00:00:00 --- log: started forth/04.02.19
00:23:03 --- join: imaginator (~George@166.70.196.201) joined #forth
00:23:45 --- quit: Serg ()
01:08:28 --- join: Serg (~z@212.34.52.140) joined #forth
02:39:51 --- quit: imaginator (Read error: 60 (Operation timed out))
02:39:51 --- quit: Serg (Read error: 104 (Connection reset by peer))
02:40:47 --- join: Nutssh (~Foo@gh-1081.gh.rice.edu) joined #forth
02:56:57 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth
03:30:47 --- quit: MarkT ("I'm outta here")
05:24:23 --- join: Serg (~z@212.34.52.140) joined #forth
06:31:31 --- quit: Nutssh ("Client exiting")
06:45:53 --- join: arke (~Chris@wbar8.lax1-4-11-099-104.dsl-verizon.net) joined #forth
06:57:51 --- quit: Serg ()
07:03:04 --- join: hovil (~hovil@CommSecureAustPtyLtd.sb1.optus.net.au) joined #forth
07:18:08 <arke> later all
07:18:23 <arke> gonna be here for a few minutes at around 1pm Pacific
07:37:52 --- join: Nutssh (~Foo@gh-1081.gh.rice.edu) joined #forth
07:45:50 <hovil> whooooooo
08:21:48 --- join: downix (~downix@adsl-214-34-245.mia.bellsouth.net) joined #forth
08:24:44 --- join: proteusguy (proteusguy@51.sub-166-153-180.myvzw.com) joined #forth
09:14:14 --- quit: Nutssh ("Client exiting")
09:37:18 --- join: _proteus (proteusguy@209.sub-166-153-179.myvzw.com) joined #forth
09:41:28 --- quit: proteusguy (Operation timed out)
09:53:39 --- join: Nutssh (~Foo@gh-1204.gh.rice.edu) joined #forth
10:20:19 --- quit: _proteus ("Leaving")
10:59:19 --- quit: Nutssh ("Client exiting")
11:20:56 --- join: I440r (~mark4@165.139.76.25) joined #forth
11:37:23 <I440r> think i fixed the isforth.clss.net font problem
11:40:29 --- quit: slava (Read error: 104 (Connection reset by peer))
11:51:20 <hovil> my head
11:51:24 <hovil> it hurts
11:51:57 <I440r> stop banging on it then
12:11:10 --- join: Nutssh (~Foo@dunwlessnat.rice.edu) joined #forth
12:18:32 --- join: wossname (wossname@HSE-QuebecCity-ppp80676.qc.sympatico.ca) joined #forth
12:26:59 --- quit: Nutssh ("Client exiting")
12:44:17 --- join: proteusguy (proteusguy@176.sub-166-153-181.myvzw.com) joined #forth
13:05:44 --- quit: hovil ("Leaving")
13:11:25 --- quit: I440r ("Leaving")
13:13:14 --- quit: arke (Read error: 104 (Connection reset by peer))
13:13:25 --- join: arke (~Chris@wbar8.lax1-4-11-099-104.dsl-verizon.net) joined #forth
13:13:35 <arke> hi :)
14:12:11 --- join: Nutssh (~Foo@dunwlessnat.rice.edu) joined #forth
14:36:17 --- join: slava (~slava@CPE0080ad77a020-CM.cpe.net.cable.rogers.com) joined #forth
14:36:18 <slava> hi
14:38:03 --- quit: qFox ("if at first you dont succeed, quit again")
14:45:21 <ianp> hey
15:22:40 --- quit: Nutssh ("Client exiting")
15:24:17 <arke> hi
15:48:55 <slava> hey arke
15:49:03 <arke> whats up slava?
15:49:11 <slava> working on the game ;)
15:49:24 <slava> artist guy sent some new additions, like a hovercraft and 747 aeroplane
15:49:56 <arke> cool\
15:49:58 <downix> what kind of game are you making?
15:50:52 <slava> downix, action game
15:51:01 <slava> & rpg
15:51:24 --- join: ASau (~asau@158.250.48.196) joined #forth
15:51:39 <ASau> Dobryjj vecher!
15:53:25 <arke> terve ASau
15:55:42 --- quit: proteusguy ("Leaving")
15:56:30 <downix> ok cool
16:00:46 <ASau> Reserved value effect:
16:01:24 <ASau> in order to 0 LIST you have to 1 40 INDEX 0 LIST
16:03:52 <ASau> 1 40 INDEX is here to use all buffers  and eliminate free ones.
16:13:29 --- quit: wossname ("^_~")
16:54:52 <madgarden> Is this be considered bad form for multi-value constants?
16:54:52 <madgarden> : box1 10 10 32 20 ;
16:54:52 <madgarden> box1 draw_box
16:57:56 --- join: Nutssh (~Foo@gh-1204.gh.rice.edu) joined #forth
16:58:47 --- quit: kc5tja ("leaving")
16:58:58 --- join: kc5tja (~kc5tja@66-91-231-74.san.rr.com) joined #forth
16:59:05 --- mode: ChanServ set +o kc5tja
16:59:50 <arke> hi kc5tja
16:59:59 <kc5tja> re
17:00:40 <arke> how are you my firneknd?
17:00:41 <arke> ack
17:00:41 <arke> fuck
17:00:45 <arke> Friend*
17:02:09 <downix> hey sam
17:06:32 * kc5tja ought to put that in a frame and hang it above my desk.
17:06:48 <arke> <topher> ssh
17:06:49 <arke> <topher> quiet
17:06:49 <arke> <topher> say fun stuff
17:06:49 <arke> <arke> ssh? into your box? sure.
17:06:59 <arke> :)
17:13:14 <madgarden> <madgarden> Is this be considered bad form for multi-value constants?  
17:13:14 <madgarden> <madgarden> : box1 10 10 32 20 ; 
17:13:14 <madgarden> <madgarden> box1 draw_box 
17:13:22 <madgarden> Anyone? Bueller?
17:14:58 <kc5tja> madgarden: That's perfectly fine.
17:15:40 <madgarden> Alrighty, thanks. Seems like it would be more work any other way.
17:17:55 <kc5tja> I see that as a proper use of factoring, which is the intention of colon definitions.
17:18:04 <kc5tja> If you use those coordinates a lot, it makes sense to factor them out.
17:20:45 <madgarden> Yep, that's exactly what I was thinking. ANd if I need to tweak them, they only change in one place.
17:21:13 <madgarden> So this means I will expose more Forth words to my animation configuration scripts.
17:21:55 <kc5tja> That's one of the greatest things about Forth too -- the invisibility of data type.
17:22:04 <kc5tja> In the above case, box1 is a set of four constants.
17:22:14 <kc5tja> But later on, you might want to *compute* those values from some other input.
17:22:22 <kc5tja> Constants?  Computations?  Forth doesn't care.
17:23:11 <madgarden> Zactly. And it makes configuration files really flexible.
17:26:33 <arke> <Th3No0b> Im going to be the next hitler
17:26:33 <arke> <Th3No0b> Im going to kill all the jews and 1 clown
17:26:33 <arke> <RageAgainsttheAmish> why the clown
17:26:33 <arke> <Th3No0b> See? no one cares about the jews
17:26:33 <arke> <RageAgainsttheAmish> lmao
17:26:44 --- join: proteusguy (~proteusgu@216.27.161.121) joined #forth
17:26:51 <arke> *** Quits: TITANIC (Excess Flood)
17:27:46 <kc5tja> HAHA -- I like the Hitler bit.  That is quite amusing.
17:27:51 <kc5tja> As is the Titanic line.
17:28:02 <ianp> good old bash.org
17:40:24 <arke> yep
17:40:29 <arke> such simple entertainment...
17:40:36 <arke> they redid their engine, its alot better now
17:40:41 <arke> especially the adding thing :)
18:05:58 --- join: hefner (~hefner@pool-141-157-60-195.balt.east.verizon.net) joined #forth
18:06:52 <hefner> hola forthers
18:07:44 <hefner> what are the smallest (in bytes..) forths people are aware of?
18:08:06 <chandler> greetings hefner 
18:08:53 <hefner> hey there
18:09:01 <kc5tja> hefner: Bare bones PygmyForth (16-bit cmForth clone for PC under DOS), without block editor or assembler, comes in at 6K to 8K in size.
18:09:31 <hefner> hmm, I thought the core of colorforth was smaller than that
18:09:56 <kc5tja> It probably is.
18:10:16 <kc5tja> I think Chuck said it weighed in somewhere around 2.5K to 4K or so.
18:10:17 <Herkamire> hefner: the core of my ppc colorforth-ish thing is under 1K PPC binary
18:10:31 <Herkamire> but it's useless without the dictionary and source
18:10:36 <kc5tja> But that also includes font data too.
18:10:47 <kc5tja> And the integrated editor -- the editor is a requirement to use ColorForth.
18:13:08 <Herkamire> with simpler/smaller hardware, a forth could easily be smaller
18:13:38 <kc5tja> Yes.
18:14:21 <kc5tja> On Forth-specific hardware, the Forth run-time environment can be a few hundred bytes in size.  But usually, it's the interactive text interpreter and editor that bulk up the system usually.
18:17:50 <Herkamire> my forth is 148K
18:17:55 <Herkamire> but a lot of that is padding
18:18:07 <Herkamire> gzip takes it down to 20K
18:18:15 <Herkamire> bzip2 takes it down to 18K
18:19:27 <Herkamire> that's the ELF file containing the meta-compiled kernel, dictionary, and sources for the assembler, compiler and editor
18:22:01 <Herkamire> oh, that contains the documentation in my shadow blocks too
18:22:11 --- quit: Nutssh (Read error: 60 (Operation timed out))
18:23:26 <Herkamire> hmmmm... quite a lot of padding it looks like. I replaced all strings of null bytes in the elf with one byte,  and the file size went down to 31K
18:23:56 <kc5tja> :)
18:38:19 <ASau> hefner, postForth, Min(i)Forth
18:40:02 --- join: ree (~jwm@ns.fasthost.net) joined #forth
18:40:14 <ASau> Dobre jitro!
18:42:40 <ASau> Herkamire, My Forth gzipped is 4392 bytes. It's padded to 16K
18:45:35 <ASau> Herkamire, 5690 really.
18:46:15 --- quit: fridge ("Client exiting")
18:54:25 --- join: fridge (~fridge@dsl-203-113-229-145.NSW.netspace.net.au) joined #forth
19:46:21 --- quit: hefner ("I'm the operator with my pocket calculator")
19:56:12 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth
20:10:19 <ASau> Alive?
20:10:47 <qFox> dead in the water
20:10:55 <qFox> sir!
20:11:19 <ASau> I'm experimenting with "ramdisk".
20:11:24 <ASau> Forth blocks in RAM.
20:12:07 <qFox> i cannot help thee with that, sorry :\
20:12:08 <ASau> It might be interesting for those, who want standalone Forth or OS.
20:12:18 <qFox> someone else might tho, just not me
20:20:13 <Herkamire> asau pretty small. what does that include?
20:20:46 <Herkamire> ASau: aren't blocks always used in RAM?
20:24:35 <Herkamire> I'm going to make a solitare numbers game in my forth
20:26:25 <ASau> Core, compiler (IF, THEN, WHILE etc), VLIST, LOAD, INDEX, LIST.
20:46:58 --- quit: downix ("Leaving")
20:51:59 --- join: LOOPHOG (jdamisch@207.191.240.70) joined #forth
20:53:00 --- quit: LOOPHOG (Client Quit)
21:11:18 --- join: imaginator (~George@166.70.196.201) joined #forth
21:14:21 <arke> hi allll
21:15:40 <Herkamire> hi arke 
21:15:54 <arke> :)
21:42:08 <ASau> Dobre jitro, arke!
21:42:36 <arke> Dobryjj vecher!
22:06:20 <ASau> Crazy idea.
22:06:58 <ASau> To make my 5690-bytes Forth 12-bit.
22:20:10 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK")
23:39:48 --- quit: ASau ()
23:41:02 <OrngeTide> haah.. this is neat. direct threaded forth in C using gcc tail recursion optimization.
23:41:35 <OrngeTide> if you enable debugging it won't work though. cuz that will disable the optimzation and you'll end up consuming gobs and gobs of stack
23:46:28 <imaginator> OrngeTide: how is direct threading down in C in that case?
23:46:58 <imaginator> down/done
23:47:14 <OrngeTide> imaginator, well you let the C compiler write the NEXT portion for you. my NEXT looks like:     return curr[1](curr+1);
23:47:37 <OrngeTide> with tail recusion optimization it actually does a jump instead of a call. 
23:48:46 <OrngeTide> of course you eventually have to return somewhere. I return in my routine for bye.
23:49:21 <imaginator> Aren't function calls costly?
23:49:51 <OrngeTide> imaginator, it's not a function call. it's a jump
23:49:58 <imaginator> oh
23:50:04 <OrngeTide> imaginator, like i said. you have to have tail recursion optimization enabled.
23:50:17 <imaginator> I see.  So it's just the initial call that starts it and then the jmps.
23:50:37 <OrngeTide> otherwise 1. you'll end up consuming progressively more stack as your program runs, 2. you would have the same problems that a subroutine threaded app would have
23:50:42 <OrngeTide> imaginator, yes:)
23:51:12 <OrngeTide> i've been looking at the asm dumps and if i keep the number of args I pass down to only a few I can have them in registers. 
23:51:36 <OrngeTide> of course it's pretty evil to write forth in C. but hey. i was bored.
23:51:59 <OrngeTide> i actually found this trick in this article: http://www.complang.tuwien.ac.at/forth/threaded-code.html .. (it's under the section of Call Threading)
23:52:22 <imaginator> In my Forth before I began working on the machine code VM, I had an array of instructions and I'd use goto *jump_table[offset].  Where offset is the code for the word.
23:52:29 <OrngeTide> wait. i mean it's under the header Continuation-passing style
23:53:47 <imaginator> cool
23:54:20 <OrngeTide> imaginator, oh. I see. i've seen something like that but instead you write a macro NEXT that is goto *(ip++);
23:54:29 <imaginator> yep
23:54:35 --- quit: madgarden (Read error: 104 (Connection reset by peer))
23:54:37 <imaginator> And I store the operands in the array too.
23:54:48 <OrngeTide> yea. that's a pretty fast way to go.
23:55:04 <imaginator> It's pretty fast.
23:55:15 <OrngeTide> how do you handle the allocation of the dictionary?
23:55:18 <imaginator> But now I'm trying to get rid of next in between.
23:55:51 <imaginator> Well I make core words have a numeric value via: enum { ADD = 1, SUB };   
23:56:26 <imaginator> Then in my run() I use the codes in a jump table.   gcc has a special && operator that you can use to get the address of a local label.
23:56:43 <OrngeTide> yea. i wrote one using && before
23:57:17 <imaginator> My current idea is to copy machine code at runtime for the execution.
23:57:18 <OrngeTide> i'm thinking of using mmap to allocate my dictionary.
23:57:45 <imaginator> My dictionary was a hash table, but I implemented a 4-bit association engine that seems to be faster, but it uses a little more memory.
23:59:17 <imaginator> http://www.xmission.com/~georgeps/var/trie_playground.c
23:59:39 <imaginator> I call it a 4-bit association engine, even though someone told me it's also known as a ternary hash trie IIRC.
23:59:59 --- log: ended forth/04.02.19