00:00:00 --- log: started forth/02.11.29
00:15:12 --- join: bwb (~bwb@ip68-4-124-131.oc.oc.cox.net) joined #forth
00:35:12 --- quit: Serg_Penguin (Read error: 104 (Connection reset by peer))
00:35:13 --- quit: bwb (Read error: 104 (Connection reset by peer))
00:36:26 --- join: Serg_Penguin (~Z@nat-ch1.nat.comex.ru) joined #forth
01:11:11 --- join: [crc] (~crc@VA1-1G-u-0216.mc.onolab.com) joined #forth
01:15:27 --- quit: onetom (Read error: 54 (Connection reset by peer))
01:16:09 --- join: onetom (~tom@novtan.bio.u-szeged.hu) joined #forth
02:12:36 --- quit: Serg_Penguin ()
02:31:19 --- join: o- (~o@nzym.fi) joined #forth
02:53:42 --- join: Serg_Penguin (~Z@nat-ch1.nat.comex.ru) joined #forth
02:54:08 --- part: Serg_Penguin left #forth
03:28:59 --- join: Serg_Penguin (~Z@nat-ch1.nat.comex.ru) joined #forth
05:08:09 <Robert> Hi.
05:08:17 <Speuler> 'morning
05:09:11 * Speuler yawns
05:12:08 <Robert> Good "morning".
05:24:30 --- quit: Speuler (Read error: 60 (Operation timed out))
05:37:22 <Serg_Penguin> here morning is no damn good
05:37:52 <Serg_Penguin> cold and high pressure hit
05:38:05 <Serg_Penguin> 2nd mostly for old men w/ ill heart
05:38:53 <Robert> :-/"
05:45:03 <Serg_Penguin> how is at u'r place ?
05:46:18 <Robert> Not that cold, a few degrees below 0.
06:08:18 <Serg_Penguin> here -25 at nite expected, and snow is like srapnel
06:09:45 <Robert> Nasty.
06:16:00 --- join: Speuler (~l@mnch-d9ba4e6c.pool.mediaWays.net) joined #forth
06:16:17 <Serg_Penguin> just from outdoors (we sit in basement, go to WC via street into main door) - it's horrible
06:16:56 <Speuler> sounds better than to shit into the river
06:17:14 <XeF4> shit in a bucket and carry it to the WC when it is full?
06:17:49 <Serg_Penguin> no, no =8-(
06:18:11 <Serg_Penguin> =8<(
06:18:21 <Serg_Penguin> < is curtain crimple on the nose
06:18:33 <XeF4> 0D
06:18:39 <Serg_Penguin> 0D ?
06:18:43 <XeF4> er =)
06:18:46 <XeF4> er =D
06:18:52 <Robert> :)
06:20:37 <Speuler> freeze-dry and powderize 
06:20:52 <Serg_Penguin> if u piss at street now u may get a fancy arc of yellow ice, and get u'r frozen pepper sound like 'tuxracer' game soundtrack
06:21:53 * XeF4 has only seen that when it was <-30
06:22:46 <Speuler> remember nanook's mother's advice
06:23:28 <Serg_Penguin> ???
06:23:50 <Speuler> that was "watch out where the huskies go, and don't you eat that yellow snow"
06:25:01 * XeF4 wants -25
06:25:21 <XeF4> not *now*, because I'm almost out of firewood
06:25:41 <XeF4> but soon
06:34:41 --- quit: Robert (forward.freenode.net irc.freenode.net)
06:35:25 --- join: lament (~lament@24.78.145.92) joined #forth
06:37:27 --- join: Robert (~Robert@robost86.tsps1.freenet6.net) joined #forth
07:00:59 --- quit: Serg_Penguin ()
07:41:50 --- join: neobrat (neobrat@0-1pool78-159.nas6.salt-lake-city1.ut.us.da.qwest.net) joined #forth
07:42:05 <neobrat> howdy, all! Hope everyone had a happy Thanksgiving!
08:13:51 --- join: ASau (~asau@158.250.48.197) joined #forth
08:14:18 <ASau> Good evening!
08:16:59 <Robert> Hello.
08:18:12 <Speuler> tach
08:18:14 <ASau> Today I've found a way to compile an app. to run it separate from Forth.
08:18:42 <ASau> Anyone interested?
08:18:59 <Robert> Hm?
08:19:09 <Robert> Tell us.
08:19:51 <ASau> You have: Forth with access to mass storage (blocks or files).
08:20:12 <ASau> You can load a binary and run it.
08:20:38 <ASau> So you want to get a binary image ready to run.
08:21:00 <ASau> Main idea is not to write a true compiler,
08:21:25 <ASau> but to write something, that compiles itself into binary image.
08:21:57 <Robert> How would it do that?
08:22:51 <ASau> If you design bottom-up, than you have one hook with forward pointer
08:22:53 <Speuler> cat interpreter.src | virtualmachine
08:22:56 <Speuler> :)
08:23:27 <ASau> (main entry point),
08:23:56 <ASau> and the rest is defined before it is used.
08:24:23 <Speuler> echo -e "\opcode\opcode\opcode.." > virtualmachine
08:24:38 <ASau> Yes, something like that.
08:25:00 <ASau> But you can use this:
08:26:16 <ASau> : : <BUILDS TARGET CREATE HERE MY , DOES> MY @ TARGET , ;
08:26:50 <Speuler> ASau: you use fig or f79
08:26:52 <Speuler> ?
08:26:52 <ASau> Your new "colon" does 2 main things:
08:27:06 <ASau> Sp.: fig.
08:27:14 <Speuler> veteran ?
08:27:50 <Speuler> or f83/ansi unwilling ?
08:28:00 <ASau> No, that is the most simple Forth I am able to find in source,
08:28:19 <ASau> so I am able to warp it any way I like.
08:28:32 <Speuler> which os do you use ?
08:28:44 <ASau> DOS :)
08:28:57 <Speuler> i might have something for you there
08:29:18 <Speuler> a metacompiling kernel, which can be extended to full forth :)
08:29:26 <Speuler> instead of the other way around
08:29:31 <ASau> I don't like all that C-written forthes of ANS standard.
08:29:52 <Speuler> loading forth source into the metacompiling kernel is actually done by command line redirection
08:29:59 <Speuler> cause kernel has no file i/o
08:30:16 <ASau> I use it already :)
08:30:42 <Speuler> kernel compiles its own source without adding any source file ...
08:31:01 <ASau> Forth src.: .... SAVE <<EOLN>><<EOF>> :)
08:31:21 <ASau> Another way:
08:31:48 <ASau> HEX xxxx IN !
08:31:50 <ASau> :)
08:32:27 <ASau> Too many ways to go without file i/o.
08:32:39 <Speuler> that still requires you to get the source into mem
08:33:05 <ASau> copy /b forth.com + my.f temp.com
08:33:07 <ASau> :)
08:33:31 <Speuler> how'd you re-create forth.com ?
08:34:02 <Speuler> i'd do that by "type kernel.src | kernel.com"
08:34:42 <ASau> CREATE (SAVE) HEX 4x C, ( ... ) E9 C, NEXT HERE - 2+ ,
08:34:54 <Speuler> (not called forth.com because it aint a forth at that point)
08:35:00 <ASau> : SAVE BL WORD HERE (SAVE) ;
08:35:04 <ASau> SAVE FORTH.COM
08:35:47 <Speuler> no. you SAVE it that way, but you don't GENERATE it
08:35:56 <ASau> Since ZX it is quite usual to write in code. :)
08:36:00 <Speuler> i GENERATE kernel this way
08:36:24 <ASau> I looked at postForth.
08:36:27 <Speuler> or you would have to c, the whole kernel
08:36:38 <Speuler> that's a far cry from metacompiling :)
08:36:43 <ASau> Asm rules :)
08:37:50 <ASau> The way of "type my.f | forth.com" is known to me.
08:38:25 <ASau> One addition: "-lf my.f | forth.com"
08:38:26 <Speuler> but probably not a kernel which is a metacompiler w/o any additions
08:38:46 <Speuler> at least, i don't know any other
08:38:57 <ASau> Read the fig installation manual.
08:39:35 <Speuler> what for ?
08:39:36 <ASau> There is a distinctive feature of figF.,
08:39:50 <ASau> one asm listing for kernel.
08:40:25 <Speuler> that's not a property for fig alone
08:41:01 <ASau> I've not said it is feature of figF. only.
08:41:37 <Speuler> so you use an assembler to create the kernel. i use a kernel to create a kernel
08:41:57 <ASau> Something has to be first.
08:42:07 <ASau> Binary kernel or asm. kernel.
08:42:19 <ASau> I prefer asm.
08:42:22 <Speuler> co-evolution
08:42:34 <Speuler> starting out from asm
08:42:55 <ASau> As I can change asm kernel and find the best suited for me.
08:43:21 <Speuler> in my case, the kernel is the assembler too
08:43:34 <Speuler> well, subset
08:43:44 <Speuler> just what's required for the kernel
08:44:28 <ASau> Of course kernel can not use the whole command list :)
08:44:34 <Speuler> the mnemonics look like  "(ip+)>w" and the like
08:44:58 <ASau> Do you use preprocessor?
08:45:01 <Speuler> nope
08:45:21 <Speuler> : (ip+)->w   opcode c, opcode c, ... ;
08:45:37 <ASau> I 've undestood.
08:45:46 <Speuler> many primitives have been "virtualized" that way
08:46:19 <Speuler> again:  ONLY the kernel is required to compile the kernel source
08:46:44 <Speuler> no src massaging by additional programs
08:46:55 --- quit: lament ("mental mantle")
08:47:23 <ASau> You're using precompiled binaries to get the kernel. AIR?
08:48:07 <Speuler> i use the result of a kernel compilation to compile a kernel. no foreign code at all
08:48:25 <ASau> That is like F83, I've seen somewhere around.
08:48:49 <Speuler> nope. f83 requires you to load the metacompiler source before you can create a kernel
08:48:58 <Speuler> i use the kernel to compile the kernel source
08:49:15 <Speuler> the kernel IS the metacompiler
08:49:28 <Speuler> and the metacompiler can be extended to become a full forth :)
08:49:47 <Speuler> F83 full forth is extended to become the metacompiler.
08:50:36 <ASau> OK. And what's your OS?
08:51:02 <Speuler> it was developed for a system w/o os, and was ported to dos later on
08:51:37 <ASau> I'm going another way: DOS -> own OS.
08:52:03 <ASau> Read: "own OS" = "Forth OS".
08:52:19 <Speuler> you probably came to the right place, there are or were many forth os activities here
08:52:30 <Speuler> at one point, 5 or so
08:52:53 <Speuler> know valdocs ?
08:52:56 <ASau> Do you have your Forth OS ready?
08:53:09 <ASau> valdocs?
08:53:18 <ASau> What's it?
08:53:29 <Speuler> i'm not of the forth os conspiration
08:53:41 <Speuler> but i like stand-alone forths
08:54:00 <Speuler> os-less forths
08:54:09 <Speuler> you could call those forths "os"
08:54:26 <Speuler> as it takes over the function traditionally offered by an os
08:54:34 <Speuler> and bios
08:54:56 <Speuler> valdocs might have been the first forth-os attempt
08:55:00 <Speuler> epson
08:55:12 <Speuler> about 20 years ago
08:55:39 <Speuler> that long after forth has been used stand-alone on different machines
08:56:08 <Speuler> what is the diff between a forth os and a stand-alone forth ?
08:56:15 <ASau> I use to equal "Operating System" and "Working Environment".
08:56:29 <ASau> So os-less Forth is Forth OS.
08:56:58 <Speuler> i don't know. i don't tend to call stand-alone forth forth-os
08:57:20 <Speuler> and i assume there must be something more to it
08:57:28 --- join: proteusguy (~username@65.191.88.177) joined #forth
08:57:32 <ASau> I met a classification of OSes where the special class of OS is mentioned
08:57:59 <ASau> One prog.lang. OS, or OS-PL.
08:58:20 <Speuler> an os should allow you to load and execute binaries
08:58:25 <Speuler> or ?
08:58:48 <ASau> E.g., stand-alone Forth, Basic on MSX, Basic on XT, Basic on ZX Spectrum etc.
08:58:51 <Speuler> would loading and compiling, running from source meet that too ?
08:59:32 <ASau> LOAD "" :)
09:00:16 <ASau> Why binaries.
09:00:30 <Speuler> distribution/exchange
09:00:46 <Speuler> loading speed
09:00:50 <ASau> IIRC, Lisp-Ms had no binaries other than LISP.
09:02:36 <ASau> (Your Forth allows you do this: xx BLOCK EXECUTE )
09:03:05 <Speuler> it would if block support would be loaded
09:03:34 <Speuler> but
09:03:46 <Speuler> then you run xx block execute again
09:03:49 <ASau> CREATE R/W HEX xx C, xx C, xx C, ( ... ) R/W
09:04:02 <Speuler> you "shell" to another program ...
09:04:09 <Speuler> yy block execute
09:04:28 <Speuler> where would that binary be loaded ?
09:04:41 <Speuler> how, if not on known address, would it be relocated ?
09:04:45 <ASau> PF, of course.
09:04:48 <Speuler> writing relocatable code ?
09:04:57 <Speuler> not the domain of usual forth
09:05:27 <ASau> Use absolute code. Or relocate in compile time.
09:06:25 <ASau> CREATE R/W HEX xx C, ORIGIN xxxx + , ( ... ) 
09:06:35 <Speuler> ok.compile prog1. run. shell. compile prog2. run   .  start prog3.  forget prog1 and 2 ?
09:07:19 <ASau> At your opinion.
09:07:42 <ASau> You may chase of both.
09:07:52 <ASau> ...choose...
09:08:50 <Speuler> wouldn't work
09:09:00 <ASau> ??
09:09:11 <Speuler> : word1 foo ;   : word2 bar ;    forget word1
09:09:17 <Speuler> word2 ????
09:09:40 <ASau> : word1 ; forget word1 : word2 ;
09:09:54 <Speuler> ah. a single-tasking system :)
09:10:12 <ASau> It is if you use only pre-defined words.
09:10:19 <ASau> Define your own.
09:11:05 <ASau> Some USER-spaces will save you. :)
09:11:24 <Speuler> how would you get rid of compiled programs, started before another one was started ?
09:11:59 <ASau> One way is to make one more USER-space.
09:12:00 <Speuler> can forget them only after the following processes have been quit too ?
09:12:15 <Speuler> FSLQ 
09:12:30 <Speuler> first start last quit :)
09:12:32 <ASau> FENCE and DP may be USER variables.
09:12:47 <Speuler> how could that help ?
09:13:21 <Speuler> problem is not solved by just compiling to several voc threads
09:13:22 <ASau> You'll have tree-organized splitted dictionary.
09:13:41 <Speuler> you have / can have those with vocabularies too
09:14:03 <Speuler> implementation dependent
09:14:21 <ASau> Everything is implementation dependent.
09:14:43 <Speuler> interesting talk but i need to cut it short. need to put food on the fire
09:14:50 <ASau> Some implementations are just close enough to see no (or small) defference.
09:14:53 <Speuler> people will come soon, 
09:15:14 <Speuler> don't want to start cooking when they've arrived already
09:15:32 <ASau> See on the net?
09:15:39 <Speuler> i'll be around
09:15:49 <Speuler> regular visitor of this #
09:16:00 <ASau> OK.
09:16:04 <Speuler> l8er
09:16:51 <ASau> Bye!
09:26:26 --- quit: neobrat ()
09:44:47 <ASau> Bye, all!
09:44:49 --- quit: ASau ("Toffee IRC client for DOS v1.0/b535")
09:51:32 --- join: lament (~lament@24.78.145.92) joined #forth
10:52:50 --- quit: skylan (Connection reset by peer)
10:53:16 --- join: skylan (sjh@207.164.213.121) joined #forth
11:21:02 --- quit: proteusguy (Read error: 104 (Connection reset by peer))
11:21:09 --- join: proteusguy (~username@65.191.88.177) joined #forth
11:24:12 --- quit: lament ("mental mantle")
11:25:05 --- quit: XeF4 ("brb")
11:26:17 --- join: XeF4 (xef4@lowfidelity.org) joined #forth
11:29:34 --- quit: [crc] (Read error: 60 (Operation timed out))
11:30:48 --- join: galexand (~galexand@adsl-18-2-209.rdu.bellsouth.net) joined #forth
11:41:38 --- join: Robert_ (~Robert@robost86.tsps1.freenet6.net) joined #forth
11:42:03 --- quit: Robert (Remote closed the connection)
11:57:06 --- join: proteus_ (~username@65.191.88.177) joined #forth
11:57:34 --- quit: proteusguy (Read error: 110 (Connection timed out))
11:59:08 --- quit: proteus_ (Client Quit)
13:25:41 --- nick: Robert_ -> Robert
13:59:06 --- join: lament (~lament@24.78.145.92) joined #forth
13:59:23 --- join: proteusguy (~username@65.191.88.177) joined #forth
14:25:41 --- quit: lament ("mental mantle")
14:53:07 --- part: Speuler left #forth
15:23:02 --- join: bwb (~bwb@ip68-4-124-131.oc.oc.cox.net) joined #forth
15:49:57 --- join: joonas (jpihlaja@kruuna.helsinki.fi) joined #forth
15:51:34 * joonas tips his hat
15:51:40 <joonas> Evening all.
16:03:17 <bwb> evening
16:03:58 <joonas> ah, there is life in #forth yet.
16:05:14 <joonas> Have you every used Valgrind?  It's the best debugger I've ever seen!
16:05:19 <joonas> every -> ever
16:07:19 <joonas> http://developer.kde.org/~sewardj/
16:09:21 <joonas> It basically runs your program in a VM that fully emulates an x86 and catches pretty much all memory related errors.
16:10:24 <joonas> Actually, it's more of a instrumenting JIT compiler as it disassembles, instruments loads/stores, and reassembles the basic blocks in the code.
16:12:53 <joonas> It's The Fonz of debuggers. :)
16:16:37 <fridge> haha
16:19:34 <joonas> One thing tho: you have to have a fast machine to use it on a "real" program, as it slows down execution a lot.
16:20:14 <fridge> I'd probably only use it if all other attempts to debug a problem failed
16:20:22 <fridge> so it wouldn't be much of an issue =)
16:21:46 <joonas> Yes definitely.  I usually don't have memory related errors at all, but for the odd really spurious bug (you know the kind I'm talking about), it's invaluable.
16:34:45 <joonas> Has any one played Core War?
16:40:10 <XeF4> once, with some example code, but never against anyone else
16:41:09 <joonas> It's still an active game.  There's a tournament going on right now, if you're interested.
16:42:00 <XeF4> too busy
16:49:56 --- quit: Fractal (Read error: 110 (Connection timed out))
16:51:46 --- join: lament (~lament@h24-78-145-92.vc.shawcable.net) joined #forth
16:52:16 <joonas> hi lament
16:56:35 <onetom> joonas: ive seen & tried cold war but never seriously tried it
16:57:03 <onetom> it was programmable in red, right?
16:57:23 <joonas> (sorry, I wasn't paying attention...)  yes, redcode.
16:58:19 <joonas> The idea is to make a program that tries to crash an opponent program.
16:59:03 <onetom> sure, i saw a dos version what also visualized the state of war
16:59:56 <lament> hi
17:00:28 <lament> hm, that's the third time someone mentions core wars in two days.
17:01:14 <galexand> last i looked at the corewar tourney it looked like it had degenerated into rock paper scissors.  i was nonplussed.
17:02:10 <lament> corewars tournaments are dominated by evolved programs
17:02:29 <lament> not hand-coded ones
17:02:29 <joonas> lament: where did you learn that?
17:02:31 <galexand> oh really?  it must have advanced since i last looked :)
17:02:38 <joonas> lol
17:02:40 <galexand> hmm let's see 1998 was a while ago
17:02:46 <lament> joonas: corewars.sf.net mostly
17:03:11 <joonas> lament: you're looking in the wrong place.  Try www.ociw.edu/~birk/COREWAR/
17:03:19 <joonas> Also www.koth.org
17:03:30 <lament> joonas: How's it the wrong place?
17:05:21 <galexand> oh maybe the 'right place' is dominated by hand coded rocks papers and scissors :)
17:05:28 <lament> hehe
17:05:43 <lament> well, the problem with corewars.sf.net is that most tournaments use some language called 'corewars'
17:05:43 <joonas> lament: I was too harsh.  It's the right place, except that the Corewars project on sf.net implements a different game that isn't what was described in SciAm.
17:05:58 <lament> but they have redcode tournaments, too
17:06:13 <joonas> to be fair, it *also* implements redcode.  well, most of it anyway.
17:06:19 <lament> Yes.
17:07:29 <joonas> You should note that sf.net also has tournaments for purely evolved warriors.  Hand written ones are disallowed from them, because they would dominate the tournaments.
17:08:25 <joonas> IMO, that's a good thing.  People genuinely interested in evolving warriors should have a venue to battle each other.
17:13:22 <joonas> Anyway, just for the record, the rock-paper-scissors analogy doesn't really hold:  Rock has pretty much vanished, being replaced by...uh, boulders.
17:13:41 <joonas> Paper is engineered to go head to head with scissors.
17:13:53 <Robert> Good evening.
17:13:59 <joonas> evening
17:14:06 <galexand> hahah i see joonas :)
17:14:24 <galexand> joonas i want to learn some game like that.   should i learn redcode or chess?
17:14:33 <galexand> i don't think i have the willpower to learn both
17:14:35 <joonas> both, definitely. :)
17:14:42 <galexand> damn heh
17:14:46 <joonas> ok, go with chess: more players.
17:15:00 <galexand> yeah i was leaning that way too its' good to be able to find an opponent in every big city
17:15:06 <galexand> er a bunch of opponents
17:15:21 <lament> hmm, have you read the review of chess as a strategy game?
17:15:32 <galexand> no
17:15:34 <joonas> no
17:15:41 <galexand> but i could put redcode on my ipaq and then you know :)
17:15:55 <lament> read it!
17:15:56 <lament> http://homepages.apci.net/~ipw4/bigmuddy/chess.html
17:16:00 <joonas> cheers
17:16:14 * joonas goes to read it
17:16:15 <lament> That review it's funny mostly because it's so true.
17:16:21 <lament> s/it's/is
17:16:39 <galexand> do you know any decent on-line introduction to actually playing chess?
17:17:06 <galexand> i mean i know how all the pieces move but i want to know what the french-dude's-name gambit is
17:17:35 <galexand> oh heh i have read the review.  it is totally accurate :)
17:19:14 <joonas> wow, thanks for that!
17:30:54 <lament> galexand: Don't learn Chess; learn Go
17:43:04 <joonas> Any opinions on the Forth Scientific Library?
17:47:47 <lament> How few words can a forth have and still be able to build the other ones from them?
17:50:56 <joonas> http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=4b1ffa1f.0112200314.3118823%40posting.google.com
17:52:17 <lament> thanks
17:55:23 <joonas> Are you building a forth implementation?
17:55:26 <lament> damn, it's an old thread and the links are all broken :(
17:55:34 <lament> Not sure.
17:55:58 <joonas> heh, sorry about that.
17:56:33 <lament> and they recommend eforth, which is of course good, but there's no freely available info about it
17:57:42 <lament> There's pretty much no free information about Forth, anywhere :(
17:58:40 <joonas> Check out the "Moving Forth" articles and "Build your own (Forth) assembler" articles by Brad Rodriguez
17:59:08 <lament> Yeah, i've seen that
17:59:15 <lament> But I'd rather learn Forth first :)
17:59:44 <joonas> Oh, in that case you should read the GForth 0.5.0 manual.
18:00:10 <joonas> Also, the book "Forth Programmer's Handbook" from Forth Inc. is good (though not free.)
18:00:24 <joonas> The gforth manual is browsable on line at:
18:00:31 <lament> Yeah, i've read it
18:00:43 <lament> Everyone says it sucks, and i believe them :)
18:00:57 <joonas> http://www.complang.tuwien.ac.at/forth/gforth/Docs-html/gforth.html
18:01:05 <joonas> Oh.
18:01:13 <joonas> I don't think it sucks.
18:02:29 <joonas> I found the "Tutorial" and "Intro to ANS forth" life savers.  And of course the word lists/descriptions.
18:02:42 <fridge> the intro was ok
18:03:03 <fridge> I thought the tutorial a bit ... can't think of the right word
18:03:10 <joonas> "condescending"?
18:03:48 <lament> Have you read "eForth and Zen"?
18:03:49 <joonas> I particularly remember the sentence that starts with "For the math whizz out there[...]".
18:04:05 <joonas> no.  do you have a link?
18:04:11 <lament> no.
18:04:19 <lament> only the first chapter is available online.
18:04:19 <joonas> found it.
18:04:33 <lament> which is pretty much an introduction, and says nothing about eforth itself.
18:05:43 <lament> I'm pretty sure that book is exactly what I want.
18:06:34 <joonas> Actually, I think I've read the intro a long time ago, but didn't find it useful.
18:06:44 <lament> Well, it's not
18:06:54 <lament> But the rest of the book actually describes eforth in detail
18:08:42 <joonas> What's your plan to do with it?
18:10:23 <lament> Nothing, since i won't ever read it :)
18:11:59 <joonas> I meant with forth?
18:12:21 <lament> Oh.
18:12:23 <lament> Not sure.
18:15:01 --- join: Speuler (~l@mnch-d9ba4e6c.pool.mediaWays.net) joined #forth
18:15:49 <joonas> hi Speuler
18:16:00 <Speuler> hello joonas
18:16:49 <Speuler> huevae huomenta :)
18:17:34 <joonas> Kiitos samoin. 
18:19:39 --- quit: proteusguy (Read error: 104 (Connection reset by peer))
18:21:42 <joonas> Speuler: you're using debian?
18:22:05 <Speuler> that's right
18:22:16 <joonas> woody?
18:22:22 <Speuler> sarge
18:22:34 <joonas> Only I'm still on potato and am considering the switch.
18:22:35 <joonas> oh
18:22:37 <Speuler> well, mix from woodz and sarge
18:22:55 <Speuler> there's possibly not much left of woody
18:23:01 <joonas> has dselect gotten any better?
18:23:05 <joonas> what do you mean?
18:23:26 <joonas> oh, on your setup.
18:23:29 <galexand> dselect sucks esp. if you're mixing dists like speuler is :)
18:23:54 <Speuler> one difference of dselect is now that you don't press space but enter on startup to skip the help page
18:24:29 <Speuler> galexand: bullexcrement
18:24:52 <Speuler> galexand: you just need to put all packages on hold
18:25:10 <Speuler> then you pick the ones you really want to upgrade
18:25:28 <galexand> speuler, heh okay :)
18:25:33 <galexand> speuler, but what about its 300 library dependencies?
18:25:38 <galexand> 90% of which are wrong
18:25:44 <galexand> heh
18:25:48 <Speuler> joonas: there are other frontends to dpkg than dselect
18:26:03 <galexand> i've never had a debian dist work right with apt-get/dselect for more than a month
18:26:11 <Speuler> galexand: you pick one, if one dep is required, it will be updated too
18:26:21 <galexand> i always wind up installing something that depends on some private build of libc that got into the devel tree anyways
18:27:07 <Speuler> galexand: first debian i used was 1.2
18:27:31 <galexand> i had a box all the way from 0.93rc6 to 2.0 i think and it is definitely possible you just abandon dselect somewhere.
18:27:32 <Speuler> have installed debian on at least 50 machines 
18:27:45 --- join: Fractal (tytptc@h24-77-171-228.ok.shawcable.net) joined #forth
18:27:50 <Speuler> can't remember a single one where it didn't work straight away
18:28:04 <galexand> er all the way to 2.1 heh
18:28:07 <galexand> oh it always works the first time
18:28:24 <Speuler> galexand: 0.93 was probably linux kernel, not debian
18:28:32 <galexand> like this box i installed from cds 2.2.  then i eventually allowed dselect to do the full upgrade to whatever came next, i assume 2.3
18:28:40 <Speuler> that was about 1992 or 93
18:28:44 <galexand> speuler, i did every upgrade between 0.93 to 2.1, i didn't jump straight in
18:28:53 <galexand> um no dude it was 1995
18:28:54 <Speuler> galexand: ah ok. 
18:29:11 <Speuler> galexand: you're more of a debian veteran than i am
18:29:20 <galexand> i had a 170MB HDD with debian 0.93rc6 on it and i copied that straight over to a 3.2G that i never reformatted. :)
18:29:21 <Speuler> first i run SLS
18:29:26 <Speuler> then slack
18:29:30 <Speuler> then debian
18:29:33 <galexand> i am not nearly patient enough for anything slackware.  i
18:29:45 <galexand> i really like the packages, the dependency system just always gets broken and i always wind up doing everything with dpkg --force-depends by hand
18:29:47 <Speuler> dabbled with suse, but returned to debian quickly
18:29:53 <galexand> but even so it's very convenient.  i still stick with debian after all this :)
18:30:19 <galexand> the trouble is that i always wind up installing one program from the devel branch and then it tries to pull in a bunch of libraries and i usually let the one program do that
18:30:36 <galexand> but then the next program tries to pull in a new version of glibc on top of the version i downloaded two weeks ago and i just go HARUMPH and forget about it. :)
18:30:40 <Speuler> i've put everything on hold
18:30:51 <galexand> i should try putting things on hold.  the hold system is very nice.
18:31:13 <Speuler> cause i don't want to upgrade those packages i didn't explicitely choose
18:31:47 <joonas> I find the procedure to hack on the source of existing packages confusing.
18:31:50 <Speuler> if i don't like the upgrade of the dep packages, i do R and forget about it
18:32:09 <galexand> i really like dpkg-source :)
18:32:10 <galexand> wheni t works
18:32:28 <Speuler> usually i use apt-get
18:32:38 <galexand> the biggest problem facing debian is that it records too much version dependency...so a lot of the devel tree, esp. the more obscure parts of it, has dependencies on versions of glibc and other libraries that are just private versions.
18:32:48 <Speuler> but for checking out new packages, i use dselect
18:33:37 <galexand> speuler, try installing nmh :)
18:33:37 <joonas> heh, #forth is hijacked by #debian :)
18:33:56 <galexand> at 2.3 or devel i don't remember which nmh is completely broken because it depends on libraries that don't exist
18:34:49 <Speuler> i used nmh a while ago
18:35:26 <Speuler> galexand: right because #forth knows what is good for you :)
18:35:43 <Speuler> galexand: actually, i do forth too 
18:36:39 <Speuler> s/galexand/joonas/
18:36:40 <galexand> i wonder if they've fixed the devel nmh package hmm
18:36:45 * galexand does forth too
18:36:48 <galexand> apt-get install gforth
18:37:21 <Speuler> debian gforth package is not complete, need apt source gforth too
18:37:27 --- quit: lament ("pathetic")
18:37:27 <galexand> i'm contemplating portinf hforth to run on arm-linux on my ipaq
18:37:31 <Speuler> or no make
18:37:45 <Speuler> nice
18:37:45 <galexand> i really like the hforth bootstrapping technique
18:37:51 <galexand> (same as eforth i think)
18:37:52 <joonas> The 0.4.9 version was complete, except that it had the old docs.
18:38:14 <joonas> What's the hforth bootstrapping technique?
18:38:35 <Speuler> didn't notice 0.4.9 in debian. thought they went straight from 0.4.0 to 0.5.0
18:39:03 <galexand> the core words for hforth, including everything needed for : are all written in a standard macro-assembler
18:39:06 <joonas> It's in potato.
18:39:16 <Speuler> i don't have debian gforth installed, instead compiled, placed into /usr/local
18:39:16 <galexand> the macro-assembler builds up the dictionary and all
18:39:51 <joonas> yes, that's a nice way to go about it.
18:41:09 <joonas> One downside is that you end up with a lot of words written in "forth" in the macro assembler... unless you have a cross-compiler written in forth too.
18:41:50 <galexand> yeah it's unfortunate.  the macro-assembler makes it relatively easy to generate normal 'forth' words (direct threaded), but it is awkward.
18:42:27 <galexand> i only like it because you can see how everything goes the first time you write the system.  otherwise it is very daunting to try to write something targetted at real meta-compilation from the start
18:42:57 <joonas> agreed.  It's daunting even afterwards.  :)
18:43:49 * joonas curses std::string to the deepest pits of hell!
18:44:18 <joonas> (I'm tracking N+1 memory leaks due to std::string constructors atm.)
18:44:44 <galexand> hahahah C++
18:44:47 <galexand> i'm sorry dude
18:45:19 <joonas> Thanks.  I'm open to all the sympathy I can get. ;)
18:45:55 <joonas> Here's a quick-and-dirty way to do tail-call optimisation in a compiler:
18:46:33 <joonas> When compiling, assume all words are non-tail-call optimisable if you don't have access to its name field.
18:47:03 <joonas> If you have access to its name field, then check a bit there to see if it is "blacklisted" from tail-call optimisation.
18:47:27 <joonas> This would include all words that take arguments in the program thread and things that touch the return stack.
18:48:25 <joonas> After compiling a word, set a flag somewhere that indicates its tail-optimisability.
18:48:49 <joonas> when compiling EXIT, check that flag and convert EXIT into your preferred form of tail-call.
18:49:54 <joonas> This method avoids most of the analysis during compilation about whether or not a call is optimisable by placing a (small) burden on the user to mark non-tail-callable words.
18:50:18 <galexand> i don't see why any word would need to be blacklisted.
18:51:25 <joonas> You wouldn't want to tail-call words that use the return stack.
18:51:30 <joonas> e.g. >R
18:52:06 <joonas> or words that have arguments, like the primitive that implements a literal.
18:52:20 <joonas> I'm assuming a threaded forth here.
18:53:13 <galexand> well i assume in every word that uses the return stack that >R and R> are balanced
18:53:15 <galexand> aren't they?
18:53:52 <joonas> Not necessarily.
18:53:58 <galexand> oh :)
18:54:01 <galexand> so just a couple exceptinos
18:54:02 <joonas> Hang on... I'm digging up an example.
18:54:14 <galexand> also i thought the literal words and stuff like that were all immediate
18:55:59 <Speuler> galexand: "immediate" refers to compile-time action.
18:56:12 <Speuler> galexand: we're looking at run-time behaviour now
18:57:48 <joonas> : (s")( -- str len)
18:57:48 <joonas>         r>  dup count   ( ip str len)
18:57:48 <joonas>         rot 1+ +        ( str len ip+len+1)
18:57:48 <joonas>         >r
18:57:54 <joonas> ;
18:58:20 <joonas> argh, insert a space before the stack effect.
18:58:27 <joonas> this would be compiled as:
18:58:35 <joonas> dc.l (s")
18:58:50 <joonas> dc.b  4, "abcd"
18:59:07 <joonas> dc.l type
18:59:25 <joonas> for the fragment s" abcd" type
19:00:03 <Speuler> joonas: ansi forgets access of return stack elements which haven't been put there by the same word which access them
19:00:17 <Speuler> s/forget/forbid/
19:00:42 <joonas> right I wouldn't do that in standard code.
19:01:35 <joonas> the point was just that this would be an example of a word inside a forth that shouldn't be tail-called.
19:02:53 <joonas> hmm.. I guess my explanation skills could do with a polish.
19:03:54 <joonas> Right, next topic please.
19:06:41 <galexand> joon, oh.4
19:06:45 <galexand> joonas, thanks
19:08:02 * joonas fetches coffee
19:16:32 --- join: I440r (~mark4@sdn-ap-010tnnashP0280.dialsprint.net) joined #forth
19:16:40 <fridge> was someone working on a forth syntax file for nedit?
19:16:46 <I440r> bongo! :)
19:16:51 <I440r> hi ppl
19:17:44 <I440r> doing syntax highlighting in forth is next to impossible
19:18:30 <Speuler> hi I440r 
19:18:36 <I440r> :)
19:18:39 <Speuler> ltns
19:20:07 <I440r> ya
19:20:21 <Fractal> It's largely because syntax doesn't really exist in forth.
19:20:31 <I440r> ive been afk, working on some code my father inherited at work
19:20:54 <I440r> plus i did a little isforth bugfixing :)
19:21:06 <Fractal> BTW, I've got an interesting namepsaces implementation working now. Let me post a link. One sec.
19:21:09 <I440r> found a bug in alias
19:21:29 <I440r> if you created an alias to an alias to a headerless word it would crash
19:23:38 <Speuler> ah
19:23:47 <Speuler> attempt to  0 @
19:23:57 <I440r> no, thats not a bug :P
19:24:17 <Speuler> as result of aliasing noname alias
19:24:25 <I440r> if you try access memory that doesnt belong to you thats a bug in YOUR code, not in the compiler :P
19:24:50 <Fractal> http://www.hcsw.org/frugal/namespaces.txt
19:25:11 <Fractal> Kind of an interesting idea, if I do say so myself. Should be easy to implement in other forths too.
19:25:59 <Fractal> Also, frugal runs on OS X and SunOS/SPARC now.
19:36:36 <galexand> wow i never thought of it that way.  0 0 ! vs *(int *)0 = 0;
19:36:41 <galexand> it's easier to crash forth :)
19:37:09 <I440r> lol
19:37:12 <I440r> 0 off
19:37:21 <I440r> off being a 0 !
19:39:23 <Fractal> You mean 0 swap !, I assume
19:39:47 <galexand> fractal, i have to say your pong ui sucks
19:39:52 <galexand> :)
19:39:58 <Fractal> Oh. Why's that?
19:40:06 <galexand> i don't like the 'intertia'
19:40:24 <I440r> pong ui ?
19:40:28 <Fractal> It'd be boring without it. :)
19:41:13 <galexand> fractal, is frugal in indirect threaded forth with a C bootstrap?
19:41:20 <Fractal> Essentially you have an x velocity as well as a y velocity. Once you think about it that way.
19:41:21 <galexand> er a direct threaded i guess
19:41:29 <Fractal> galexand : Yes, more or less.
19:41:49 <galexand> my y velocity is zero so i don't want to think of it that way :)
19:41:50 <Fractal> You can read more about the implementation in docs/INTERNALS
19:42:04 <galexand> thanks fractal
19:42:09 <Fractal> NP
19:42:16 <galexand> didn't even notice docs/ :)
19:42:48 <Fractal> Actually, there are a lot of docs/ about stuff. docs/USING is decent.
19:42:49 <galexand> i like the fact that you appear to have immediate words, it makes your forth more like a C bootstrapped forth and less like an interpretter for some coincidentally stack-based language
19:43:10 <Fractal> :)
19:43:27 <Fractal> What forths don't have immediate words?
19:44:00 <onetom> fridge: use vim. it has a forth syntax hilite file ;)
19:44:10 <Speuler> Fractal: cmforth
19:44:22 <galexand> fractal, interpretted coincidentally stack-based languages :)
19:44:46 <fridge> onetom: I just switched after realising how mouse dependant nedit is
19:45:01 <fridge> which is okay on a desktop, but I hateusing the trackpad thing on this laptop
19:45:09 <onetom> hehe
19:47:23 <Fractal> Are you sure cmforth doesn't use immediacy?
19:47:46 <Speuler> Fractal: it uses another approach
19:47:47 <galexand> cmforth uses a FORTH dictionary and a COMPILER dictionary and "immediate words" get put in the COMPILER dictionary
19:48:02 <galexand> pygmy forth copied that approach
19:48:09 <Speuler> right
19:48:26 <Speuler> compiler voc is only searched during compilation
19:48:27 <Fractal> Oh, I see. That makes sense.
19:48:47 <Speuler> "immediate" words may be defined twice
19:49:12 <galexand> ooh i didn't even think of that implication
19:49:32 <galexand> so we could make a version of FOR in the FORTH dictionary :)
19:49:43 <Speuler> that's the equiv of state-smart words
19:50:25 <galexand> heh i remember years ago reading about state-smart words on MISC-list but this is the first time someone's said it and i've been able to figure out what they meant
19:50:56 <Speuler> galexand:   : foo   compiling if  dothis else dothat then  ; immediate
19:51:14 <Speuler> : compiling  state @  ;
19:51:27 <galexand> hmmm i'm tempted to make a reverse-polish language that is compiled in the classic sense.
19:51:29 <Speuler> therefore state smart
19:53:02 <Speuler> in general it is recommended to avoid creating state-smart words
19:53:11 <galexand> yeah i see that
19:55:12 <Speuler> you could consider literal numbers to be state-smart
19:56:02 <I440r> yea you should avoid literals!!!
19:56:04 <I440r> hehe
19:56:12 <Speuler> but   : foo   7 5 +  ;   versus   [7] [5] +   would look kind'f strange :)
19:57:11 <Speuler> I440r: no, avoid creation of literal numbers :)
19:57:39 <I440r> define
19:57:39 <I440r> 1 constant 1
19:57:47 <Speuler> you used a literal :)
19:57:53 <Speuler> what about
19:58:00 <I440r> 2 constant 2
19:58:00 <I440r> ....
19:58:00 <I440r> 999999 constant 999999
19:58:05 <Speuler> here dup <> 1+ literal 1
19:58:12 <Speuler> here dup <> 1+ constant 1
19:58:22 <I440r> :)
19:59:30 <Speuler> but,
20:00:53 <Speuler> <bs><bs><bs><bs>
20:01:20 <I440r> :)
20:01:29 <galexand> i440r, but that depends on compile-time literals...
20:01:35 <galexand> er run-time rather *shrug*
20:01:48 <galexand> what is <> ?
20:01:59 <Speuler> : <>  = 0= ;
20:02:14 <galexand> oh != :)
20:03:38 <I440r> <> goes way back to the beginnings of basic. maybe even further :)
20:04:51 <Speuler> "less or more"
20:05:01 <Speuler> "but not equal"
20:05:39 <galexand> yes i used <> in basic.  then i turned 12.  !=  :)
20:06:58 <galexand> why do most animals have a head and a tail?
20:07:09 <galexand> i think even worms are directional
20:07:22 <Speuler> see urchins ?
20:07:31 <Speuler> sea ...
20:07:45 <galexand> hmmm top and bottom
20:08:00 <galexand> :)
20:08:19 <galexand> i don't actually know what a sea urchin is i'm thinking of like a star fish
20:08:33 <Speuler> galexand: a sphere with spikes
20:08:40 <Speuler> dark to black
20:08:54 <Speuler> hole on downside
20:09:00 <galexand> hole = head :
20:09:02 <galexand> )
20:10:16 <galexand> i want my kids to have eyes on their feet
20:10:27 <galexand> and i'm not gonna have kids until that's possible
20:12:15 --- quit: skylan (Read error: 104 (Connection reset by peer))
20:13:19 <Speuler> i've been eating raw sea urchins
20:13:24 <galexand> really?
20:13:37 <Speuler> not kidding you
20:13:42 <galexand> tasty?
20:13:57 <Speuler> first was strange. the next one was ok
20:14:04 <galexand> awesome.  where do you get them?
20:14:09 <Speuler> from the sea
20:14:13 <galexand> awesome
20:14:25 <galexand> hmmm where can i get sushi at 11pm in a crappy city in the south
20:14:29 <Speuler> raw = alive ...
20:14:51 <Speuler> at least, reflexes still present
20:14:56 <galexand> but but
20:15:13 <galexand> can you eat the spikes?
20:15:26 <Speuler> if there was a tail, i'd noticed it
20:15:45 <galexand> i wonder how hungry i'd have to be to just pick up a sea urchin and eat it
20:15:53 <galexand> like an apple!
20:15:53 <Speuler> nope. you crack them open, add some drops of lemon, and suck them empty
20:16:14 <galexand> why?
20:16:25 <Speuler> t'is nourishing
20:16:28 <galexand> cool
20:16:32 <Speuler> why does man eat ?
20:16:36 <galexand> is this as an alternative to going to the grocery store or what?
20:16:46 <galexand> "i don't feel like paying for lunch let's go to the beach"
20:16:51 <Speuler> right
20:17:02 <Speuler> sicily
20:17:23 <galexand> neat
20:17:29 <Speuler> many people live from the sea there
20:18:31 <Speuler> can add them to stew/soup etc too
20:18:40 <Speuler> kills the reflexes
20:18:49 <galexand> neat
20:18:54 <Speuler> (with hot soup, that is)
20:19:57 <galexand> yeah
20:22:20 --- join: skylan (sjh@Rockcliffe25.tbaytel.net) joined #forth
20:23:07 <Speuler> i guess you can charge a fortune for sea urchin soup in a 5 star restaurant
20:23:42 <galexand> yeah probably
20:24:03 <Speuler> look what you pay for snails in a french restaurant
20:24:59 <Speuler> or frog legs
20:25:36 <Speuler> or fly wings
20:25:40 <Speuler> ahem
20:28:52 <galexand> fly wings?
20:29:21 <Speuler> offer it as speciality, and people will buy it
20:32:00 <galexand> it makes as much sense as fish eggs i guess
20:35:11 <Speuler> less, i guess. fish eggs are nourishing. i doubt fly wings are
20:35:28 <Speuler> they may stick between your teeth though
20:41:37 <Speuler> g'd night
20:43:41 <I440r> nite nite
20:43:46 <I440r> :P
20:43:46 --- quit: I440r ("BitchX-1.0c19 -- just do it.")
21:22:59 --- quit: fridge ("Confucius say: Man who meows ate pussy!")
21:48:31 --- join: fridge (meldrum@zipperii.zip.com.au) joined #forth
21:55:08 <joonas> bye
21:55:10 --- quit: joonas ("ircII2.8.2-EPIC3.004+Kasi --- Bloatware at its finest.")
22:00:41 --- quit: galexand ("Leaving")
22:21:30 --- quit: fridge (Success)
23:27:24 --- quit: TreyB (Read error: 110 (Connection timed out))
23:50:04 --- join: delta__ (~lazy@h-64-105-46-245.MCLNVA23.covad.net) joined #forth
23:59:59 --- log: ended forth/02.11.29