00:00:00 --- log: started forth/02.11.12 02:35:58 --- quit: pyromania (Read error: 110 (Connection timed out)) 02:49:03 --- nick: proteus_ -> proteusguy 03:07:17 --- join: proteus_ (~username@65.191.88.177) joined #forth 03:07:19 --- quit: proteusguy (Read error: 54 (Connection reset by peer)) 03:09:34 --- quit: proteus_ (Client Quit) 03:28:39 --- quit: onetom (Read error: 60 (Operation timed out)) 04:02:56 --- join: pyromania (~pyromania@dialup-243.159.220.203.acc01-high-pen.comindico.com.au) joined #forth 04:41:05 --- join: onetom (~tom@novtan.bio.u-szeged.hu) joined #forth 06:11:46 --- join: Herkamire (~jason@wsip68-15-54-54.ri.ri.cox.net) joined #forth 06:12:18 --- join: tathi (~josh@wsip68-15-54-54.ri.ri.cox.net) joined #forth 06:36:58 Hi all 06:38:00 mornin :) 07:21:04 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 07:29:44 morning 07:37:34 --- join: I440r (~mark4@sdn-ap-005tnnashP0179.dialsprint.net) joined #forth 07:42:30 What is c, again? 07:42:36 create & , 07:42:36 > 07:42:36 ? 07:43:19 ok 07:43:25 you know what , (comma) does ? 07:43:39 its stack comment is ( n1 --- ) 07:43:52 Um , vaguely :) 07:43:58 it allots 4 bytes of memory and writes N1 into the allocated memory 07:44:01 puts n1 at the current place in memory 07:44:06 right, ok 07:44:18 the comma words are teh words taht actually COMPILE the data into the dictionary 07:44:25 if you do 07:44:27 hm 07:44:31 here 12 , 07:44:33 sounds key 07:44:47 you will compile the 12 into memory 07:44:54 that allocates and writes 12 into the memory.. 07:45:05 you will also leave the address it was compile at on the stack (here) 07:45:12 cool 07:45:17 makes sense yeah 07:45:22 here isnt there anymore 07:45:22 as you compile data forth maintains a pointer to where the NEXT item will be compiled to 07:45:30 you can get the address by using the word here 07:45:38 yeah, i got 'here' down' 07:45:39 down 07:45:54 What's c, then? and what's immediate? 07:45:55 so i could comment the above as follows 07:46:05 here \ get the address where the next itme is compiled to 07:46:22 c, is like , except it only writes ONE character. one byte 07:46:27 w, writes 2 07:46:30 Ah, so it's just optimized 07:46:32 character comma 07:46:36 word comma 07:46:37 and comma 07:46:47 no. its used totally different 07:46:56 you COMMA in 32 bit data 07:47:04 you W-COMMA in 16 bit data 07:47:11 and you C-COMMA in 8 bit data 07:47:44 here 5 c, 'H' c, 'E' c, 'L' c, 'L' c, 'O' c, 07:47:48 count type 07:48:20 erg 07:48:31 'H' aint 32 bits? 07:51:31 It occupies 32 bits on the stack. 07:51:44 But, "c," only copies the low 8 bits to the heap. 07:51:56 what he said :) 07:52:30 Hmm... 07:53:04 * Robert thinks about what to do next. 07:53:40 crazy 07:53:54 and all i need in normal ascii is 8 bit? 07:53:58 Who? Me? I AM NOT CRAZY! 07:54:07 Eh, yes. 07:54:18 heh yeah, awesome 07:54:25 32 bit seems wasteful now that i think of it 07:54:36 ianni yes all ascii characters are 8 bits 07:54:55 if ascii was 32 bits we whould have a HUGE character set :) 07:55:00 is that 65535^2 or 65535? 07:55:01 Ehehe 07:55:09 oh, 2^8 07:55:41 $FFFFFFFF is 32 bits $FFFF is 16 bits 07:55:46 figure it out :) 07:55:50 :) 07:55:56 in isforth say $ffffffff u. 07:55:58 2^32 different characters.. MMMM 07:56:33 erm, ASCII's technically just 7-bit, isn't it? 07:56:33 :) 07:56:44 tathi - yeah, taht sounds familiar 07:56:55 Well... 07:57:09 If you count the old standardized ASCII, sure. 07:57:24 But people never follow standards, even if they're there. 07:57:52 So now we got a bunch of more or less sane 8-bit character sets. 07:58:16 Which cause more problems than you can imagine if your last name starts with an Ö. 07:58:25 heh 08:02:34 : hello-world ." Hello World" cr bye ; 08:02:36 or 08:02:40 int main (void) 08:02:42 { 08:02:52 fprintf(stdout,"Hello World\n"); 08:02:55 return 0; 08:02:57 } 08:03:02 hmm c sux 08:03:26 int main(void){puts("Hello, world!\n");return 0} 08:03:26 :P 08:03:50 yes. and the C version is STILL more visuallyu cluttered 08:03:58 all you did was CRAM it all into one line 08:04:06 the forth version isnt CRAMMED 08:04:12 : hello-world 08:04:18 ." Hello World" 08:04:23 cr bye ; 08:04:33 WAY less verbose 08:04:39 c sux 08:06:25 :) 08:13:20 brb, my siberian attack POODLE is nagging me :P 08:13:26 walkies --> 08:19:50 wonder hahah 08:35:05 ack :) 08:43:36 --- quit: pyromania (Read error: 60 (Operation timed out)) 09:01:36 .( Hello, World!) cr bye 09:03:10 cant turnkey that :) 09:03:27 but yes, it shows how forth can be FAR less verbose :) 09:08:42 I440r: so how does your turnkey thing work? it creates an elf whith the VM and the heap, and it knows what word to call to get the whole thing running? 09:08:55 I440r: do you save the stacks too? 09:10:27 turnkey saves out the list space (where all the code and colon goes) 09:10:36 it does NOT save any of the headers 09:11:29 to do a turnkey you need to set the initialization vector (can be done by patching into default) 09:11:41 : my-main defers default run-program bye ; 09:11:50 or by doing 09:11:59 : my-main run-program bye ; 09:12:13 ' my-main is default \ completely bypass the usual forht init 09:12:26 the prefered way in isforth is 09:12:30 ' my-main is quit 09:13:51 quit is a defered word thats normally defered to (quit) 09:14:11 default (the defered initialization chain) calls quit after all has been initialized 09:14:23 actually thats a lie heh 09:14:53 if you look in isforth.asm you see it init memory, init forts ip to... 09:14:56 dd init 09:14:58 dd quit 09:15:05 init runs default 09:15:08 or something 09:16:34 i got that backwards lol 09:16:37 dd quit 09:16:39 grr 09:16:42 dd default 09:16:44 dd quit 09:17:14 and default calls init - i hafta change this slightly, "init" doesnt init anything except the terminal so its badly named 09:18:51 ok cool 09:19:00 Oh the dogs they had a meeting, 09:19:04 they came from near and far. 09:19:12 and some dogs came by motor bus, 09:19:15 and some by motor car. 09:19:22 on entering the meeting hall, 09:19:24 ' my-main is main 09:19:26 would be nice 09:19:27 each dog could take a look, 09:19:33 where he had to hang is asshole, 09:19:40 up high upon a hook. 09:19:46 and when they were all seated, 09:19:54 each canine son and sire, 09:20:02 some dirty bulldog sone of a bitch 09:20:14 jumped up and hollared FIRE!!!!!!!!!!! 09:20:27 all was in a panic, twas hell upon to look 09:20:33 each doggie grabed at random 09:20:38 an asshole from a hook 09:20:44 now the assholes were all mixed up 09:20:51 which made each doggie sore, 09:20:56 to have to wear another dogs ass, 09:21:01 he'd never worn before 09:21:08 and that is why, until this day, 09:21:11 a dog will drop a bone. 09:21:16 to go and smell another dogs ass 09:21:20 to see if its his own! 09:22:54 HAha. 09:31:11 --- quit: XeF4 ("pois") 10:13:05 --- join: neobrat (~neobrat@h-64-105-21-62.DNVTCO56.covad.net) joined #forth 10:13:41 howdy all 10:14:45 Hi :) 10:14:59 a little quiet at the moment eh? 10:19:04 a stunned silence from I440r's poetry 10:19:08 ;) 10:19:31 * neobrat is sad that he missed poetry hour :) 10:19:45 Heh. 10:19:51 thank god it didn't go on for an hour ;) 10:20:02 hehe 10:20:08 that good huh? 10:20:25 humph 10:20:29 Heh. 10:20:59 Sometimes I wonder if certain persons are 40 or 4 years old :) 10:23:02 lol 10:23:14 the cow is of the bovine ilk 10:23:18 one end is MOO the other milk 10:23:20 :P 10:26:00 oh my 10:26:08 maybe im not that sad :) 10:26:17 :) 10:26:53 i would do the ryme of the ancient mariner but i dont have all of that memorized yet heh 10:27:02 and it might scare some of you kiddiez away :) 10:27:27 i only know "cultured" poetry.... 10:27:39 ...There once was a woman from venus.... :) 10:29:16 lol 10:30:23 anyone have an EASY (don't care about speed) algorithm to check if a string is an integer OR a float? 10:30:25 neobrat does your irc client support privmsg ? 10:30:39 ya, just don't ever notice it 10:30:43 neo a FORTH string ? 10:30:53 just scan it for a . :) 10:31:03 hmm no. . is used to signify a double isnt it 10:31:05 hrm 10:31:22 ive no idea. i dont think theres ANY real application that REQUIRES floating point 10:31:33 Not in FORTH 10:31:41 -- sorry should have specified 10:31:44 aha 10:31:49 then search for a . :) 10:31:52 or something heh 10:31:53 the algorithm could be, but the concept i need for anything 10:31:54 --- join: pyromania (~pyromania@dialup-160.159.220.203.acc01-high-pen.comindico.com.au) joined #forth 10:34:08 neobrat: is it difinately an int or a float? or could it be any text? 10:35:09 any text is possible 10:36:14 and should be able to handle multiple possibilities: (ie -1.3e56) 10:36:27 or -3.4+6i 10:37:30 you want immaginary numbers too? 10:37:41 yep 10:37:54 but that can be an after thought if i get the rest 10:40:21 in C strtod will do everything except the immaginary part I think 10:51:27 thanks, ill have to look at that 11:12:26 --- join: XeF4 (xef4@lowfidelity.org) joined #forth 11:38:08 --- quit: neobrat () 12:17:30 bigforth is cool 12:20:07 --- join: neobrat (~neobrat@h-64-105-21-62.DNVTCO56.covad.net) joined #forth 12:25:07 oooo, has midi 12:30:17 ugh 12:32:39 the code is so scary 12:34:11 create masks $55555555 , $aaaaaaaa , $33333333 , $cccccccc , 12:34:27 $0f0f0f0f , $f0f0f0f0 , $00ff00ff , $ff00ff00 , 12:34:32 0 var shift 12:34:38 : revbits 12:34:43 1 !> shift 12:34:46 5 12:34:46 for 12:34:55 dup shift u<< swap shift u>> 12:35:06 5 r@ - 3 u<< masks + 12:35:16 dup 4+ @ swap @ 12:35:21 rot and -rot and or 12:35:29 shift 2* 1> shift 12:35:32 nxt ; 12:35:40 erm that nxt should be undented 2 spaces 12:35:47 can anyone see of a way to make that smaller >? 12:37:49 erh thats shift 2* !> shift 12:38:00 grr i REALY need to learn to type 12:38:23 how many bits are there total? 32? 12:38:46 and do you want to code smaller or faster? 12:38:54 yes 12:39:00 ill show you the original code 12:39:04 : revbits 12:39:20 dup 1 u>> $55555555 and swap 1 u<< $aaaaaaaa and or 12:39:41 dup 2 u>> $33333333 and swap 2 u<< $cccccccc and or 12:40:04 dup 4 u>> $0f0f0f0f and swap 4 u<< $f0f0f0f0 and or 12:40:28 dup 8 u>> $00ff00ff and swap 8 u<< $ff00ff00 and or 12:41:01 dup 16 u>> $0000ffff and swap 16 u<< $ffff0000 and or ; 12:41:15 which is about 80 bytes bigger :) 12:42:09 dup shift tuck u>> swap u<< ? 12:43:29 nah thats no smaller 12:45:19 and what exactly is this code trying to do...? 12:45:42 reverse all the bits 12:45:56 %10110101 would be 10101101 12:46:01 etc 12:46:23 its based on some "obfuscated" c code 12:46:41 why anyone ever writes obfuscated c code is beyond me, its ALL obfuscated :) 12:47:14 is u<< a shift or a rotate? do you have a rotate? 12:47:21 a shift 12:47:28 theres u<< and << 12:47:29 rotate would make this much simpler 12:47:32 and u>> and >> 12:47:36 unsigned and signed 12:47:55 when shifting there should be no difference between signed/unsigned 12:48:07 when shifting _left_ 12:48:13 when shiftuing LEFT you mean 12:48:19 there isnt realy 12:48:24 but i have seperate words heh 12:48:28 true, i stand corrected ;) 12:48:34 are they aliases? :) 12:48:43 no 12:48:54 << and u<< are not identical 12:49:02 why not? 12:49:10 but they can be used in place of each other 12:49:21 look at the INTEL instruction set :P 12:49:43 i rarely use anything except u>> and u<< 12:49:50 ive never seen a need for >> yet 12:49:53 erm 12:49:55 sets the flags differently or something? 12:49:56 << i mean 12:50:04 nope 12:50:17 lsl and lsr as opposed to asl and asr 12:50:29 I don't give a crap about the instruction set. 12:50:34 asr and lsr are functionally identical 12:50:42 they do exactly the same thing. why have different definitions? just pick the best one 12:50:44 but not teh same opcode 12:50:44 oh right 12:50:57 silly silly x86 architecture... 12:51:03 neither is best heh 12:51:34 if get a better computer then ;) 12:51:49 lol 12:57:51 so do you have a reason to reverse all the bits in a word? 12:57:58 or are you just doing this for kicks? 12:58:18 ok i shaved 5 bytes off :) 12:59:06 rhm that should have been 9 bytes - im confused! 12:59:33 do you want speed or compactness? 12:59:57 compactness heh 13:00:07 i never optimize for seed when im playing :) 13:00:11 double check this: 13:00:24 but this was a speed improvement too 13:00:50 : revbits ( n -- n ) >r 32 0 for r@ 1 i lshift & i rshift step r> drop 32 1 for over i lshift | step ; 13:02:03 hang on 13:02:40 i may have missed something, but the idea is there 13:07:42 actuallym u miss the point heh 13:07:59 : revbits 0 32 for 2* over 1 and + swap 2/ swap nxt nip ; 13:08:04 thats alot smaller and alot faster 13:08:50 so why not do that? 13:09:07 im trying to make the code i originally typed here smaller :) 13:09:23 bah -- go with what works :) 13:09:31 they ALL work heh 13:13:37 you're trying to make it much better without rewriting it??? 13:18:20 lol yes and no :) 13:18:58 I think it's more fun with rotates 13:22:16 : >>> rotate-right ; 13:22:23 : revbits ( x -- ) 13:23:13 1 <<< 13:23:56 its teh obfuscated code im trying to obfuscate even MORE :) 13:24:03 while making it slaller 13:24:14 dup $55555555 and 2 <<< 13:24:14 if i can get this change to work ive just shaved 30 more bytes off 13:25:03 --- join: wossname (wossname@HSE-QuebecCity-ppp81919.qc.sympatico.ca) joined #forth 13:25:37 s/\$55555555/$aaaaaaaa/ 13:27:45 scratch that. 13:27:50 I can't program over irc 13:28:29 me either heh 13:30:04 looks like it's 68 bytes in PPC assembly language (using rotates) :) 13:32:51 here $0000ffff , $ffff0000 , $00ff00ff , $ff00ff00 , $33333333 , $cccccccc , $55555555 , $aaaaaaaa , ( reversed order of pairs) 13:32:55 : revbits 13:32:58 1 5 13:33:03 for 13:33:13 2dup u<< >r tuck u>> r> 13:33:29 r@ 1- 3 u<< masks + 13:33:38 dup @ swap 4+ @ 13:33:43 rot and -rot and or 13:33:45 swap 2* 13:33:50 nxt drop ; 13:34:20 does isforth have a rotate? 13:34:29 no 13:36:07 after compiling the above i have 504898 bytes of list space free. after compiling teh original code i have 504734 bytes free. that makes my code 356 bytes smaller ? 13:36:19 err no 13:36:27 i know why it came up that high heh 13:36:29 lol 13:36:46 well, here's a 16 bit version with rotates, it is just one more line for 32 bits but I don't feel like writing it out and figuring out where the masks go and which way to rotate it. 13:36:50 : 16revbits 13:36:50 thats 164 bytes smaller 13:36:52 2 <<< 13:36:55 dup $aaaaaaaa 2 <<< swap $55555555 and or 13:36:57 dup $cccccccc 4 >>> swap $33333333 and or 13:37:00 dup $c3c3c3c3 8 >>> swap $3c3c3c3c and or ; 13:37:41 have you tested that ? 13:37:55 only on paper :) 13:38:17 does gforth have a rotate? 13:38:29 dunno 13:38:56 1 2 3 rot .s <3> 2 3 1 ok 13:39:13 ianni: no :) I want to rotate bits 13:39:17 oh 13:39:18 no not rot 13:39:25 rotate bits not rotate items on the stack heh 13:40:11 if I had a forth with rotate I would test it and give you a 32 bit version 13:45:27 * neobrat is extremely relieved... 13:45:33 I thought I broke my compiler... 13:45:36 *phew* 13:45:48 Stupid pointer error when recoding my dictionary lookup 13:46:30 lol 13:48:12 i did that a couple of times last week 13:48:30 i did some bug fixes and had to spend a couple of days bug fixing my bug fixes heh 13:49:30 bugs suck 13:49:35 I've been getting my programms < 1K 13:49:42 so it never takes days to find a bug 13:49:52 it's great having the whole program on the screen at once 13:50:09 that would be nice 13:52:51 heh :) we're adding rotate to fpos 13:55:06 lol 13:55:21 lol 14:02:13 hmm...took me longer to find the documentation than to actually write it though... 14:15:21 rotate through the carry flag? or with it? 14:16:15 --- join: deego (~user@63.126.179.199) joined #forth 14:16:35 --- part: deego left #forth 14:20:44 --- quit: neobrat () 14:32:40 ok, this is tested and works: 14:32:41 : revbits 14:32:41 1 <<< 14:32:41 dup $aaaaaaaa and 2 <<< swap $55555555 and or 14:32:42 dup $cccccccc and 4 <<< swap $33333333 and or 14:32:44 dup $f0f0f0f0 and 8 <<< swap $f0f0f0f and or 14:32:46 dup $ff00ff00 and 16 <<< swap $ff00ff and or ; 14:33:05 : <<< rotate-left ; 14:38:03 which is basically the same as the original code which i optimized by 164 bytes! 14:38:17 in fact i think the above is larger :) 14:38:27 whats teh 1 <<< for at the start :) 14:41:16 your full of crap 14:42:04 ? 14:42:42 I have 5 rotates instead of 10 shifts 14:42:45 : revbits 14:42:45 dup 1 u>> $55555555 and swap 1 u<< $aaaaaaaa and or 14:42:45 dup 2 u>> $33333333 and swap 2 u<< $cccccccc and or 14:42:45 dup 4 u>> $0f0f0f0f and swap 4 u<< $f0f0f0f0 and or 14:42:45 dup 8 u>> $00ff00ff and swap 8 u<< $ff00ff00 and or 14:42:45 dup 16 u>> $0000ffff and swap 16 u<< $ffff0000 and or ; 14:43:03 I have 8 ands instead of your 10 14:43:09 right 14:43:14 i missed that :) 14:43:22 Herk: does ppc have hardware rol 14:43:23 now i understand the 1 <<< :) 14:43:24 wadda hell rya playin w? 14:43:25 ? 14:43:31 XeF4: yes 14:44:05 onetom: writing a short forth program to reverse the bits in a 32 bit cell 14:45:03 Herkamire: umean: ...76543210 -> 01234567... ? 14:45:08 how many bytes does the above compile to in your forth ? 14:45:19 im assuing 32 bits per XT 14:45:44 if those are bit numbers yes :) 14:46:06 sure they r 14:46:57 how many bytes does the above compile to in your forth ? 14:47:33 over 300 for sure 14:47:38 mine compiles to 189 bytes 14:47:56