that and switch back to ~bsd.
21:03:45 <gilbertdeb> Herkamire, I knew you'd like something about plan9. I just wasn't sure what.
21:03:48 <jstahuman> gilbertdeb: it could be better, like anything, but it beats a linux distro.
21:04:01 <gilbertdeb> yep. especially the man pages.
21:04:13 <gilbertdeb> the only better man pages I've seen are for solaris.
21:04:19 <gilbertdeb> the worst were for IRIX. 
21:04:23 <gilbertdeb> pits.
21:04:26 <jstahuman> gilbertdeb: have you tried openbsd?
21:04:36 <jstahuman> gilbertdeb: try open/net
21:04:37 <gilbertdeb> I have an unbootable cd here.
21:04:53 <gilbertdeb> and I did install openbsd ones on a MAC se/30.
21:05:14 <jstahuman> gilbertdeb: good thing about openbsd is they put more emphasis on quality rather than conformance. You won't see something like PAM in openbsd.
21:05:26 <gilbertdeb> I left the machine open and a winNT4.0 book I had fell from my bookshelf and broke the monitor's neck :(
21:06:37 <gilbertdeb> it was a sureal experience. I've been scarred since.
21:07:37 <Herkamire> gilbertdeb: that's really funny.
21:08:00 <jstahuman> did you see any bill gates daemons?
21:08:10 <gilbertdeb> :)
21:08:24 <Herkamire> gilbertdeb: I like a bunch of things about plan9. I'm just not sure what I can get from it. It seems to be trying to solve a problem that I don't have.
21:08:31 <Herkamire> I don't have a network of computers or multiple users.
21:08:53 <gilbertdeb> Herkamire, you can run it on a single computer as well.
21:09:00 <gilbertdeb> ... as a single user. thats what I do.
21:09:01 <jstahuman> plan9 was meant for workgroups
21:09:14 <gilbertdeb> yes it was.
21:09:18 <jstahuman> a typical corporate or research setting
21:09:23 <gilbertdeb> in every sense of the word 'group'.
21:10:31 <jstahuman> unix tried, but couldn't
21:10:41 <jstahuman> NFS is a nightmare for instance
21:10:53 <jstahuman> and so was NIS
21:10:58 <gilbertdeb> yes. try switching off one of the shares...
21:10:58 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth
21:11:17 <gilbertdeb> jstahuman, but unix follows the worse is better edict.
21:11:21 <jstahuman> gilbertdeb: well, it's more than that. it's just buggy.
21:11:37 <gilbertdeb> hehehe
21:11:41 <jstahuman> gilbertdeb: depends who you ask
21:11:57 <gilbertdeb> perhaps the thing to do is to create great software, throw in some bugs and release it.
21:12:08 <gilbertdeb> MS has done very well doing that. Unix has lived long too :)
21:12:13 <kc5tja> gilbertdeb: Not true.
21:12:20 <gilbertdeb> oh?
21:12:24 <kc5tja> gilbertdeb: Where'd you get that idea regarding fonts in Unix?
21:12:25 <gilbertdeb> so worse is not better?
21:12:32 <jstahuman> gilbertdeb: unix provides enough primitives to write good software, and those primitives that are broken are easy enough to fix, it's just politics get in the way.
21:12:38 <gilbertdeb> fonts? I don't think much about fonts at all.
21:12:43 <gilbertdeb> that wasn't me about fonts kc.
21:13:03 <kc5tja> <gilbertdeb> kc5tja, the more fonts you have in *nix, the longer it takes to boot.
21:13:07 <gilbertdeb> ah yes that was me!
21:13:08 <gilbertdeb> :)
21:13:33 <gilbertdeb> I imported all the fonts I could from my wordperfect cd and got those problems.
21:13:41 <gilbertdeb> I converted them to TTF's.
21:13:57 <jstahuman> one of the larger problems is people not knowing how to use unix the right way. so we end up with crap like NFS
21:14:30 <kc5tja> How does this affect boot times?
21:14:44 <jstahuman> probably the ye olde xfs service
21:14:51 <kc5tja> Unix doesn't even deal with fonts at all until you launch X11, and even then, only when you actually reference a font.
21:14:55 <gilbertdeb> kc5tja, it spent a lot of time getting the xfontserver running.
21:15:00 <kc5tja> Oh . . .
21:15:02 <kc5tja> there's your mistake.
21:15:09 <kc5tja> :)
21:15:09 <gilbertdeb> yeah.
21:15:21 * kc5tja doesn't use a font server anymore.  XFree86 has built-in TT font support.
21:15:27 <kc5tja> Well 4.x does at least.
21:16:04 <jstahuman> well there are fonts for the console, but there are very few, and you can only load a few at a time
21:16:05 <gilbertdeb> it was on 3.356(?) I think.
21:16:49 <jstahuman> i am interested in the exokernel OSes
21:17:27 <gilbertdeb> yeah me too. but I'm afraid it is going to be SO much work.
21:17:32 <jstahuman> i don't like the one monolithic kernel idea
21:17:42 <kc5tja> jstahuman: Yes, as am I.
21:18:00 <jstahuman> there's something missing though, not sure what
21:18:00 <kc5tja> Exokernels are a very intriguing idea which I feel is being wholesale neglected today.
21:18:48 <kc5tja> gilbertdeb: Well, it is; just because we're moving code of a traditional OS out of supervisor space doesn't mean the code ceases to exist.
21:18:49 <gilbertdeb> kc5tja, but in the end don't they lead to distributed OSes?
21:19:12 <jstahuman> when I think about, I want to map an exokernel subsystem to the actual hardware for complete parallelism
21:19:18 <kc5tja> gilbertdeb: If you want a distributed OS, write a distributed OS.  There is nothing about an exokernel which forbids this.
21:19:54 <gilbertdeb> yes I'm saying that don't exokernel oses favor that the most?
21:20:36 <jstahuman> so that if all exokernel subsystems ran in parallel on their own hardware, so much less work would be needed for context switch/synchronization, with scalability to boot
21:20:56 <gilbertdeb> jstahuman, plan9. a distributed os :)
21:21:09 <jstahuman> but i am talking the same hardware
21:21:22 <gilbertdeb> how? on the same box?
21:22:02 <jstahuman> so you'd have an exokernel module running alongside the memory controller, the disk subsystem, and so on
21:22:39 <jstahuman> and if a chip could have multiple cores, with an exokernel CPU module running along each one
21:22:57 <gilbertdeb> are there any books yet on exo-kernel oses?
21:23:56 <kc5tja> The problem is, how do you put that stuff in hardware?
21:24:04 <jstahuman> one of the bottlenecks with modern OSes is all the synchronization overhead, especially with SMP since memory is shared
21:24:05 <kc5tja> The definition of an exokernel is that it provides protection only, not policy.
21:24:27 <kc5tja> Resources aren't finite -- a harddrive can appear in the system tomorrow, and maybe two days later, a new video card with new GPUs and stuff.
21:24:39 <jstahuman> kc5tja: pretty much the same way it's done today with hardware, something similar to bios
21:24:44 <kc5tja> And then there's the issue that, by putting it in hardware, does it not, by de facto, become policy?
21:24:59 <kc5tja> jstahuman: Yes.  And look how often BIOS is actually used.  :)
21:25:39 <jstahuman> kc5tja: well, this answers the question how, the usage would be the requirement
21:27:09 <jstahuman> but as the result the main CPUs become just like any other devices, without much code for juggling things around
21:27:32 <kc5tja> The only thing that would appear in hardware is the ROM containing the protection services of the device.
21:27:37 <kc5tja> e.g., software
21:27:49 <kc5tja> It can't possibly appear in hardware; otherwise, the system becomes too limiting.
21:28:39 * kc5tja just isn't comfortable with the idea of hardware-enforced protection schemes.  Many issues arise which would make said hardware designers frightened and run away.  :)
21:28:50 <kc5tja> Like, how does one maintain protection-related state?
21:28:59 <kc5tja> Page tables in the CPU are bad enough.  :)
21:29:11 <jstahuman> it's more like a software module running on top of every hardware subsystem, all modules forming your exokernel effectively
21:29:48 <gilbertdeb> kc5tja, are there any books on exo-kernel oses?
21:29:49 <gilbertdeb> at all?
21:29:55 <kc5tja> gilbertdeb: Not a single one.
21:30:05 <jstahuman> gilbertdeb: there's plenty of good information on the web though
21:30:10 <gilbertdeb> ack.
21:30:29 <gilbertdeb> but one doesn't necessarily need more than one computer for an exo-kernel OS right?
21:31:16 <kc5tja> jstahuman: OK, that makes some additional sense.  But, if we're going to do this, why not just go all the way to Chuck's ideal of massive quantities of super-cheap processing nodes all running Forth and communicating using common protocols?  :)  It is a proper superset of the idea, after all.  :)
21:31:18 <jstahuman> gilbertdeb: one CPU is all that's enough
21:31:45 <jstahuman> kc5tja: I thought his idea is to have a multi-core chip?
21:31:58 <kc5tja> jstahuman: Chuck has more than one idea at a time.  :)
21:32:08 <gilbertdeb> hehehe.
21:32:20 <gilbertdeb> I won't say anything about Forth Mysticism.
21:32:24 <kc5tja> The reason he's working on these chips at all is because he feels that the general-purpose computer is a futile endeavor.
21:32:31 <jstahuman> what's the 25x reference? a multi-core chip? or a chips of chips/
21:32:42 <gilbertdeb> kc5tja, the GP machine is mostly an entertainment box.
21:32:45 --- quit: njd ("BitchX: try our lowfat flavor too!")
21:33:14 <kc5tja> jstahuman: The 25X is a CPU chip with a matrix of 5x5 X18 CPUs.
21:33:38 <jstahuman> so a multi-core chips
21:33:45 <kc5tja> Yes.
21:33:45 <jstahuman> similar to what IBM is doing
21:33:58 <kc5tja> However, only the border processors have access to the I/O pins on the chip.
21:34:20 <kc5tja> Each processor runs from its own supply of memory, and communicates with the other cores via a special, high-speed 18-bit wide parallel bus.
21:34:41 <jstahuman> well, it certainly makes more sense than SMP
21:34:51 <kc5tja> It's basically a systolic array.
21:35:15 <kc5tja> But I envision that it could easily emulate an F21 in software at near-full-speed.
21:35:22 <jstahuman> are they all abstracted, or completely open?
21:35:43 <kc5tja> As with all Forth engines, all details of the hardware are accessable to the programmer.
21:36:37 <jstahuman> when are we going to see one?
21:37:00 <TreyB> When someone pays to tape one out and do a run.
21:37:07 <kc5tja> Who knows?  If ever?  I do know that he finally received commercial backing for development (hence the removal of the 25X material on his ColorForth page) though.
21:37:23 <TreyB> Really?  I hadn't followed it recently.
21:37:26 <kc5tja> But as far as how long it'll take to see first silicon, I don't have that information.
21:37:47 <jstahuman> so a perfect application is beowulf on the chip
21:37:55 <TreyB> Not exactly.
21:38:02 <kc5tja> It's not really powerful enough to do that type of stuff.
21:38:11 <kc5tja> It's killer application is an FPGA-replacement, I think.
21:38:24 <gilbertdeb> kc5tja, I'm sure enough signatures can be gathered for a 'commemorative' forth chip.
21:38:37 <gilbertdeb> spam nntp :)
21:38:42 <kc5tja> gilbertdeb: Do you have any idea how much it costs to fab a chip?
21:38:55 <gilbertdeb> well they are numbers aren't they?
21:38:55 <kc5tja> Assuming it works in first silicon?
21:38:59 <gilbertdeb> lets do the math and see.
21:39:15 <kc5tja> I encourage you to do the research.  :)
21:39:31 <gilbertdeb> but as it stands now, how much do you think it would cost?
21:39:51 <jstahuman> that's why we have near monopoly on larger chips
21:40:07 <jstahuman> but forth chips are simple though
21:40:24 <gilbertdeb> jstahuman, who has the monopoly?
21:40:40 <jstahuman> well intel almost has monopoly on PC chips for instance
21:41:15 <jstahuman> AMD has only recently become a sizeable competitor, and others like VIA and Transmeta are heh, nearly invisible
21:41:28 <gilbertdeb> and MIPS is dying ...
21:41:38 <gilbertdeb> and I know next to nothing about SPARC
21:41:47 <jstahuman> i say bring back Zilog
21:42:13 <TreyB> AMD makes an ultra low power chip based on the MIPS core.
21:43:11 <gilbertdeb> jstahuman, zilog is still alive and well.
21:43:37 <gilbertdeb> TreyB,  whats it called?
21:43:38 <gilbertdeb> the 29k?
21:43:40 <jstahuman> yes, i've heard
21:43:47 <TreyB> Alchemey
21:43:49 <jstahuman> embedded apps
21:43:56 <gilbertdeb> jstahuman, rumors of its death have been greatly exagerrated :)
21:44:21 <jstahuman> there are some nice things done for commodore 64 too
21:44:34 <jstahuman> contiki
21:44:48 <kc5tja> Zilog sucks
21:44:58 <jstahuman> a commodore TCP/IP stack, VNC, web browser, etc.
21:45:00 <kc5tja> The Z-80 was a huge mistake.
21:45:09 <kc5tja> Not commercially, but technologically.
21:45:28 <gilbertdeb> kc5tja, again, WIB
21:45:40 <kc5tja> Excuse me?  
21:45:45 <gilbertdeb> worse is better.
21:45:46 <kc5tja> (Forgive me, but I speak English.  :))
21:46:00 <kc5tja> Funny, the 6502 model is doing at least as well, if not better, than the Z-80.
21:46:09 <gilbertdeb> I think I'm going to get to say it a lot, so I'm introducing WIB as the official abbreviation.
21:46:24 <kc5tja> Last I checked, no Z-80 has ever ran 500MHz.  :)
21:46:33 <gilbertdeb> hehehe.
21:46:35 <jstahuman> it's more like good enough is better than
21:46:49 <gilbertdeb> wib geibt
21:46:58 <gilbertdeb> jstahuman, yours looks like a german name!
21:47:00 <gilbertdeb> too long.
21:47:24 <Herkamire> gilbertdeb: what's WIB?
21:47:30 * kc5tja laughs
21:47:35 <kc5tja> Herkamire: Worse Is Better
21:47:37 <gilbertdeb> Worse Is Better.
21:47:41 <Herkamire> gilbertdeb: might your introduction include an explanation?
21:47:47 <Herkamire> ahh
21:48:00 <gilbertdeb> it is my new mantra :)
21:48:27 <Herkamire> like my old pappy used to say: "90% of everything is crud."
21:48:30 <kc5tja> gilbertdeb: Anyway, to answer your question, on a moderately modern fab, it can cost up to $500,000 to run a single wafer.
21:48:46 <gilbertdeb> Hmmmmm.
21:48:51 <kc5tja> I'm not sure what it'd be for an older fab line, but those are dying by the wayside on a daily basis.
21:48:53 <gilbertdeb> can I have that in software please? :D
21:49:34 <kc5tja> This is why MachineForth is so appealing to me -- I can get the relative benefits of a hardware native Forth, using traditional processors, even if it is 2.2x slower than what it theoretically can achieve.
21:49:45 * Herkamire wonders if he will ever have the sort of machine he dreams of
21:50:09 * kc5tja will, someday, but it won't ever run as fast as I'd like it to be.
21:50:24 <Herkamire> I don't care so much about speed.
21:50:39 <kc5tja> Either it'll use discrete component TTL chips (just for the fun of assembling it), in which case I'll be really cramped for speed, or I'll end up using synchronous designs in FPGA chips, where I'll still be limited to 33MHz or so.
21:50:40 <Herkamire> fast enough for responsive graphical interfaces.
21:51:05 <Herkamire> what's FPGA again?
21:51:16 <gilbertdeb> kc5tja, lets all build a MOVE machine :)
21:51:18 <TreyB> Field Programable Gate Array
21:51:22 <kc5tja> Well, the Commodore 64 did reasonably well with GEOS on a 1MHz processor.  But I still won't go under 8MHz if I'm doing GUIs.
21:51:28 * Herkamire doesn't know much about hardware BTW
21:51:30 <kc5tja> Herkamire: Field Programmable Gate Array
21:51:53 <TreyB> kc5tja: Go ahead and spell "Programmable" correctly, why don't you.
21:52:14 <kc5tja> Programmable -- p r o g r a m m a b l e -- programmable.  ;D
21:52:35 <gilbertdeb> eh can you use it in a sentence?
21:52:41 <gilbertdeb> j/k
21:52:57 <gilbertdeb> hmmm I came up with another one. b/p for (bad pun) :)
21:53:14 <kc5tja> gilbertdeb: My programmable VCR caused my entire season 3 of Star Trek: Deep Space Nine episodes to be overwritten with Oprah Winfrey and Jerry Springer episodes.
21:53:26 <gilbertdeb> :O
21:53:27 <Herkamire> hmmm... past my bedtime. nice chatting boys.
21:53:31 --- quit: Herkamire ("zzzZZZ")
21:53:37 <jstahuman> there's tivo for that
21:53:39 <gilbertdeb> sacre bleu!
21:53:43 <kc5tja> Wow, early for him.  I distinctly remember him staying up later.  :)
21:53:46 <jstahuman> get with the program, people
21:53:50 <kc5tja> jstahuman: I'm old fashioned.
21:54:11 <kc5tja> jstahuman: I even still use my Amiga 500 now and again.  And it only has a 68000 in it running at 7.1590909MHz.
21:54:35 <jstahuman> old fashioned is fine, overwritten tapes is ont
21:54:50 <TreyB> s/ont/not/
21:54:58 <jstahuman> tnx
21:55:06 <kc5tja> Hey, he asked me to use programmable in a sentence, so I did.  See, that was another one.  :)
21:55:34 <gilbertdeb> :)
21:55:46 <kc5tja> At any rate, I just caught up to 2002 for O&M.
21:55:55 * kc5tja will take a break for a while.
22:01:03 <TreyB> G'night, all.
22:01:20 <gilbertdeb> 'mornin
22:03:41 --- join: Serg_Penguin (Serg_Pengu@212.34.52.140) joined #forth
22:03:53 <Serg_Penguin> hi !
22:05:21 <kc5tja> re
22:05:24 --- join: skylan (sjh@vickesh01-4877.tbaytel.net) joined #forth
22:26:54 --- quit: jstahuman ("leaving")
22:32:35 --- quit: gilbertdeb (Read error: 104 (Connection reset by peer))
22:32:37 --- join: gilbertdeb_ (~gilbertde@fl-nken-u2-c3b-45.miamfl.adelphia.net) joined #forth
22:32:56 <kc5tja> Wow.  One second difference -- impeccable timing.
22:32:56 <kc5tja> :)
22:34:40 --- part: gilbertdeb_ left #forth
23:04:16 <Serg_Penguin> kc5tja: how are you and your editor ?
23:04:38 <Serg_Penguin> i was trying to write same thing in gp-forth, but hit some trubble
23:20:01 <kc5tja> My editor?  Vibe?  It is working quite well for me.  :)
23:20:05 <kc5tja> (Sorry, am reading a web comic)
23:20:15 <kc5tja> What kind of trouble are you hitting?
23:21:25 <Serg_Penguin> no, not yours
23:21:39 <Serg_Penguin> my attemp to rewrite the control scheme 4 gp-forth
23:21:40 <Serg_Penguin> brb
23:24:40 <kc5tja> Well, I figured that much.  :-)
23:26:11 <Serg_Penguin> actually, i rewrote all the app to compile itself on every run from source
23:26:14 <Serg_Penguin> and...
23:26:18 <Serg_Penguin> variable boom
23:26:26 <Serg_Penguin> and it explodes ;((
23:27:52 <kc5tja> Variable boom?
23:28:02 <Serg_Penguin> any variable really
23:28:30 <Serg_Penguin> and messy text runs all over the screen, w/ nasty beeping
23:28:43 <Serg_Penguin> it even _hung_ my XP !
23:28:53 <kc5tja> Honestly, I've never had XP hang.
23:29:01 <kc5tja> You're as bad with XP as I am with Linux.  :)
23:29:18 <kc5tja> (Yes, I've hung Linux twice before, though on fairly old kernels; my roommate has done it with 2.4.19 though)
23:30:03 <Serg_Penguin> not really cumpleetly hung, but i had to reboot (normally, via Start) to make it sane again
23:30:40 <Serg_Penguin> and i can't figure what the hell is wrong
23:31:01 <Serg_Penguin> i took old half-done editor and converted it to self-loader
23:31:02 <kc5tja> Oh, I completely hung Linux.  :D  It was rather funny actually.
23:31:09 <Serg_Penguin> it worked as be4
23:31:16 <kc5tja> I forget what I did to do it, but it involved 802.11b and x-chat.  :D
23:31:31 <Serg_Penguin> after that, i cleared main source file..
23:31:37 <Serg_Penguin> and wrote in it:
23:32:27 <Serg_Penguin> CREATE #xxx 5 c, c" # c, c" a c, 20 c, 20 c, 20 c,
23:33:06 <Serg_Penguin> #xxx COUNT TYPE - just 4 test
23:33:11 <Serg_Penguin> it boomed !
23:33:35 <Serg_Penguin> then i reduced it just to VARIABLE BOOM - it kept booming
23:33:59 <kc5tja> Man, I'm not sure what Forth you're using, but it sure sounds awfully buggy.
23:34:09 <Serg_Penguin> but old editor had many VARIABLE's and CREATE's - and it worked !
23:34:33 <Serg_Penguin> and i did not change loader stub !
23:34:48 <Serg_Penguin> and i never seen such err's be4 ;(((
23:36:32 <kc5tja> I don't know.
23:36:44 <kc5tja> That's extremely erratic behavior for a Forth system.
23:36:56 <kc5tja> When you type VARIABLE BOOM, is that before or after you load the editor's source?
23:39:55 <Serg_Penguin> it's the only line in "editor source", after i wiped previous, working OK !
23:40:35 <Serg_Penguin> the stub is (from memory):
23:41:02 <Serg_Penguin> FORGET FORTHERR \ last word, redefined later
23:41:09 <Serg_Penguin> USE DOS.FRL
23:41:17 <Serg_Penguin> FORGET FINDFIRST
23:41:40 <Serg_Penguin> : RC " INCLUDE ASCII.FRT" LOAD ;
23:41:44 <Serg_Penguin> RESTART RC
23:41:49 <Serg_Penguin> SAVE ASCII.COM
23:41:55 <Serg_Penguin> eof
23:42:09 <kc5tja> I really don't follow that code.
23:42:22 <Serg_Penguin> it _worked_ w/ old editor, and i did not change it
23:42:54 <Serg_Penguin> it creates ascii.com, what loads all the rest from ascii.frt at every start
23:43:30 <Serg_Penguin> what you don't get ?
23:48:43 <kc5tja> All of it.  :)
23:48:57 <kc5tja> Like, I remember what LOAD does, only because you explained it to me the last time my brain exploded.  :)
23:49:33 <kc5tja> And I can figure out that USE DOS.FRL includes a library module of some kind, resumably one which enables DOS access.
23:50:43 --- join: w1k1 (~w1k1@pD9E595E6.dip.t-dialin.net) joined #forth
23:52:36 --- join: rO| (~rO|@pD9E595E6.dip.t-dialin.net) joined #forth
23:52:48 <Serg_Penguin> FORGET FORTHERR \ last word, redefined later
23:52:55 <Serg_Penguin> just to save space
23:53:09 <Serg_Penguin> it's redefined later, making errors non-fatal
23:53:26 <Serg_Penguin> and giving full string w/ error, not just line num
23:53:29 <Serg_Penguin> ...
23:53:38 <Serg_Penguin> FORGET FINDFIRST
23:53:54 <rO|> howdy serg
23:54:08 <Serg_Penguin> i need just a few words from DOS.FRL, not ones above FINDFIRST
23:54:33 * Serg_Penguin is mega fine in life, but very bad at coding now
23:54:48 <rO|> GOOD morning.
23:55:03 <Serg_Penguin> RC INCLUDE's ascii.frt
23:55:33 <Serg_Penguin> RESTART RC does what it says - newly generated Forth system executes RC first, to load ascii.frt
23:55:45 <Serg_Penguin> and SAVE is obvious
23:55:59 <Serg_Penguin> brb
23:56:14 <rO|> GOOD morning, #
23:56:20 <kc5tja> RESTART definitely wasn't obvious to me.
23:56:26 <kc5tja> Not without the additional context.
23:56:35 <kc5tja> But, I'm still not seeing how this relates to your problem.
23:56:47 <kc5tja> I thought you were trying to get the editor to work from within Forth, but from the DOS command-line?
23:57:29 <Serg_Penguin> i'm not seeing trubble too
23:57:46 <Serg_Penguin> don't get last phrase
23:58:11 <Serg_Penguin> i just run 'ascii.com' and i see black screen on what i can paint
23:58:22 <Serg_Penguin> no Forth, at leas 4 dummy user
23:59:59 --- log: ended forth/03.07.27