00:00:00 --- log: started forth/01.06.09
00:03:03 --- quit: nate37 (night)
00:20:13 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
00:20:29 <futhin> hello?
00:20:43 <futhin> anybody know why one would want to metacompile the compiler??
00:21:17 <futhin> i.e code the forth compiler in assembly, then rewrite the compiler in forth and use the old compiler to compile a new compiler???
00:28:15 --- quit: adu (Ping timeout for adu[adsl-63-201-92-89.dsl.snfc21.pacbell.net])
00:39:30 --- quit: futhin (.)
01:47:16 <Trey> ping
02:19:44 * aaronl is away: sleep
02:20:18 * Trey is away: sleeping [4:20]
05:34:10 --- join: cleverdra (jfondren@1Cust9.tnt3.florence.sc.da.uu.net) joined #forth
05:50:07 --- quit: cleverdra (Leaving)
08:21:57 --- join: cleverdra (jfondren@1Cust64.tnt3.florence.sc.da.uu.net) joined #forth
08:42:22 --- quit: cleverdra (Leaving)
09:01:41 --- join: edrx (edrx@copacabana-ttyS3.inx.com.br) joined #forth
09:02:08 --- quit: edrx (Signed off)
10:02:04 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
10:43:28 <Trey> clear
10:43:31 <Trey> ooops
10:57:33 --- join: I440r (mark4@purplecoder.com) joined #forth
10:57:48 <Trey> Howdy I440r
10:57:52 <I440r> hi!
10:58:51 <I440r> has tcn been in here lately ?
10:59:20 <Trey> I only just signed in myself.
10:59:27 <I440r> k :)
10:59:44 <I440r> i dont think he will have been, he doesnt have much net access at the moment :)
10:59:58 <I440r> which makes it difficult to do isforth development with him hehe
11:00:00 <futhin> would anybody here know chuck moore's email addy?
11:00:10 <I440r> hmmmm
11:00:15 <I440r> google search it
11:00:16 <Trey> Jeff Fox would, if he has one.
11:00:26 <I440r> he prolly doesnt have one
11:00:32 <I440r> he is legally blind u know...
11:00:40 <Trey> "Meddle not in the affairs of dragons..."
11:00:43 <futhin> jeff fox comes on here?
11:00:48 <Trey> Not that I know of, futhin.
11:00:53 <I440r> brb gotta move my car
11:00:55 <futhin> i'll just email elizabeth rather
11:00:57 <Trey> Flooding?
11:01:10 <Trey> futhin: good luck :-)
11:01:41 <futhin> trey: it should work, she and chuck started the forth, inc. company
11:02:22 <futhin> hopefully they are still friends or she can speak for him or something :)
11:03:25 <Trey> Sure, but what do you want from Chuck?
11:03:49 <futhin> i just want to talk to him about forth's future, visions, ideas, etc..
11:03:56 <futhin> just want to explore some things
11:04:19 <Trey> You should probably have a better understanding of the history and issues, first.
11:04:41 <futhin> true, i'm not going to email anybody right now :)
11:06:17 <I440r> just join comp.lang.forth
11:06:24 <I440r> you can post messages in there to her
11:06:30 <I440r> she reads everything in there
11:06:30 <futhin> trey: are you working on any projects?  any os devving?
11:06:42 <Trey> Nothing public.
11:07:10 <Trey> I think I may add PCI bus support to Enth, just for the experience.
11:07:36 <futhin> well, i want to talk to chuck about forth as related to something else, which i won't mention ;)
11:07:58 <futhin> so posting it to comp.lang.forth won't be any good :)
11:08:17 <Trey> If you have a -real- project in mind, with a serious business proposition, he might listen.
11:08:19 <futhin> trey: you are working on an os?
11:08:47 <Trey> I have thoughts about an OS, but I haven't written any code.
11:09:09 <Trey> I've done some work on a native code Forth compiler for x86.
11:09:39 <Trey> I'd like to write the OS in Forth, compiling the required code at boot/load time.
11:09:43 <futhin> somebody needs to write a native code Forth compiler for windows :)
11:09:50 <I440r> trey optimizing ?
11:09:53 <Trey> Someone already has.  Lots of someones.
11:10:10 <Trey> I440r: I plan on doing opitimzations, but I haven't gotten that far yet.
11:10:13 <I440r> not a native code compiler i dont think
11:10:18 <I440r> i dont like optimizing compilers
11:10:19 <I440r> at all
11:10:43 <futhin> hm.. i thought native code == assembly/machine language.   code a forth compiler in assembly and use the windows "interrupts"  to do the graphics, etc..
11:10:50 <Trey> I plan to optimize at the machine code level.
11:11:04 <I440r> there should be a 1:1 correlation between source and object created and with optimizations a given piece of code can be compiled into different object code in different places
11:11:06 <I440r> you los that
11:11:20 <I440r> thats what a native forth compiler creates
11:11:32 <I440r> it doesnt compile to forth, it compiles to asm
11:11:52 <I440r> skip carter used to come on here every now and again
11:11:53 <futhin> if you use a c wrapper then it's not really a native code compiler?
11:12:02 <I440r> ugh
11:12:12 <I440r> why bastardize forth by wrapping it up in c
11:12:17 <I440r> thats why im writing isforth
11:12:21 <I440r> because i HATE c
11:12:22 <futhin> if you code the forth compiler in C it isn't a native code compiler
11:12:32 <I440r> and every lame ass forth compiler written in c is NOT a forth compiler
11:12:44 <I440r> its just one more piece of shit written in c
11:13:01 <futhin> but all the compilers for Windows are coded in C..  afaik nobody has coded a forth compiler for windows in assembly
11:13:12 <I440r> tom zimmer ?
11:13:24 <I440r> all the forth compilers for linux are coded in c too
11:13:28 <Trey> I440r: I plan on doing the following - generate largish machine code definitions through agressive (but programmer controled) inlining.
11:13:32 <I440r> they are all piles of hippo dung
11:13:43 <futhin> win32for == tom zimmer ??   win32for was coded in c
11:14:06 <I440r> didnt know that, thats just one more reason not to like tom zimmer :)
11:14:08 <Trey> I440r: then I plan to add peephole optimization to eliminate redundant moves/stores/etc.
11:14:15 <futhin> i440r: heheh :)
11:14:55 <I440r> actually tom was never a realy good forth coder...
11:14:59 <Trey> I440r: then I plan to add a processor specific scheduling pass.
11:15:03 <I440r> one look at fpc will show that
11:15:09 <I440r> trey ugh :)
11:15:16 <futhin> i440r: he can't help it,  coding in assembly FOR windoze means you've got to do a whole bunch of stuff in order to make dll calls and whatnot..  very few people code windows applications in assembly :)
11:16:09 <I440r> same in  linux
11:16:28 <Trey> I440r: performance matters.
11:16:46 <futhin> huh.. weird..   win32for is in the /forth/compilers/native/windows/ section...
11:17:11 <futhin> can you make a native forth compiler in c?  i'm confused what the definition of a native compiler is :)
11:17:40 <Trey> "Native" compilers generate native code (assembly/machine language).
11:17:46 <futhin> yeah
11:17:48 <futhin> that's what i thought
11:17:50 <Trey> As opposed to threaded code.
11:18:04 <Trey> (well, non subroutine threaded code)
11:18:13 <futhin> huh?
11:18:32 <futhin> don't all compilers use ITC and DTC and token?  even native compilers?
11:18:40 <Trey> No.
11:19:17 <Trey> Did you read those articles I pointed you at?
11:19:38 <futhin> i thought that threading was just a method of multitasking or executing the code..  you can implement threading on the assembly level can't you?
11:19:58 <Trey>  http://www.zetetics.com/bj/papers/index.html
11:20:16 <futhin> yeah i read some of that
11:20:28 <Trey> Read _all_ of it.  Twice.
11:20:33 <futhin> heh
11:20:36 <Trey> Really.
11:20:37 <futhin> probably three times for me :)
11:20:52 <futhin> i only read some of it because it started going over my head
11:21:28 <Trey> You'll need a fairly good grasp of the issues before you can get much farther.
11:21:46 <Trey> At least if you want to talk about implementations.
11:21:53 <futhin> wtf.. the guy talks about ITC, DTC and token.   can't you code the whole threading thing at the assembly level, isn't it then "native code" ?   argh
11:21:56 <I440r> futhin thats a different sort of thjreading
11:22:00 <Trey> You don't need to know much of that stuff if you just want to write Forth code.
11:22:13 <I440r> native forth is UNthreaded
11:22:24 <futhin> is that faster or slower?
11:22:25 <Trey> I440r: well....
11:22:52 <I440r> hmm i suppose its not tho is it...
11:23:10 <Trey> Subroutine threaded code gives you a starting point for a native code implementation.
11:23:16 <futhin> the bradford guy talks about threading, he makes it sound like all compilers should thread or something....
11:23:20 <I440r> do native forth compilers optimize each word seperatly or take the whole program and optimize based on past code too
11:23:52 <Trey> I440r: "native" forth compilers (IMHO) use the CPU as the inner interpreter.
11:24:31 <Trey> After you do that, you start working on various "optimizations".
11:25:04 <futhin> quote: "Threaded code" is the hallmark of Forth. A Forth "thread" is just a list of addresses of routines to be executed
11:25:19 <futhin> from that url
11:25:31 <Trey> Conceptually correct, and a good place to start.
11:30:52 <I440r> if you look at the way forth : definitions are normally compiled, they look something like the following in asm....
11:30:55 <I440r> some_word:
11:30:57 <I440r>  call nest
11:31:02 <I440r>  dw foo
11:31:04 <I440r>  dw blah
11:31:07 <I440r>  dw unnest
11:31:38 <I440r> the call to nest pushes the current interprative pointer onto teh return stack and pops the return address of teh call INTO ip
11:31:42 <I440r> ip now points at foo
11:31:57 <I440r> every coded definition ends witgh a next... INCLUDING nest
11:32:26 <I440r> next takes the item ip points to
11:32:30 <I440r> and jumps to that address
11:32:35 <I440r> so we wouold jump to foo
11:32:38 <I440r> and execute foo
11:32:56 <I440r> it also increments ip to point to the blah so
11:33:17 <I440r> the next at the end of foo causes us to jump to blah
11:33:21 <I440r> etc etc etc
11:33:24 <I440r> thats threading in forth
11:33:37 <I440r> actually, thats direct threading
11:33:56 <I440r> i dont like indirect threading, its wastefull imho
11:34:07 <I440r> anything else is not forth
11:34:17 <I440r> subroutine threading is NOT FORTH grrrr
11:34:31 <futhin> heh
11:40:06 --- quit: I440r (brb maybe :))
11:49:47 --- join: I440r (mark4@purplecoder.com) joined #forth
11:50:12 --- mode: ChanServ set mode: +o aaronl
11:50:33 --- mode: ChanServ set mode: +o clog
11:50:36 --- mode: ChanServ set mode: +o futhin
11:50:38 --- mode: ChanServ set mode: +o lar1
11:50:42 --- mode: ChanServ set mode: +o Trey
11:50:44 --- mode: ChanServ set mode: +o I440r
11:51:31 <Trey> I440r: the choice of threading model depends on many factors.  ITC doesn't necessarily mean wastefull.
11:51:55 <I440r> a pointer to a pointer to the code instead of a pointer to the code ?
11:52:00 <I440r> sounds it to me :)
11:57:26 <futhin> i440r:  i maybe be wrong about win32for being coded in c
11:58:31 <futhin> question:  is it better to do "if <stuff> exit then <blah>"  or "if <stuff> else <blah> then" ??
11:58:45 <I440r> i would say th elatter
11:58:53 <futhin> ok
11:58:58 <I440r> you should never have more than one exit point from a definition if you can avoid it
11:59:05 <I440r> plues its smaller :)
11:59:27 <futhin> hm.. not really smaller: exit or else
12:00:21 <I440r> oh yae hehe actualy the 2 exit point method would be smaller :)
12:00:33 <I440r> but i would still go for the one exit point only method
12:04:22 <futhin> would you two (i440r and trey) mind looking at some MUF code for me and changing it to fit your styles of coding?  :)
12:05:04 * Trey doesn't have forth coding style yet.
12:05:19 <I440r> hehehe
12:05:30 <futhin> trey: haven't you coded in forth? :)
12:05:52 <futhin> i440r: can i send you some MUF code (it's a forth-like lang)
12:06:02 <I440r> what does it do ?
12:06:07 <I440r> sure dcc away
12:06:16 <Trey> futhin: not really.
12:07:11 <I440r> i got it
12:07:14 <futhin> i440r: there's a second piece
12:07:15 <I440r> dont need to send again :)
12:07:21 <I440r> its called the same thing tho
12:07:29 <futhin> i440r: it's a different one.. 165-mod.m
12:07:38 <futhin> i sent you 165.m
12:07:43 <I440r> oh ehhe
12:07:46 <I440r> try again :)
12:08:15 <I440r> connection refused ???????????
12:08:16 <I440r> wtf
12:08:31 <I440r> i dont like
12:08:44 <futhin> 165-mod.m is cleverdra's style.. 
12:08:48 <I440r> blah if else stuff then
12:08:59 <I440r> i would do blah ! if stuff then ;
12:09:07 <I440r> er not even  not ! grrr
12:09:08 <futhin> although i modified 165-mod.m myself a bit
12:09:13 <I440r> ive been brainwashed by c grrr
12:09:32 <futhin> i440r: huh? what are you talking about?
12:09:34 <I440r> also i reserve ( and ) for scak comments
12:09:40 <I440r> i use \ for line comments
12:09:52 <I440r>    checkargs if
12:09:53 <I440r>    else me @ "You need to type: order <petname> <commands>" notify exit then
12:10:05 <I440r> there is no if part, only an el;se part
12:10:09 <I440r> BAD 
12:10:18 <futhin> yeah.. well, check the 165-mod.m :)
12:10:31 <futhin> checkargs 0= if me@ "you need to.." notify
12:10:47 <futhin> checkargs 0= if me@ "you need to.." notify exit then
12:11:10 <futhin> why would you use !  when that's for variables and variables are bad? :)
12:12:21 <I440r> you can get rid of the 0= on one of those at least
12:12:31 <I440r> chackargs if exit else ..... then
12:12:43 <I440r> but that is the same as the original
12:12:52 <I440r> so the 0= is better
12:12:53 <futhin> there's a part in 165-mod.m that needs to be fixed.. the "continue"  word is in the wrong word..
12:12:54 <I440r> or...
12:13:13 <I440r> you could have checkargs return the opposite conditions
12:13:18 <I440r> and totally eliminate all this
12:13:23 <I440r> checkargs if "blah" then
12:14:01 <futhin> check 165-mod.m :)
12:14:19 <futhin> but i'm just more concerned about the different ways of coding forth..  i just want to see other people's styles.. 
12:14:29 <futhin> then i can adopt something for myself..
12:17:07 <futhin> how do i create a multiword exit?  
12:17:30 <futhin> like suppose i have a word that calls another word that evals something false and wants to break out of both words
12:20:20 <I440r> you would have to either mess with the return stack or have the top level exit based on a returned value
12:22:04 <Trey> I need this machine to do a little Enth development.  I'll return later.
12:22:19 --- quit: Trey ()
12:31:05 <futhin> i440r: you messing with the muf code?
12:35:59 --- part: futhin left #forth
12:36:00 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
12:41:46 --- quit: I440r (Ping timeout for I440r[purplecoder.com])
12:43:43 * aaronl is back (gone 15:21:15)
13:08:55 --- join: I440r (mark4@purplecoder.com) joined #forth
13:09:14 <I440r> damned eth0 keeps dying grrr
13:21:01 <futhin> chuck moore quote: "So people who draw stack diagrams or pictures of things on the stack should immediately realize that they are doing something wrong. Even the little parameter pictures that are so popular"
13:21:27 <I440r> \?
13:22:12 <futhin> http://www.ultratechnology.com/1xforth.htm
13:22:51 <futhin> he talks about stack and stuff
13:23:03 <futhin> quote :
13:23:04 <futhin> A Forth word should not have more than one or two arguments. This stack which people have so much trouble manipulating should never be more than three or four deep.
13:23:04 <futhin> But as to stack parameters, the stacks should be shallow. On the i21 we have an on-chip stack 18 deep. This size was chosen as a number effectively infinite
13:31:15 <aaronl> I440r!
13:31:24 <aaronl> I440r: I've been playing with 68000
13:31:44 <aaronl> I440r: my Handspring Visor has a 33mhz embedded 68k, and i've been coding for it
13:32:07 <I440r> cool :)
13:32:18 <I440r> i like 68k alot
13:32:37 <aaronl> me too
13:32:43 <aaronl> it has a good selection of registers
13:32:58 <I440r> and a good selection of instructions :)
13:33:05 <I440r> i loke movem alot :)
13:33:10 <aaronl> one problem i have though is i don't think the conditional jumps are well-documented
13:33:13 <aaronl> i LOVE movem
13:33:14 <I440r> and move address,address :)
13:33:27 <aaronl> anyway, exactly what flags do instructions like JGE test?
13:33:40 <aaronl> and what's the difference between a jump and a branch? is a branch just a relative jump?
13:33:47 <I440r> erm i forget what flags 68k has
13:33:55 <I440r> but i know that info is doccumented..
13:33:57 <aaronl> i don't know offhand either
13:34:03 <aaronl> i'd like to get the docs for that tho
13:34:10 <aaronl> in my 68020 manual it doesn't seem to say
13:34:16 <aaronl> just "jumps if greater or equal", etc
13:34:25 <aaronl> is a branch faster than a jump?
13:34:29 <I440r> the 020 was the first 68k with a vbr wasnt it ?
13:34:39 <I440r> erm dunno... i wouldnt have thunked so
13:34:47 <I440r> its been years since i did any :)
13:34:56 <aaronl> i dunno. i don't actually  have a 68020. i just use the book for 68000 info
13:35:07 <I440r> :)
13:35:26 <aaronl> the chip itself is some dragonball -- embedded 68000-compatible chip with display controller and other cool stuff that i don't use
13:36:01 <aaronl> hmm, the addressing modes are cool
13:36:08 <aaronl> IIRC move can do mem-mem moves
13:36:11 <aaronl> which is useful
13:36:39 <I440r> yes
13:36:46 <I440r> movem and move can do memory to memory
13:36:48 <I440r> erm
13:36:50 <I440r> i think movem can
13:36:56 <I440r> maybe not..
13:37:01 <aaronl> afaik movem does registers
13:37:06 <I440r> thats just register erm its been too long :)
13:37:18 <I440r> i know movem does registers...
13:37:23 <I440r> movem is how you do pushes and pops
13:37:30 <I440r> there are no push or pop instructions :)
13:38:26 <aaronl> if i wrote in C syntax: d1 = *(*(sp+1)+1)
13:38:33 <aaronl> i would need multiple instructions, right?
13:38:49 <aaronl> or could i use memory-memory indirect?
13:39:22 <I440r> erm hehe i dunno man its been over 10 years since i did 68k :)
13:39:28 <aaronl> ya so i wrote a subset of Base64 in asm for fun
13:39:33 <aaronl> but i don't think you would wanna see
13:39:37 <aaronl> i used AT&T syntax
13:39:51 <aaronl> that seems the status quo for 68k
13:40:55 <I440r> yukk
13:40:57 <I440r> hehe
13:41:30 <aaronl> i actually kinda like it :P
13:41:42 <I440r> ur sick
13:41:45 <I440r> u need help :P
13:41:59 <aaronl> at least i don't like DOS :P
13:42:08 <aaronl> anyway, i don't see what's so bad about AT&T
13:42:26 <aaronl> you specify size on the instruction instead of on the operands
13:42:32 <aaronl> which seems to me like the right way
13:43:21 <aaronl> as for operand order, it really doesn't matter to me
13:43:39 <aaronl> oh, I440r, can you explain to me why there is ANY POINT WHATSOEVER in little endian?
13:53:16 <I440r> not realy :)
13:53:31 <aaronl> heheh
14:10:20 --- join: edrx (edrx@200.240.18.66) joined #forth
14:12:23 * aaronl is away: I'm busy
14:13:52 <I440r> ed!
14:13:59 <I440r> arron op everyone :P
14:14:10 <I440r> aaron even grrr
14:14:17 <I440r> mental note... myst learn to type!
14:15:48 --- quit: edrx (Signed off)
14:32:47 --- mode: aaronl set mode: +o I440r
14:33:23 --- mode: aaronl set mode: +o futhin
14:33:42 * aaronl is back (gone 00:21:20)
14:45:51 <futhin> no point in having op if i ain't on the perm op list ;)
15:10:47 <aaronl> ok 4 4 +                                                                        
15:10:47 <aaronl> ok .                                                                            
15:10:47 <aaronl> 8                                                                               
15:10:49 <aaronl> Heh
15:10:55 <aaronl> sparcs are so cool
15:18:07 <futhin> why? :)
15:19:07 <futhin> dammit, i'm reading stuff chuck moore said and damn, i totally identify with his philosophies
15:19:14 <futhin> it's sweet
15:22:04 <aaronl> suns have FORTH in rom
15:40:37 --- join: nopcode (yap@p3EE3E41D.dip.t-dialin.net) joined #forth
15:40:39 <nopcode> hey
15:40:42 <nopcode> anyone awake ?
15:42:47 <futhin> yes
15:43:07 <futhin> everybody is awake except clog and lar1 ;)
15:43:14 <nopcode> great :)
15:43:23 <nopcode> i need a small, pluggable forth system like ficl
15:43:29 <nopcode> but i need floating point math
15:43:37 <nopcode> any ideas ?
15:44:25 <futhin> implement the floating point math yourself?
15:44:44 <futhin> why do you need "a small, pluggable forth system"  ?
15:45:03 <nopcode> for putting it into quake 3 :)
15:45:27 <nopcode> to have a proper language there
15:46:45 <futhin> you could fake floating point math.. 
15:46:56 <futhin> : currency <# # # char . hold #s char $ hold #> type cr ;
15:46:56 <futhin> In this case, this:
15:46:56 <futhin>      200012 currency
15:46:56 <futhin> will print this:
15:46:56 <futhin>      $2000.12
15:47:09 <futhin> that's for 4th
15:47:21 <futhin> but might be applicable
15:47:24 <nopcode> well :D
15:47:25 <futhin> <---  new to forth
15:47:35 <nopcode> yeah but i need lots of interfacing to C floats
15:47:51 <futhin> aaronl: you there?
15:47:55 <aaronl> yes
15:48:13 <futhin> hm, can you help nopcode? :)
15:48:18 <nopcode> :)
15:48:32 <I440r> nop ive seen quake/doom style games written in 100% asm and they were faster than the shitty c code
15:48:35 <futhin> nop: i don't know that much, i'm still learning and stuff
15:48:36 <I440r> c is a pile of shit
15:48:40 <I440r> grrr :P
15:48:52 <nopcode> l440r bla, you don't get anywhere with 100% asm
15:49:07 <I440r> bullshit 
15:49:19 <I440r> and thats talking from 20 years of experience coding for a living
15:49:26 <nopcode> why would it be faster with opengl drivers rendering everything ?
15:49:27 <I440r> asm is faster, easier to maintain and faster to write
15:49:34 <I440r> c just has the advantage that ANY idiot can code it
15:49:44 <nopcode> lol ? what about portability ?
15:49:51 <futhin> assembly is portable
15:49:52 <I440r> who cares
15:50:01 <futhin> but most people who code in c don't care about portability
15:50:05 <I440r> if i write jack of all trade code its going to be a master of mone
15:50:09 <nopcode> l44or every openminded person... i have asm coders found not to be
15:51:03 <futhin> nopcode:  "portability" is just a bullshit word.  most idiots who code in C aren't coding for portability
15:51:12 <futhin> besides, assembly is portable
15:51:17 <nopcode> futhin is it ?
15:51:42 <nopcode> i don't think so =)
15:51:46 <futhin> sure.. depends on what kind of portability you want
15:51:57 <nopcode> source-level of course
15:52:00 <futhin> if you want to make os-specific calls then you have to use c-hooks or something
15:52:19 <nopcode> futhin what about running x86 assembler code on sparc cpus ?
15:52:37 <futhin> but if you wanted architecture portability, you could write code in virtual architecture assembly and then write a translator to translate the assembly code to the new architecture
15:52:54 <nopcode> thats the idea of c. doh.
15:53:25 <futhin> naw, you'd get the benefits of assembly
15:53:43 <futhin> plus C sucks :P
15:53:44 <nopcode> no you wouldnt
15:54:02 <nopcode> you couldnt even do register allocation yourself as you dont know how many registers the target machine will have
15:54:21 <futhin> it's more effective to code for a virtual architecture assembly because you use a different registers model
15:54:36 <futhin> which then makes it easier to translate to any other architecture
15:54:43 <nopcode> futhin what about processors with register stacks ? completely different.
15:55:38 <futhin> naw.. it's not that hard to translate registers to stacks
15:55:40 <nopcode> thats bullshit, you guys should feel lucky that there is a lowlevel language which is fairly portable and has decent implementations... c...
15:55:57 <futhin> there's some good algorithms out there on translating registers to stacks
15:56:08 <futhin> that haven't actually been implemented afaik
15:56:14 <nopcode> i personally wouldnt want to rewrite the whole of a, say, database system in ia64 assembly :)
15:56:33 <futhin> well, the algos have been tested.. but not implemented seriously afaik
15:56:46 <futhin> nopcode: naw, you don't, you just write a translator
15:57:04 <nopcode> futhin you wouldnt make use of the ia64 architecture then. a c compiler would.
15:57:06 <nopcode> instantly.
15:57:26 <futhin> what is a c compiler ?  it's a translator
15:57:39 <futhin> somebody else took the time to make that translator
15:57:44 <nopcode> yeah, but based on a halfway portable language
15:57:49 <futhin> and who says you can't make an assembly translator?
15:58:11 <futhin> based on a halfway portable language? what the hell are you talking about?
15:58:15 <nopcode> c :)
15:58:31 <futhin> no, they had to code the damn compiler in assembly or some other language before they had access to c
15:58:40 <nopcode> sure, that were the hard times
15:58:59 <futhin> so uh
15:59:09 <futhin> what are you messing with forth for?  you want to implement something in q3??
15:59:22 <nopcode> yeah
15:59:33 <futhin> why do you want to implement forth in q3?  how will it be used?
15:59:41 <nopcode> i want to extend / replace the internal scripting language
15:59:49 <nopcode> to allow for decent scripting in quake3
16:02:12 <futhin> why not use ficl and fake the floating point math?
16:02:27 <nopcode> because i need interfacing to standard c floats
16:02:34 <futhin> or implement it in a few lines of forth code 
16:02:42 <nopcode> few lines ?
16:02:51 <futhin> why do you need to interface to standard c floats?
16:02:51 <nopcode> the float handling should be ieee compliant
16:02:56 <nopcode> because q3 uses them :)
16:03:31 <futhin> i'm not quite sure what you want to do, why don't you interface to the standard c floats inside ficl then?
16:03:53 <nopcode> because to do that properly, i'd have to write interfaces to the whole of math.h
16:04:04 <nopcode> someone probably already did that, that's why i asked
16:04:52 <futhin> http://www.mindspring.com/~esessoms/ficlx/
16:04:53 <futhin> check there
16:04:58 <futhin> i dunno if that's what you wanted
16:06:18 <futhin> check http://ficl.sourceforge.net/ficl.html  and search for float
16:06:22 <futhin> there's a float.c
16:06:40 <futhin> <-- is clueless as to what you need :)
16:07:06 <nopcode> futhin hey
16:07:18 <nopcode> that wasnt there last time i checked, thx, that might actually solve all my problems :)
16:07:34 <futhin> heh
16:07:35 <futhin> ok :)
16:07:51 <futhin> may the forth be with you
16:12:53 <nopcode> =)
16:17:40 --- quit: nopcode (Leaving)
16:35:40 * aaronl is away: weirdstuff
17:35:17 --- join: Trey (bowser@ns.TreySoft.com) joined #forth
17:35:37 <Trey> Hmmm.
17:35:45 * Trey searches for cleverdra
17:35:58 <futhin> haven't seen him here
17:36:01 <futhin> today
17:36:17 <futhin> where are you searching?
17:36:22 <futhin> other irc nets?
17:40:30 <Trey> Just here.
17:40:36 <Trey> I had an Enth question for him.
17:41:42 <Trey> I want to know if Enth has words for accessing the x86 I/O address space (specifically, 32 bit reads and writes).
17:42:12 <Trey> I found PC@ and PC!, but they appear to use 8 bit accesses.
17:42:22 <Trey> I'll check back later.
17:42:28 * Trey is idle: life, etc.
18:03:56 <futhin> what is with the bullshit that jeff fox is blind?!?!?!
18:23:23 --- join: TheBlueWizard (TheBlueWiz@ip-216-25-202-50.vienna.va.fcc.net) joined #forth
18:23:23 --- mode: ChanServ set mode: +o TheBlueWizard
18:23:33 <TheBlueWizard> hiya all
18:27:05 <futhin> hey
18:27:34 <TheBlueWizard> hiya futhin
18:35:01 <futhin> it's interesting reading jeff fox's and chuck moore's stuff
18:35:02 <futhin> papers
18:35:29 <TheBlueWizard> on ultratechnology's website, eh?
18:35:33 <futhin> yup
18:35:45 <futhin> i think i see where they fucked up in getting forth widely accepted :)
18:36:13 <TheBlueWizard> how so?
18:36:23 <futhin> some of the stuff on ultatech's site is a little propaganda-like grrr..  
18:36:24 <futhin> um
18:36:52 <futhin> how..  i think they use too many disparaging remarks and they approach the issue in the wrong way
18:37:06 <futhin> they are a little too arrogant and it shows :)
18:37:55 <TheBlueWizard> hehe...I notice that many Forth boosters tend to be fanatic...as if all the world's woes can be solved with a few definitions of Forth words :)
18:37:57 <futhin> like this jeff fox page i'm looking at is "low fat" this, "low fat" that.. it's excessive and besides the point
18:38:23 <futhin> i know i love forth and i've only coded 1 thing in a forth-like language :P
18:38:25 <futhin> but
18:38:48 <TheBlueWizard> then again there is a culture that holds bloatware in utter disdain....I am partly in that camp myself :)
18:39:31 <futhin> i hate bloatware.  i hate windows and X.  I hate anything that can't handle my blazing fast 133 mhz pentium computer with 16 megs of ram
18:40:09 <futhin> and secretly, i think that the 386 computers with 2 or 4 megs of ram are adequate for EVERYTHING  if the code is good
18:40:56 <TheBlueWizard> hehe....and my Amiga 3000 runs AmigaOS, which requires less than 512K space itself (I have 8 meg on it though, and it runs at 25 MHz)
18:41:53 <TheBlueWizard> I believe it is entirely possible to write a decent GUI based OS on 386/486 machines, though a bit of graphics acceleration may be needed though
18:42:01 <futhin> naw
18:42:02 <futhin> heh
18:42:31 <futhin> one of the pages on ultratech mentions a gui in 1k of code or something like that
18:42:34 <futhin> er
18:42:37 <futhin> not 1 k
18:43:35 <TheBlueWizard> Amiga has custom chips that handles various things like bit blitting, sound generating, etc...frees up the CPU for doing something important, like cracking CIA's secret ;)
18:43:36 <futhin> yes 1k..  the itv 4os gui is 1k
18:43:45 <futhin> but basically a few k of code for a gui
18:44:59 <TheBlueWizard> 1k?  maybe possible for text based "GUI" (TUI? :)....but if it really uses bitmapped display, I greatly doubt it....have to see it to be convinced
18:46:08 <futhin> quote
18:46:09 <futhin> I have been telling people for years that a GUI can be done in a few K and that it doesn't need to eat many megs or most of the computing power of a computer. I like to point to the 1K GUI in iTV's 4OS. I like to compare it to Microsoft Windows. 
18:46:16 <futhin> --- jeff fox
18:46:19 <futhin> low fat computing
18:46:45 <TheBlueWizard> *very* low fat computing hehe
18:46:48 <futhin> heh
18:47:18 <TheBlueWizard> is the source code available for that 1k "GUI"?
18:47:24 <futhin> hm. no idea :)
18:47:51 <TheBlueWizard> hehe
18:48:41 <TheBlueWizard> if no source code is available, then either it is proprietary (Jeff wouldn't want to give his secrets away...) or that he is bluffing
18:49:22 <TheBlueWizard> either way, the prospective customers would have hard time trying to believe....
18:49:33 <futhin> i doubt the code is available, after all, it belongs to the iTV Corporation.
18:49:39 <futhin> yeah
18:49:44 <futhin> that's one thing they are screwing up
18:49:58 <futhin> i wouldn't mention the unbelievable stuff :P
18:50:45 <TheBlueWizard> there's more?
18:52:03 <futhin> no
18:52:12 <futhin> i'm talking about in general
18:52:17 <TheBlueWizard> oh
18:52:19 <futhin> like their f21 chips
18:52:23 <futhin> or whatever
18:52:27 <TheBlueWizard> ah
18:52:35 <futhin> some of the specs are "unbelievable" to the general people
18:52:51 <TheBlueWizard> I read a bit of material on F21 stuff....I have to admit it is really interesting
18:53:06 <futhin> there was an echo timer thingie.. one of the chips can go 10 gigahertz or something
18:53:41 <TheBlueWizard> that F21 stuff seems credible to me....(except perhaps for "lunatic" statements that I haven't the opportunity to check out....)
18:54:20 <futhin> heh
18:56:04 <TheBlueWizard> but I do buy the claims re: F21...I myself have had been dabbling in "Forth architectures" for a while, so it isn't off base
18:56:09 <futhin> check http://www.ultratechnology.com/echo.htm
18:56:11 <futhin> echo timer
18:56:36 <futhin> 10 gigahertz
18:57:05 <futhin> respect ma 10 gigahertz foo!
18:57:18 * TheBlueWizard is now looking at the webpage
19:06:17 <TheBlueWizard> I'm no HW guy, but I find his claim dubious....done with .8u micron technology?  I personally don't buy it at all....for starter, how does one check it using a more conventional/proven technology?  And 10 GHz signal would be more in microwave or higher ranger, and thus would be more difficult to contain (i.e. it tends to get dissipative real fast, unless you are using right materials)
19:12:58 <futhin> uh. it's electricity.  the echo of the electricity hitting the switch is causing the timing.. 
19:16:02 <TheBlueWizard> are you something of a HW guy?
19:16:29 <futhin> naw
19:16:36 <futhin> but he made it :P
19:16:58 <TheBlueWizard> ha ha ha
19:17:11 <futhin> why should i find it dubious?  i'm no expert.  most people are morons too
19:17:28 <futhin> seriously, i know quite a few smart people, but they are dumb as hell!
19:17:32 <TheBlueWizard> ha ha ha true
19:18:09 <futhin> like, if a smart person buys into the hype of C and OO and "portable"  is he a smart person??   :)
19:18:34 <TheBlueWizard> I know my nick suggests that I should be all-knowing on everything....which I ain't :P
19:18:59 <futhin> naw.. your nick suggests you are the wizard of blue stuff
19:19:05 <futhin> :P
19:20:41 * TheBlueWizard laughs uproariously
19:21:45 --- mode: futhin set mode: +o Trey
19:22:47 <futhin> as far as i know.. everybody is stupid as fuck.  including me, i'm stupid as fuck.  there's probably only 5 people in this world who aren't as stupid as fuck
19:24:43 <TheBlueWizard> lemme guess...Albert Einstein, Donald Knuth, a plain old rock, and I don't know of other two :P
19:25:00 <futhin> naw.. i'm talking about right now, in the present
19:25:06 <futhin> i don't know donald knuth
19:26:51 <futhin> well.  it's probably not 5 people.. maybe 1 or 2..  child prodigies or something
19:27:15 <futhin> who knows
19:27:54 <TheBlueWizard> you don't know Donald Knuth?!?   He is the author of "The Art of Computing" books....he is one real sharp bright dude around
19:28:11 <futhin> there's edsger djikstra.. he'
19:28:16 <futhin> there's edsger djikstra.. he's pretty ph33rabl
19:28:19 <futhin> e
19:29:44 <TheBlueWizard> I know that guy too...famous for his "No gotos" dictum...and D. Knuth demurred..."Not so fast", and he gave a counterexample :)
19:30:04 <TheBlueWizard> now you see why Donald is so smart!
19:30:15 <futhin> heh
19:31:14 <futhin> i judge a person to be smart or not by whether the person can keep his mind open and still process the information coming at him.  the wider the mind is open, the harder it is to process the info and the smarter he is
19:31:48 <futhin> some of my computer friends are "smart" but..    one of them took a look at forth and dissed it outright
19:32:17 <TheBlueWizard> so, by this logic, the smartest guy has his mind infinitely open that he can't process anything....so that is equivalent to being a rock!
19:32:28 * TheBlueWizard bows low to a plain old rock
19:32:28 <futhin> naw
19:32:29 <TheBlueWizard> ;)
19:32:35 <futhin> no
19:32:43 * TheBlueWizard laughs uproariously
19:33:11 <futhin> that's not quite what i meant... the guy with the mind open widest AND still be able to process the information..  generally, if you can't process the info, you shut your mind down a little bit
19:33:43 <futhin> and then there's ambition which i like to tie into smartness
19:34:14 <TheBlueWizard> yeah :)
19:34:30 <futhin> and clarity
19:34:38 <futhin> "it takes a genius to simplify"
19:35:34 <futhin> i'm looking that the ultratech site and they didn't do such a good job communication-wise with the unarmed public :)
19:36:00 <TheBlueWizard> Einstein said: "Make it as simple as possible, but not too simple" :)
19:36:24 * TheBlueWizard blinks at the choice of word "unarmed"
19:37:13 <futhin> unarmed==unprotected from the barage of words
19:37:18 <futhin> or something like that
19:37:25 <futhin> unarmed==unprotected
19:37:53 <TheBlueWizard> uh, I get it
19:44:25 <futhin> <TheBlueWizard> I know that guy too...famous for his "No gotos" dictum...and D. Knuth demurred..."Not so fast", and he gave a counterexample :)
19:44:30 <futhin> where did you read that??
19:46:43 <TheBlueWizard> I read about that some years ago...it was a famous argument....rather well known in computer science community
20:10:22 --- join: adu (andrew@adsl-63-201-89-133.dsl.snfc21.pacbell.net) joined #forth
20:40:09 <TheBlueWizard> one reference: http://www.softpanorama.org/Lang/Java/Tijplus/tijplus03.shtml  another one (search on "Knuth"): http://www.cs.ttu.edu/fase/v9n12.txt   and apparently the most definitive one: http://www.cis.temple.edu/~ingargio/cis71/software/roberts/documents/loopexit.txt
20:40:15 <TheBlueWizard> hope that helps
20:40:20 --- mode: TheBlueWizard set mode: +o adu
20:43:05 <TheBlueWizard> as you can see, this was famous...be sure to carve that into your geek genes ;)
20:44:33 <futhin> heh
20:45:01 <futhin> at the moment as far as i'm concerned edsger dijkstra is cooler than knuth :P
20:49:51 <TheBlueWizard> I think Knuth is cooler....he wrote TeX, and had done a lot of CS stuff...I am truly impressed
21:06:04 <I440r> TBW!!!!
21:06:42 <TheBlueWizard> hiya I440r!!!!  ltnc (c==chat)
21:07:04 <adu> i440R!!!!!!
21:07:23 <adu> yey!
21:12:50 <I440r> :)
21:12:52 <I440r> adu!!!
21:12:53 <TheBlueWizard> so...how're things?
21:12:55 <I440r> im back btw :)
21:12:59 <I440r> back in lafayette
21:13:15 <TheBlueWizard> so you're done with your contract work eh?
21:13:36 --- topic: set to 'gotta find another contract asap :)' by I440r
21:13:58 <adu> cool
21:14:08 <adu> lafayette?!? i live near there!
21:14:16 <I440r> lafayette indiana ?
21:14:25 <TheBlueWizard> heh re: topic
21:14:27 <adu> o n/m i think we talked about this before
21:14:41 <I440r> hehe damned xchat 
21:14:55 <I440r> i go to type in channel and it types in topic :P
21:14:56 * TheBlueWizard lives too far from anyone to be of interest :P
21:15:05 <adu> ahaha
21:15:06 --- topic: set to 'do drop >in' by I440r
21:15:37 <TheBlueWizard> so...are you planning to resume working on your opus (aka IsForth)?
21:16:06 <I440r> already am doing to a degree, trying to colaberate with tcn over email but he dont got much net access at the moment
21:18:00 <adu> cool!
21:18:40 <I440r> fixed a bug in do loops :)
21:18:47 <I440r> hehe but thtas about it...
21:18:55 <I440r> tho tcn actually has it compiling :P
21:19:03 <adu> nice
21:19:10 <adu> i want to write an os in forth
21:19:26 <TheBlueWizard> :)
21:28:17 <I440r> :)
21:28:21 <I440r> forth IS an os :P
21:29:32 <TheBlueWizard> hehe....it can get tricky trying to decide whether it is an OS or not...for example some people thinks Emacs is an OS unto itself :)
21:32:10 <I440r> it is
21:32:16 <I440r> but it lacs a decent editor
21:32:35 * TheBlueWizard laughs
21:33:17 <TheBlueWizard> by the way, I upgraded my Amiga 3000....finally :)
21:34:08 <I440r> hehe cool!
21:35:54 <aaronl> Hehe
21:36:01 <aaronl> i"m setting up a SparcStation2 right now
21:37:31 <TheBlueWizard> I upgraded mine to 8 meg, and hooked up 2G HD to it as well.  Also replaced CMOS battery (the old one leaked acid)....this puppy should last another 5 years, hopefully.  I bought it in '91...amazing machine....AmigaOS and its GUI is renowned for its efficiency
21:38:27 <I440r> i loved my amiga 1200
21:38:42 <I440r> it could draw 16 million pixles persecond with the blitter
21:38:54 <I440r> i dont think a p3 could come close to that
21:40:38 <TheBlueWizard> I don't have A1200 ....mine runs on '030/25MHz
21:41:15 <I440r> yea mine was an 020 at 12 mhz :)
21:42:00 <adu> I440r: so is there something i can do to isforth?
21:42:02 <TheBlueWizard> yeah....and I tried to explain to futhin that graphics accelerator may be needed for decent GUI operations on 386/486s...
21:42:26 <TheBlueWizard> I thought A1200 has a '040....
21:42:40 <I440r> nope
21:43:10 <TheBlueWizard> hmm....interesting
21:43:43 <TheBlueWizard> and best of all, I got my A3000 autographed by 7 former Commodore engineers!
21:43:56 * TheBlueWizard is on cloud 9
21:44:25 <I440r> COOL!!!
21:45:24 <TheBlueWizard> yeah!  among the signees is Dave Haynie (I went to Dave Haynie's house...there was a party going on...and that's how I got it autographed)
21:45:43 <TheBlueWizard> s/ee/er/
21:46:03 <I440r> :)
21:46:06 <I440r> gate crasher :P
21:46:34 <TheBlueWizard> nope...NCAUG people went up to there
21:46:41 <TheBlueWizard> http://www.ncaug.org
21:49:14 * TheBlueWizard just checked out the URL he just mentioned, and it now no longer said anything about Haynie's party....hmmm
21:56:47 <futhin> thebluewizard: i ran windows on a 386 and it wasn't bad..  so with tight code it could be better.  granted, a graphics accelerator would probably be needed for movies and whatnot
22:01:13 <TheBlueWizard> if you see what Amiga can do graphics wise, you'd understand
22:01:53 <TheBlueWizard> I am not saying it would be next to impossible...just that it wouldn't feel smooth with 386 stuff w/o graphics accelerator
22:02:13 <I440r> hehe
22:02:25 <I440r> they do gfx accel with 386 now :P
22:02:33 <I440r> voodo3 isnt a graphics card
22:02:39 <I440r> its just an accelerator realy
22:03:44 <TheBlueWizard> erm...IIRC 386 machines tend not to have PCI stuff, so you can't just put the newest stuff onto old mobo...and even if you can, there'd be bottleneck (slow RAM accesses, etc)
22:04:24 * TheBlueWizard inserts "[...] and expects fast operations" somewhere above
22:13:20 --- quit: adu (IRCStep)
22:14:39 <TheBlueWizard> such is life :)
22:32:41 <TheBlueWizard> got to go....bye all
22:32:44 --- part: TheBlueWizard left #forth
22:55:40 --- quit: futhin (sleep)
23:01:43 --- join: ree (jwm@twisted.goodnet.com) joined #forth
23:18:59 --- quit: ree ()
23:59:59 --- log: ended forth/01.06.09