00:00:00 --- log: started forth/04.06.07 00:29:03 yeah 00:29:04 in halo too 00:37:13 --- quit: arke (Read error: 104 (Connection reset by peer)) 00:37:31 --- join: arke (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 00:38:07 --- part: arke left #forth 00:38:31 --- join: arke (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 00:38:56 --- topic: set to ':NOTICE: Kestrel project update! See http://www.falvotech.com/weblog for more details.' by arke 01:29:50 --- join: cedricshock (~cedricsho@209-181-58-5.eugn.qwest.net) joined #forth 01:36:50 hrm. 01:36:57 does gforth provide any curses stuffs? 01:37:23 Hm, rtfm. :) I have no idea 01:38:08 ^_^ 01:40:08 guess i could use isforth 01:40:20 but that would mean firing up vmware again 01:40:22 and i don't wanna 01:41:24 Or, worse, install a real system! 01:41:42 i have a real system 01:41:50 networking doesn't work right now on it though 01:42:03 Oh, why? 01:42:04 (VMWare is actually booting my SuSE partition :)) 01:48:13 Just a question. 01:48:30 arke. Why not hack curses into gforth yourself? 01:49:20 Robert: my current means of accessing the net doens't work on linux (WUSB11) 01:49:26 ASau: because I don't want to :) 01:49:36 ASau: and also, I need TCP/IP 01:49:46 AFAIK, Gforth provides libc/solibc API. 01:50:05 arke, take a look at http server written in Gforth. 01:50:15 There exists one. 01:50:32 Maybe even not single one. 01:51:07 ok, cool 01:51:11 any idea what its called? 01:52:40 cool, there was one written in bigFORTH as well 02:01:41 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 02:02:35 Hi fox 02:04:01 arke. Look at Paysan's pages. 02:04:07 IIRC, he wrote it. 02:04:21 yeah, I am. 02:04:23 In german too :) 02:05:50 mornin 02:08:20 --- join: crc (crc@0-1pool176-8.nas6.philadelphia1.pa.us.da.qwest.net) joined #forth 02:09:50 --- join: doublec (~doublec@coretech.co.nz) joined #forth 02:11:03 * arke is away: sleep --- night! 02:27:48 --- quit: Herkamire ("goodnight") 02:40:56 is it possible to rotate a byte? 02:41:31 i mean, shift it so that the bits shifted off on the left, shift back in on the right (or the other way around) 02:41:44 but then not for a 32bit number, but for 8bit 02:49:25 : ROL DUP 2* SWAP [ HEX ] 80 AND 0<> OR ; 02:49:39 : ROL DUP 2* SWAP [ HEX ] 80 AND 0<> OR 0FF AND ; 02:49:54 --- quit: ASau () 02:55:18 --- quit: crc (Nick collision from services.) 02:55:39 --- join: crc (crc@0-1pool176-8.nas6.philadelphia1.pa.us.da.qwest.net) joined #forth 03:41:18 --- quit: doublec ("Leaving") 03:42:11 arke: You could almost certainly call curses from gforth using fflib.fs. See http://www.shockfamily.net/cedric/fflib.html 04:22:03 --- quit: cedricshock ("Leaving") 04:26:12 --- quit: crc ("Hacking...") 06:06:13 --- join: imaginator (~George@georgeps.dsl.xmission.com) joined #forth 07:12:11 * arke is away: schoool!!! aaaarghrgurgkubjkjbtklk43126@#G>:%@L:K$LKrelkngK3GG:%KI@:%VLREN 07:16:16 man 07:16:20 i feel arke there 07:18:31 --- quit: I440r_ (Read error: 60 (Operation timed out)) 08:03:23 --- quit: Serg () 08:08:34 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 08:59:56 --- quit: tathi ("leaving") 09:08:53 --- quit: imaginator ("rest") 09:15:36 --- join: Topaz (~top@exten-halls-131.soton.ac.uk) joined #forth 09:21:00 --- join: Serg[GPRS] (~z@193.201.231.126) joined #forth 09:21:38 re from home 09:23:49 just installed Mirc but still hope to code something oldsch00l-intro-sized ;) 09:26:09 i want city nets to be wireless 09:26:36 to roam from one rented room/provider zone to another, w/o expence related to cabling 09:27:09 and cheaper than cellphone ,) 09:35:50 whee 09:36:50 yeah ! 09:37:21 or i want cheaper data-only cell stations for local providers ;)) 09:37:43 plug 2 tails, add users and have fun 09:39:55 --- join: ASau (~asau@158.250.48.204) joined #forth 09:40:04 Dobryjj vecher! 09:40:16 re 09:42:13 mobile provider raised price up to 1.2$ per megabyte ;(( 09:42:38 * Serg[GPRS] looks at counter 09:54:33 what was it before? 09:55:09 i'm uncertain, but kinda like _10_times less 09:55:52 hmm 10:07:40 I would just wait until I found a cheaper method =) 10:08:44 i spent 35k this session ;) 10:09:03 Kbytes, not K$ ;)))) 10:10:43 at Saturday i gone to free state lawer ,) 10:11:02 helped them a little w/ faulty Win98 10:12:03 traded a CD of two magz/few years for a recovery stuff 10:12:29 nice guys ,) crackers 10:23:35 1,3359375 RUbles spent ;) 10:35:14 so. lets compare. what filetype explanation would i prefer to read. http://astronomy.swin.edu.au/~pbourke/dataformats/bmp/ oooooorr... http://www.w3.org/TR/1996/REC-png-19961001.html 10:39:07 flip a coin 10:40:23 sup forthers 10:43:32 --- quit: Serg[GPRS] ("maybe BRB in half/hour") 10:50:09 heh, use libpng if you want to open png files 10:56:10 --- join: cedricshock (~cedricsho@209-181-58-5.eugn.qwest.net) joined #forth 10:59:12 --- quit: ianp ("Hey! Where'd my controlling terminal go?") 11:04:06 is there a word that converts a little endian number to big endian? 11:06:10 : flip split swap join ; 11:06:11 hehe 11:07:04 split? 11:07:16 it was in fpc, its in isforth :P 11:07:23 oh 11:07:41 well appearantly the size field can be skipped so i wont have to bother :p 11:08:41 i do have a question though 11:09:06 if i create a variable of 2 bytes (in a 32bit implementation, where one cell = 4bytes) 11:09:18 and i store something to this variable, 15 data ! 11:09:25 does it store 2 bytes or 4 bytes? 11:09:29 yes 11:09:38 because it seems, that when fetching from it, it fetches the 2 bytes, instead of the 4 bytes 11:09:46 create foo 2 allot 11:09:54 theres a 2 byte variable 11:10:08 but you need to have a 16 bit fetch word (usually w@) 11:10:08 aye but when you do: 0 foo ! 11:10:10 foo w@ 11:10:17 nope - that will NOT work 11:10:20 it would be w! 11:10:21 hmmm is that implementation dependent then? 11:10:24 yes 11:10:28 ah pity 11:10:29 okay 11:10:39 then i'll go by it the hard way, just to make certain :p 11:10:48 i noticed win32forth does this 11:11:39 neither fetch or store are size smart, they only know address and data, they dont know that the variable was cretaed as 16 bits 11:11:41 or perhaps just for 1 byte vars... 11:11:52 1 byte vars would be c@ and c! 11:11:57 yeah but... hold on 11:12:18 hm 11:12:34 ignore this part. i dont know what...how...who :\ 11:12:52 lalalalala nice weather aint it? 11:12:53 :) 11:37:16 needs more isforth 11:38:06 can isforth do asynchronous I/O? 11:42:59 --- join: networm (~networm@L0664P22.dipool.highway.telekom.at) joined #forth 11:51:13 --- join: Sonarman (~matt@adsl-64-169-93-183.dsl.snfc21.pacbell.net) joined #forth 12:39:51 hmmm 12:39:59 dunno :) 13:01:41 --- join: Herkamire (stjohns@h000094d30ba2.ne.client2.attbi.com) joined #forth 13:01:53 --- join: wossname (wossname@Toronto-HSE-ppp3698988.sympatico.ca) joined #forth 13:14:16 --- part: gl left #forth 13:17:08 ok. i'm working on some code to create an image. now i need to create a buffer before actually writing the image. this buffer of course varies according to dimensions of the image. 13:17:33 am i forced to create a static buffer for this purpose? 13:17:53 or can i specify dimensions and create a buffer when the word is invoked? 13:21:08 yup 13:21:13 :| 13:21:17 just create the buffer when you need it 13:21:21 oh 13:21:23 how? 13:21:45 well just make a word that takes in the size and creates a variable 13:21:52 treat hte variable as the buffer 13:22:01 but how will i name it? 13:22:06 i mean, give it a name 13:22:12 name what? 13:22:16 the picture or the buffer? 13:22:26 the buffer 13:22:33 : createbuffer variable buffer ; 13:22:49 cant you name the variable to be "buffer" ? 13:22:54 or do you need to have more than one buffer? 13:23:00 wont that merely create a buffer of 4 bytes? 13:23:06 (or 1 cell, anyhow) 13:23:25 just a sec 13:23:54 and one buffer is enough 13:26:34 you can use ALLOCATE and FREE and RESIZE like malloc() and free() and realloc() 13:26:43 inside a word? 13:27:00 hmmm and i'd have to save the location manually? 13:27:00 yup 13:27:20 ALLOCATE'll return the address of the buffer 13:27:31 so allocate and not alloc? 13:27:52 in ANS it's called ALLOCATE 13:27:55 VARIABLE BUFFER ALLOT 13:28:03 use ALLOT 13:28:05 is there a common name for ROT ROT? 13:28:06 ALLOCATE is for weenies 13:28:07 oh sorry i did mean, allot not allocate. 13:28:07 :P 13:28:11 futhin: stfu :) 13:28:21 Sonarman: heh :P 13:28:32 i.e., rotate into the other dir 13:28:39 networm: -ROT 13:28:49 he doesn't know the buffer size at edit-time, so he can't use aLLOT 13:28:50 ah, thanks 13:29:03 Sonarman: which one is prefered by chuck moore ALLOT or ALLOCATE? :P 13:29:08 * qFox chuckles... 13:29:08 he can get the buffer size 13:29:21 he doesn't need the buffer size at edit time 13:29:27 oh, RIGHT 13:29:28 heh 13:29:43 : CREATEBUFFER VARIABLE BUFFER ALLOT ; (pass size of buffer to it) 13:30:02 okokok, say i have this word; : createbuf ( width height -- ) * (allot?) ; or what 13:30:20 and variable wont allocate 4 bytes to it? 13:30:22 qfox: the word i just typed is valid code 13:30:30 ok 13:30:44 you pass the number of bytes you need 13:30:48 to CREATEBUFFEr 13:30:49 variable will just be the name you use for the buffer 13:30:52 and how will i use this buffer in a word before i create it? 13:31:09 because its xt needs to be known at compile time 13:31:11 you also need a word to store all the info to the buffer 13:31:14 futhin: I think you mean CREATE not VARIABLE 13:31:18 using a do loop or something 13:31:25 but if i create a new buffer at runtime, there will be a new xt, wont there? 13:31:56 Herkamire: sure CREATE can be used.. (i'm a bit rusty) 13:32:13 : variable create 0 , ; 13:32:22 VARIABLE buffer 13:32:24 --- quit: networm ("Client exiting") 13:32:25 : init-buffer ( w h -- ) HERE buffer ! * ALLOT ; 13:32:35 qFox: the buffer will be created when the CREATBUFFER is run. the name "buffer" will give the xt 13:32:37 --- join: lalalim_ (~lalalim@p508AA79D.dip.t-dialin.net) joined #forth 13:32:52 hmmmmm 13:32:54 <--- hella rusty w/ forth coding :P 13:33:03 i love all this confusing stuff btw :\ 13:33:08 >:D 13:33:24 hmm, but then with my init-buffer word above, there's no way to free the memory 13:33:46 unless your Forth provides a factor of FORGET that lets you specify the new end of the dictionary 13:33:57 qFox: you don't normally add new words to the dictionary at runtime 13:34:07 for now, lets pretend i dont really care about what happens to the buffer. if i need new memory, i'll close and open a new win32forth window. the vm can clean up for me 13:34:22 qFox: you'd more likely use ALLOCATE which will return an address to the memory 13:34:28 Herkamire> i understand that, but i think its obvious that it is desired in a cirumstance like this. 13:34:46 what circumstances? 13:34:52 I highly doubt it 13:34:55 well if you run the program once then you don't need to worry about freeing the memory 13:35:13 Herkamire> creating a program that creates a new image with variating dimensions... 13:35:39 just FORGET the first word of the program after you're done and that'll clear the buffer 13:36:10 s/clear/free 13:36:41 qFox: I don't see why the program would add words to the dictionary at runtime 13:36:55 its not adding words as such, its adding a variable.. 13:37:02 you'd just have three variables: buffer width height 13:37:16 actually the width and height would be a parameter, but w/e 13:37:28 parameter to what? 13:37:30 --- quit: cedricshock ("Leaving") 13:37:33 yeah, and buffer isn't THE buffer, it's a pointer to the buffer which you allocate at runtime without needing to define a word at that time 13:37:34 to the word creating the image... 13:37:39 Herkamire: it would be values passed on the stack 13:38:01 right 13:38:04 but... wouldn't you want to know the dimentions afterwards too? 13:38:21 that can be put in variables, but thats not an issue really 13:38:23 just dup them on the stack if you want them afterwards 13:38:30 width @ height @ new-image buffer ! 13:38:32 either will work 13:38:33 variables = lame :P 13:38:49 variables are not lame 13:38:59 * qFox chuckles... will try 13:39:12 proper factoring destroys the need for most variables 13:39:35 : new-image ( width height -- addr ) * pixel-depth * allocate ; 13:39:50 futhin: no it doesn't 13:40:04 chuck uses lots of variables 13:40:11 i would have to agree with Herkamire. 13:41:09 he doesn't use as much variables as a bad forth coder does 13:42:00 right 13:42:13 sometimes it's better to use the stack 13:42:34 with a forth cpu that's certainly the case 13:43:01 using the stack makes the code reentrant too 13:43:14 but there are a great many cases where a variable is the best way 13:43:16 --- join: arke_ (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 13:43:19 --- quit: arke (Read error: 104 (Connection reset by peer)) 13:43:21 often it's a choice between that and stack juggling 13:44:05 reentrant makes my ass twitch 13:44:14 heh 13:44:17 technically the stack is faster, but with most forth implementations this makes hardly no difference... 13:44:19 --- quit: lalalim (Read error: 110 (Connection timed out)) 13:44:21 j/k 13:44:24 if any at all 13:44:26 i need to make some sort of multi tasker for isforth 13:44:42 And you need to code the resolver! 13:44:44 * Robert runs away 13:44:44 : create.buffer ( width height -- ) here buffer ! * 3 * 54 + allot ; 13:44:49 qFox: what do you mean technically? 13:44:51 proper factoring of code leads to minimal stack juggling therefore less need for variables 13:45:16 futhin - not always 13:45:24 Herkamire> because in a vm, like win32forth, its all memory, so its basically not faster then a hardware stack would be on a cpu 13:45:45 sometimes you just need to totally rethink your code to get rid of variables or excessive stack juggling 13:45:54 then = opposed to what 13:46:03 futhin: right. that works great, but sometimes you need to access 6 items in a random order. 13:46:35 i440r: yup. thats implicit in even attempting to code in Forth PERIOD 13:46:44 I totally agree that you should try to factor and refactor to minimize stack juggling and variables 13:47:00 but you can't get away without either 13:47:12 its implicit in attempting to code 13:47:41 Herkamire: a bad forther will end up trying to jugglie 6 items, a good forther will limit it to 3 items without going overboard with variables 13:47:42 I would asume in most forths (hardware or otherwise) a variable would be faster than accessing the 3rd stack item 13:48:14 futhin: agreed :) 13:49:21 herk no 13:49:49 i think accessing any stack item in isforht is faster than a variable 13:50:01 I440r: why? 13:50:08 some_variable: 13:50:11 call dovariable 13:50:14 dd xyzzy 13:50:25 dovariable pops the return address off the stack 13:50:36 erm no 13:50:42 oh I see 13:50:44 lol 13:50:52 pop top of stack to eax 13:50:54 push ebx 13:51:01 mov data pointed to by ebx into ebx 13:51:21 mine just compiles a LIT 13:51:22 erm i mean data pointed to by eax eve 13:51:31 a variable compiles a lit ? 13:51:35 my const does taht 13:51:51 my compiler doesn't know the diffirence between variables and constants 13:51:58 the goal isn't strictly speed, the goal isn't strictly smallest footprint, the goal is striving for the simplest program which automatically leads us to better solutions and ultimately the program will be fast and small and bloat free and beautiful 13:52:01 neither does my 8051 forth 13:52:10 a variable is a constant address 13:52:18 futhin: here here! 13:52:26 right. 13:52:32 the goal isnt negative bloat 13:52:37 its negative obfuscation 13:52:52 bloated slow software is prefereable to the crap most people puke out 13:52:56 my compiler treats (green) numbers, hex numbers, variables, and constants exactly the same. it just calles LIT 13:53:01 as long as its readable 13:53:27 i think colorizing forth is a backwards step 13:53:30 i440r: simplest program = zero obfuscation 13:53:34 THAT obfuscates it 13:54:02 because in forth the developer of the compiler can chose ANY set of colors he wants for whatever he wants 13:54:12 dudes... obfuscation is a word that only makes things harder to understand you know... 13:54:14 :\ 13:54:17 colorforth is instant gobbldegook 13:54:32 from latin obfuscare - ob meaning intense 13:54:37 fuscare meaning darkness 13:54:42 I440r: in a few keys, you can have my forth display the sources without color 13:54:47 then just say darkness 13:54:48 :p 13:55:03 were coders, we like the word obfuscated 13:55:07 it makes us sound smart 13:55:08 qFox: but obfuscation is such a cool word 13:55:16 what he said! 13:55:25 qfox: obfuscation means something different than darkness 13:55:32 but but but 13:55:37 that's like trying to get me to say "chick peas" instead of "garbanzo beans" 13:55:38 [22:54:34] fuscare meaning darkness !!! 13:55:52 if its obfuscated its wrapped in intense darkness. its hidden. its not easilly descernable 13:56:10 or however thats spelt :P 13:56:27 desce... you did that one on purpose ;p 13:56:32 the word obfuscation is derived from doesn't necessarly mean the same thing 13:56:47 discernable 13:56:49 i believe that sentence misses a few words 13:56:50 hehe 13:57:09 doesn't necessarily mean the same thing as obfuscation 13:57:45 futhin i think the people who invented the word "obfuscated" meant exactly what we use it for 13:58:26 if i wasnt quite sure they were already dead, i'd say they ought to be shot 13:58:28 ;) 13:58:59 (oh and i am just joking about this whole thing eh..) 14:00:52 I haven't even thought about writing reentrant code 14:03:26 *gasp* 14:03:30 blasphemy :P 14:03:47 i thought well factored code was automatically reentrant for the most part 14:03:59 especially if you avoid variables and stuff 14:04:02 no it's not 14:04:18 not if you change data along the way 14:04:34 not if you use variables 14:04:37 or arrays etc 14:07:59 if i use 14:07:59 value buffer 14:07:59 : create.buffer ( width height -- ) here to buffer * 3 * 54 + allot ; 14:08:12 will i successfully free hte buffer using FREE buffer ? 14:08:40 no 14:08:48 you can't free allot 14:08:54 ok 14:09:04 use allocate for dynamic memory allocation 14:09:12 i'll just close and reopen when i get Error: cb Out of System memory! 14:09:14 :p 14:09:24 or just allot enough memory for your biggest image, and never change buffer 14:09:38 nah thats ok. this is my preferred method 14:09:51 whateve 14:09:52 r 14:09:54 :) 14:10:12 hm 14:10:18 well this sux... 14:10:45 either use dynamic memory allocation, or allot enough for your bigest requirement. 14:10:49 that goes for any programming 14:10:52 it appears i cannot create a variable of.. 375k 14:11:01 what? 14:11:07 (to create a 500x250 px img) 14:11:19 a variable is one cell 14:11:20 what forth are you using? 14:11:23 win32forth 14:11:34 Herkamire> it should allot bytes, not cells 14:11:36 Herkamire: you can ALLOT to the variable 14:11:50 futhin: yes, but that's stupid 14:11:51 qFox: just use CREATE 14:11:53 just use create 14:12:12 CREATE BUFFER 30535105150110 ALLOT 14:12:16 or something like that 14:12:32 ok, and how do i get around this 14:12:33 Error: Need a NAME to create! 14:12:36 at runtime 14:12:40 I would hope there is some way to get win32forth to give you more memory 14:12:57 (this is caused by CREATE btw) 14:12:57 qFox: give it a name 14:13:00 i am. 14:13:06 : create.buffer ( width height -- ) create buffer * 3 * 54 + allot ; 14:13:12 qFox: no you're not 14:13:26 CREATE is not an immediate 14:13:55 : create.buffer create * 3* 54 + allot ; create.buffer buffer 14:14:25 s/3\*/3 */ 14:14:27 but this still means that the buffer is created at compile time, not runtime 14:14:39 or it would mean i'd have to manually create the buffer before using the word 14:14:46 qFox: if you want to create buffers at run time _________USE ALLOCATE!!!!!!!!!!!!!!!!!!!!!!! 14:14:47 either way was what i was trying to avoid 14:14:56 ALLOCATE ALLOCATE ALLOCATE 14:14:59 THEN DONT FUCKING TELL ME TO USE ALLOT OR CREATE 14:14:59 ffs 14:15:25 I sait 3 times, either create a big buffer ahead of time with ALLOT or do it at runtime with ALLOCATE 14:16:13 I bet ALLOCATE won't give you the out of memory errors either 14:16:53 not all forths have allocate 14:16:58 whats the stack decription for allocate? ( size -- adr ) ? 14:17:04 win32forth does, i checked 14:17:10 so does isforth 14:17:12 --- quit: wossname ("oleh") 14:17:37 : create.buffer ( width height -- ) * 3 * 54 + allocate to buffer ; 14:17:39 will that do? 14:17:40 yeah, not all forths have it, but if you want dynamic memory allocation, use it 14:17:48 qFox: yes 14:17:50 ok 14:18:10 i hate that fucking word "to" 14:18:15 its fucking incomplete 14:18:18 me to 14:18:19 its a fucking STORE operation 14:18:23 !> <-- thats the fucking word 14:18:45 ans is so fucking limp wristed sometimes 14:18:59 ans the GHEY forth 14:19:04 it doesnt bother me and it will allow me to use buffer as a normale variable (eg. returning its adr) 14:19:40 ???? I don't think so 14:20:38 I never use TO, so maybe I'm wrong about how it works 14:20:51 0 value buffer 14:20:51 : create.buffer ( width height -- ) * 3 * 54 + allocate to buffer ; 14:21:02 buffer acts as a constant does 14:21:07 value is another fyucking dumbass name 14:21:09 to is the way to store data to the variable 14:21:21 well, or whatever alternative I440r thinks of ;) 14:21:30 to will work on variables, constants, values and deferred words 14:21:37 right. and BUFFER will give you it's value, not the address you can use to change it 14:21:57 so now when i use buffer, it will behave like a constant. once it has the address of the newly allocated space, it'll behave like a regular variable, as far as the user is concerned 14:22:16 your restricting your create.buffer to creating only ONE buffer 14:22:25 thats ok, i only need one 14:22:27 you shouldnt do the store operation inside it 14:22:44 my problem with VALUE/TO is that it hides the fact that there is a level of abstraction. it makes it look like it's a constant. 14:22:58 this makes your code less readable, and bugs harder to find 14:23:46 I440r: it's good factoring if (and only if) he only uses it for that one variable anyway 14:24:08 he even has "buffer" in the name. I think it's pretty clear 14:24:23 Herkamire> how do i free ALLOCATEd space? unalloc? 14:24:41 although... if he only cals creat.buffer in one place in the code, I too would take the "to buffer" out 14:24:50 FREE I think 14:25:10 I've never used it, but that's my recolection 14:26:13 hm anyone know another page with the ansi specs? http://www.taygeta.com/forth/dpansf.htm doesnt work for me atm 14:31:13