00:00:00 --- log: started forth/04.06.10 00:29:36 --- join: htp123 (~tehsux@S010600055d233ab7.gv.shawcable.net) joined #forth 00:38:03 --- quit: I440r (calvino.freenode.net irc.freenode.net) 00:38:03 --- quit: dpb9cpu (calvino.freenode.net irc.freenode.net) 00:38:03 --- quit: cmeme (calvino.freenode.net irc.freenode.net) 00:38:08 --- join: I440r (~FooBlah@168-215-246-243.gen.twtelecom.net) joined #forth 00:39:51 hiya i440r :) 00:41:57 --- quit: htp123 ("*") 00:48:23 --- join: htp123 (~tehsux@S010600055d233ab7.gv.shawcable.net) joined #forth 00:50:40 --- quit: Serg () 00:52:58 --- join: dpb9cpu (~dpb@lexx.daves.net) joined #forth 01:13:15 --- join: Serg31561 (~z@212.34.52.140) joined #forth 02:46:17 --- join: hash (~hash@port-204-54-210.fastadsl.net.nz) joined #forth 02:55:15 --- quit: arke_ (Read error: 104 (Connection reset by peer)) 02:55:24 --- join: arke__ (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 02:56:06 --- quit: htp123 ("*") 03:05:09 --- quit: scope (Read error: 110 (Connection timed out)) 03:21:30 --- part: hash left #forth 03:45:52 Morning, #forth. 03:46:02 morning Robert :) 03:46:20 Hi, angel. Did it work? 03:46:29 did what work? 03:46:37 DTC -> STC 03:46:41 mostly 03:46:48 Still buggy? 03:46:50 hunting the last couple bugs now. 03:47:03 it's almost interactive... it's printing out startup messages and the like 03:47:15 currently just dying on the first loop it encounters. 03:47:33 I think I recognize that situation... 03:48:08 * Robert yawns away for a shower. 03:50:50 found the problem, now to fix it. 03:59:46 * Serg31561 is snarfing various ZX Spectrum forths from http://www.worldofspectrum.org/ 03:59:56 --- nick: Serg31561 -> Serg 04:00:38 i gotta learn to convert taped progs into disks 04:01:16 hehe 04:01:50 i'm trying to fix a broken UNTIL routine in my new STC FORTH core. 04:03:46 * Serg uses own control structs 04:03:49 then again, writing a forth core in 1 day is pretty decent. 04:03:58 yeah ! 04:04:11 and not too many bugs either... 04:04:15 just a couple left. 04:04:24 most of my control structures seem to work now. 04:04:42 i need to improve them soon though... i need a ?DO and a FOR/NEXT 04:04:48 i got some docs on writing own forth but did not took the challenge 04:04:59 this isn't my first... 04:05:14 url ? 04:05:18 i've been toying with rpn languages for most of a decade... i invented one before i ever heard of forth. 04:05:19 url of what? 04:05:37 of your forths, of coz 04:06:02 hehehe, most of them aren't available... an old snapshot of the DTC version of this OS is available on my webpage: http://caladan.nanosoft.ca/cyvos.php 04:08:28 yeah ! 04:08:33 eh? 04:08:39 cool ;) 04:08:44 glad you like it :) 04:08:57 screenshots only, did not run 04:09:26 *nods* 04:09:31 well, i didn't see you reboot either. 04:09:44 * Serg has lotsa boxes around 04:09:59 fair enough. 04:10:12 * Serg is sysadmin 04:10:59 my idea-fix is Unix in Forth ;) 04:11:12 worrie. 04:11:21 but all i did code is litle utils 04:11:40 a game in one-liners w/ Forth console, etc 04:12:10 http://cryptomancer.narod.ru , in Russian 04:12:20 gonna make it bi-lingual ;) 04:19:21 A very good idea. 04:19:59 but CSS stuff sucks - blocks creeps apart 04:20:29 * Robert knows how he would have solved that 04:20:45 tables ? hell no ! 04:21:53 Nah 04:21:57 Plain HTML. ;) 04:22:09 i want it side-by-side ! 04:25:27 oh, figured it out. 04:25:34 tripping over my own return address. 04:25:36 gotta love that. 04:25:50 i hate words that screw around with the return stack in unpleasant ways :P 04:26:03 i used allot instead of allocate 04:26:09 i deserve a gold star for figuring that out 04:26:19 ? 04:26:46 Klaw: Write unallot, and we'll give you a gold star. 04:27:25 couldn't I just restore here ? 04:27:35 Heh. 04:27:37 move the dictionary pointer back or whatnot? 04:27:52 Sure. If you haven't allocated anything else since that. 04:30:42 it'll take writing entire new forth w/ 2 stage RAM indirection and garbage collector 04:31:32 2 stage RAM indirection? 04:31:40 photon induction 04:31:46 hmm... 04:32:02 several addres spaces 04:32:23 one for words headers, other for words itself, their bodies 04:32:55 And what's the point of that? 04:32:57 so you can throw out words from middle and defrag mem, adjusting pointers in headers chain 04:33:06 Ah. 04:33:12 what's wrong with cmove 04:33:25 That's pretty smart. I think I'll steal the idea. 04:33:30 how u gonna fix all calls ? 04:33:36 cmove allocate 04:34:07 Robert: it was done long ago to bypass dos 64k limits 04:34:25 btw not sure for GC, 04:35:25 but it was capable of using all 640 04:50:11 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 06:20:05 --- quit: Robert (Read error: 54 (Connection reset by peer)) 06:23:30 --- join: Robert (~snofs@c-bf5a71d5.17-1-64736c10.cust.bredbandsbolaget.se) joined #forth 06:55:08 --- quit: Serg () 07:28:59 --- join: cmeme (~cmeme@216.184.11.30.unused.swcp.com) joined #forth 09:07:29 --- join: Herkamire (~stjohns@h000094d30ba2.ne.client2.attbi.com) joined #forth 09:41:55 --- join: O3BEPH (~z@193.201.231.126) joined #forth 09:42:19 --- nick: O3BEPH -> Serg[gprs] 09:43:11 * Serg[gprs] is torturing ZX emu 09:43:37 * solar_angel is torturing herself with a mix of evil x86 assembly and scotch bonnet pepper sauce. 09:44:50 Mmmm... x86. 09:45:51 err cam you make an mpg of that for me ? 09:45:54 }:) 09:46:22 what? 09:46:24 nice smiley ;) 09:47:09 _horny_ smiley ,) 09:47:19 evil grin (tm) 09:47:31 I Know Something You Don't Know (TM) 09:47:48 i wanted to see an mpg of you torturing yourself with x86 and pepper sauce lol 09:48:00 it's not much to look at, honestly. 09:48:06 lol 09:48:15 Then paint your impression of it. 09:48:24 me sitting on a couch with a laptop, chips&sauce, coding the night away. 09:48:25 one of zx forhs refuses to load from tape ;( 09:48:30 except that it's really the afternoon and i have no concept of time. 09:51:39 hahaha, that would make one weird mpg :p 09:58:22 lol 10:05:09 --- quit: warpzero ("Tried to warn you about Chino and Daddy Gee, but I can't seem to get to you through the U.S. Mail.") 10:08:17 --- join: warpzero (~warpzero@dsl.142.mt.onewest.net) joined #forth 10:15:16 s" and c" use the PAD right? 10:15:43 or do they actually reserve space for this? 10:27:17 qFox: s" compiles a string 10:27:42 it doesn't really have anything to do with PAD 10:27:56 although the implementation might alter pad in it's parsing 10:28:59 it copies the string somewhere safe (not pad) and compiles instructions to load up the address and cound of the string. 10:29:31 and when you use s" in interpretation mode? 10:29:41 you don't 10:30:08 and when you do? 10:30:12 :) 10:30:20 how else are you supposed to get a counted string? 10:30:29 to give as argument (filename for example) 10:30:39 you don't. the behavior is undefined 10:30:45 huh 10:31:29 --- part: Serg[gprs] left #forth 10:31:44 well... anyways, if i use s" or c" i dont have to worry about it getting overwritten? 10:31:58 huh, looks like I did use s" at interpret time with gforth. the ANS reference says it's undefined. 10:32:32 works for win32forth as well you know.. 10:33:23 but again, say you have a code that works with a file 10:33:31 how would you pass the filename on? 10:33:32 oh, the FILE wordset defines the interpret version of s" 10:33:39 so I guess it is ok to use at interpret time. 10:33:46 stupid stupid ANS 10:33:50 lol 10:33:56 but how would you? 10:34:07 dont tell me you've never done that before? 10:34:29 s" /dev/fb/0" w/o open-file assert constant file 10:35:38 I haven't done much with ans in a long time 10:35:50 and when I did I did't do much with files 10:36:47 well, i'm not quite sure what /dev/fb/0 is, but its still not a filename. what if you need the filename for output? 10:37:10 what do you mean it's not a filename? 10:37:42 well. say this word processes a file. : process.file do.it ; 10:37:57 how would you pass on the fileNAME, while in interpreting mode 10:38:16 (i have to go bbl) 10:38:24 s" /path/to/your/file" process.file 10:46:06 --- join: Serg[gprs] (~z@193.201.231.126) joined #forth 10:48:28 did one have xperience w/ forth on speccy ? 10:49:03 i finally loaded some Forth, but it seems to be in compile mode.... 10:49:18 and i dunno how to go to interpreting mode 10:59:00 --- quit: Serg[gprs] ("good nite to all !") 11:14:14 arke__: Alive? 11:23:55 --- quit: solar_angel ("sleepage.") 12:27:55 but Herkamire: [19:31:59] huh, looks like I did use s" at interpret time with gforth. <-- that implied that you dont use s" (or didnt untill now) 12:27:57 --- quit: warpzero (Read error: 60 (Operation timed out)) 12:28:01 oh 12:28:07 ignore. 12:28:37 i somehow read "works", instead of "i did use" 12:28:39 ffs :\ 12:31:47 now I use herkforth 12:31:56 except for occational code examples 12:32:55 herkforth doesn't have string support yet 12:33:36 and yet I somehow managed all 4 times I wanted a string 12:34:41 hehe 12:36:35 I'll have strings someday 12:36:49 and other data besides source 12:37:44 I plan to have direct support for various kinds of tables 12:37:45 hmmm s" and c" wouldnt seem to be very difficult to implement 12:38:07 it's a little annoying if you don't have seperate code and data space. 12:38:10 but still not that hard. 12:38:51 you just compile a branch that jumps over the length of the string. then copy the string to HERE 12:39:13 (and encrement HERE) 12:39:31 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 12:39:45 exactly, although i'm not sure why you would branch 12:39:52 oh right, and you have to compile LITs to load addr and u 12:40:09 qFox: so it doesn't try to execute the string as instructions 12:40:17 hmmm 12:40:51 the way i see it, get the input buffer pos, read chars of it untill you encounter " (there is no escape char for s" and c" afaik, so thats easy) and copy those bytes 1 on 1 to here 12:41:22 you need a branch 12:41:27 because you're in the middle of compiling a word 12:41:35 isnt a branch an if -- then type of thing? 12:41:48 yeah, except without the conditional 12:41:57 if branches only if TOS is zero 12:41:59 oooh 12:42:03 hmmmm 12:42:13 but s" would be an immidiate then 12:42:19 right. 12:42:31 it would compile a branch, the data, and LITs 12:42:59 but other then that it would act the same as it would at interpretation time 12:43:21 ohwell 12:43:35 you can use it basically the same. 12:44:13 anyways, if s" and c" dont use the pad, i'm happy. i was about to allocate space to copy the data to this space before continueing, because i was afraid it'd get lost otherwise 12:44:18 but then i realized that it might not use PAD :p 12:44:22 s" myfile" w/o open-file should give you the same results as : blah s" myfile" w/o openfile ; blah 12:44:23 hence my question 12:44:32 yeh 12:44:49 I don't know abot PAD 12:44:59 s" does put it somewhere that WORD does not overwrite it though. 12:45:20 otherwise this wouldn't work: s" myfile w/o open-file 12:45:41 (it would try to open a file called open-f 12:45:59 oops, missing quote. I meant: s" myfile" w/o open-file 12:58:02 ugh i hate that sort of code 12:58:11 create myfile ," filename" 12:58:22 myfile w/o open-file 12:58:24 MUCH better 12:58:44 i hate s" with a passion 12:58:55 specially when they do s" blah blah" evaluate 12:59:39 hmmm well i dont always need to store the filename 12:59:52 s" d:\whatever.txt" file-open drop 12:59:56 will work for me 12:59:57 eh 13:00:00 s" d:\whatever.txt" r/o file-open drop 13:01:07 that way i'll have the fileid, which is all you need for file words 13:01:36 but i'm talking to pass on the filename to a word. ," is not really ... "userfriendly", if you will 13:03:38 err were not TALKING about the user now are we, were talking about the CODER!!!! lol 13:03:48 :P 13:03:54 and s" is horrible 13:04:51 s" compiles the string inline but the (s") doesnt DO anything with it the way (.") does. all your doing is injecting an string of characters into the execution stream that you now have to branch arround 13:05:03 the ," method will be faster 13:05:17 and factored better 13:06:25 actually i was talking about the user... 13:07:04 like when you have a program, and somebody else loads that program into forth, and then wants to execute the word that processes a file 13:07:49 i suppose you could do it like: PROCESSFILE d:\whatever.txt ,but thats not really forthish. d:\whatever.txt PROCESSFILE will not work either.... 13:19:36 --- join: Teratogen (~leontopod@67.40.69.97) joined #forth 13:33:08 --- join: lalalim (~lalalim@p508ABE0B.dip.t-dialin.net) joined #forth 13:41:42 Teratogen: Hi, MINIX. 13:41:51 And hi, lalalim 13:51:10 --- quit: lalalim_ (Read error: 110 (Connection timed out)) 14:35:23 --- quit: tathi ("leaving") 14:39:49 MINIX!@ 14:39:58 Yes, hi 14:55:50 Teratogen, using minix? 14:56:53 hi 14:57:54 slava: He's a frequent #minix troll. :) 14:58:39 i have two new docs 14:58:42 http://www.jedit.org/factor/compiler-use.txt 14:58:45 http://www.jedit.org/factor/compiler-impl.txt 15:00:16 OK, bookmarked for reading later (perhaps). :) 15:00:46 I'm looking for Forth compiler information, so thanks. 15:01:12 its only vuagely related to forth 15:01:23 my compiler is a different design 15:01:38 might give ideas nonetheless :) 15:02:05 Yes, I don't care who I steal from. 15:04:09 my compiler is mainly to avoid the inefficiencies of writing an interpreter in Java. my 'compiled' code is about as fast as a DTC forth. 15:04:19 Also, I don't know if I want to use pure Forth (if there were such a thing). 15:04:51 biggest difference between factor and forth is factor has runtime type info. 15:05:16 this adds a lot of complexity. i'm working on type inference at the moment to make it go faster 15:09:17 My design goals are simply to make programming relatively simple while keeping the compiler somewhat efficient. 15:09:35 And to not use a lot of memory. ;) 15:09:59 factor is pretty bloated, 10mb with the entire library loaded, and 20mb if you compile everything :) 15:10:45 --- join: blockhead (default@dialin-109-tnt.nyc.bestweb.net) joined #forth 15:10:46 it won't be any worse than a lisp system though. 15:11:00 since it is at about the same level of 'high-levelness' 15:18:35 - interpret-only - prevent the word from being compiled. Also prevents 15:18:35 words that call this word from being compiled. 15:18:44 so, it first processes a word to find this word? 15:18:51 or how do you do that? 15:19:12 (slava) 15:19:39 qFox, what forth calls 'compile mode' (inside a : ... ;) is 'parsing' for me. 15:19:55 qFox, when : ... ; is read, a new interpreted word definition is added 15:20:01 so you first parse a word in compilemode before writing it? 15:20:01 the user can then use "..." compile or compile-all or whatever 15:20:42 when it sees : it reads a word name, and the definition which is a 'word thread', until the ; 15:20:45 this is stored in the dictionary 15:21:05 if you want the word to go faster, you compile it later, which turns it into bytecode (and presumably native code if you have a recent JVM) 15:22:06 yes but i mean, that line i pasted, that implies that you look at the contents of a word definition before actually defining it, to see whether that interpret-only word is used or not, right? 15:22:38 interpret-only is simply a word that sets a flag in the header of the last defined word 15:23:12 hmmm it sounds a bit confusing to me but allright 15:23:31 if you ignore the compiler then its like forth 15:23:46 (OT. its 0.23h here, and there are still commercials on tv about european elections that were held YESTERDAY...) 15:23:54 hehe 15:24:03 oh i'm sorry i was under the impression this was a forth implementation? 15:24:22 no 15:24:26 its similar 15:24:31 ouch. ok then nevermind at all :) 15:24:35 my bad 15:24:36 but you can't do a "forth" in java... 15:24:47 well, i've done it in mirc 15:24:53 if it can be done in mirc... ;) 15:24:57 does mirc provide direct memory access for @ and !? 15:25:03 no i had to emulate it 15:25:13 that's a bit silly :) 15:25:19 but in the forth environment, you wouldnt know the difference 15:25:38 i just have named variables 15:25:39 same can be done in java, i'm sure 15:25:44 3.14 "pi" set 15:25:45 "pi" get 15:25:51 uhu 15:25:54 and a namestack for pushing new lexical environments 15:26:09 so its kind of your own lang with a forth twist? 15:26:14 in java 15:26:37 yes 15:26:45 neat :) 15:26:57 for example i only have + - * /, not different forms for integers/floats/etc. its runtime-typed 15:27:16 Sorry, got distracted. Heard a Brazilian station on the radio. :) 15:27:22 but you do work with floats, not only int? 15:27:33 fixnum, bignum, float, ratio, complex 15:27:45 1 sqrt . 15:27:45 #{ 0 1.0 } 15:27:56 that should be -1 sqrt :) 15:28:20 imaginary number theory? :) 15:28:31 yup 15:28:34 you know euler's formula? 15:28:35 nice 15:28:38 no 15:28:43 you know number e? :) 15:28:58 i just remember that imaginary number stuff. x^2=-1 15:28:59 yes 15:29:03 e pi i * ^ . 15:29:03 #{ -1.0 1.2246467991473532E-16 } 15:29:17 that should be #C{ -1.0 0 } but pi is the float approximation not the exact number obviously:) 15:29:37 hmmm, what does i represent? 15:29:49 i -1 sqrt = . 15:29:50 t 15:29:55 i is sqrt(-1) :) 15:29:59 ah ok 15:30:48 so e^(pi*SQRT(-1))=0 ? :) 15:30:54 = -1 15:31:05 it is? 15:31:07 hmkay 15:31:15 #{ -1.0 1.2246467991473532E-16 } means -1.0 + i * very small roundoff error 15:31:33 #{ a b } is the syntax for a + b*i ( #{ is a parsing word as you might guess :) ) 15:31:34 imaginary numbers have been too long for me to know more then that they exist :) 15:32:49 Hm, I've been wondering about implementing floating-point support in Forth. Do you know of any simple code to read/print floating-point numbers? 15:33:48 in forth its easy but thats probably not what you meant :p 15:34:17 Yes, preferably Forth. 15:34:38 But assembly language routines would be appreciated, too. 15:34:58 well, the extended float instruction set has F. 15:35:03 i'm sure that you can figure that one out ;) 15:35:10 (ANSI) 15:35:41 http://www.taygeta.com/forth/dpans12.htm <-- the whole thing 15:35:42 Robert: you inquired the state of my health earlier? 15:35:57 arke__: Wondering how your project is going. 15:36:13 qFox: I mean how to implement it. 15:36:56 right :) 15:37:10 never done floats myself so cant help ya there 15:37:20 Sinner. 15:37:41 Robert: remember Rkg? I studied old notes of mine, and I have decided that RkG will be its interface, but in a redesigned way. 15:37:53 No, I don't. What is it? 15:38:13 RkG was supposed to be like a network application maker thingy 15:38:40 What does that have to do with a Forth system? 15:38:41 anyway, it'll provide graphics and other stuffs in a networked fashion 15:38:59 Robert: RkG will be the means oif writing non-trivial Forth programs on the system 15:40:32 "non-trivial programs" makes me think of good old Smerdy. 15:40:42 ack 15:40:43 forgive me :( 15:40:45 --- nick: arke__ -> arke 15:40:46 :D 15:40:49 Robert: lemme restate that 15:40:54 No problem.. 15:40:59 what is rkg 15:40:59 Robert: programs which do not use simple I/O :) 15:41:04 slava: read above 15:41:33 i can't find it :) 15:41:36 :-P 15:42:46 slava: RkG = the means in my system to do non-simple I/O in a networked fashion (think X11, but not quite) 15:43:02 hmm cool 15:44:07 slava: used to be written in C, using curses as an interface, X11 planned, but I changed that 15:44:25 arke, so like a vt100 but higher level protocol? 15:44:37 slava: I'm sure if you look at logs from the november 2003 era you might see me talk about it 15:44:50 slava: kinda. It's hard to explain. 15:44:53 arke, is it available for download? 15:45:01 slava: no. 15:45:06 It doesn't exist except on paper 15:45:17 (notes I made during class for the initial plan) 15:46:07 hmmm 15:46:14 is f** the only word to power something? 15:46:24 x ^ y 15:46:51 --- join: Kryptech (~root@12.147.148.157) joined #forth 15:47:14 qFox, i prefer ^ :) 15:47:18 what is forth? 15:47:28 yes but ^ doesnt seem to be hmm 15:47:34 defined. 15:47:41 so thats why i'm asking :) 15:47:46 neither is ** for that matter 15:47:47 qFox, : ^ f** ; :) 15:47:51 Kryptech> a programming language 15:47:53 Kryptech: hello there, UNIX C code. 15:47:54 slava> hate. 15:48:15 coder... 15:48:22 thats what I meant :) 15:48:31 UNIX C coder... 15:48:43 Kryptech: Forth is the way the mice found the answer to life and the universe (which happens to be 42) 15:48:44 what does it's syntax look like? ;-) 15:48:55 Kryptech: Forth doesn't have syntax 15:49:08 arke: wtf? 15:49:14 arke..? 15:49:18 how can it not have syntax? 15:49:31 Kryptech, the syntax is user-definable 15:49:35 well in some remote way is probably right 15:49:37 Kryptech, its just whitespace-separated words 15:49:48 wow! 15:49:58 what does a forth program look like? 15:50:13 and it doesn't even have to be whitespace separated :) 15:50:20 Kryptech: however you want it to. 15:50:49 Kryptech> http://galileo.phys.virginia.edu/classes/551.jvn.fall01/primer.htm 15:50:50 how are the primitave operators implemented? (sp?) 15:50:50 start reading 15:51:17 Kryptech, either in another language (C, assembly) or cross-compiled from another forth 15:51:17 for a second there i thought i lost that link :\ 15:51:22 Kryptech, there are forth implementations written in forth 15:52:06 hmm... what are forth's limits? 15:52:17 gui, sockets, what/ 15:52:18 ? 15:52:18 the programmer 15:53:13 well there is no real wide accepted standard. well there are, ansi and fig, but everyone seems to want to spit on standards 15:53:30 so, can you use gui and sockets and threads and stuff? 15:53:35 * blockhead spits on the standard :D 15:53:37 Kryptech, its turing-complete and can interface with your operating system, so of course 15:53:51 okay... 15:54:01 if you're in windows, you can but it'll cost you some trouble. in unix/linux its probably easier 15:54:16 * Kryptech is linux! ;-) 15:54:25 i kinda got that 15:54:49 oh, sorry... that was obvious earlier, wasnt it? ;-) 15:55:08 what is the learning curve for forth? 15:55:22 fast. its easy 15:55:29 * blockhead hands out beers all around. His forth's outer ineterpreter is finally working 15:55:33 yay! 15:55:40 plus if you're used to RPN, you basically already halfway :p 15:55:40 Kryptech: depends on you. For some, it's y = sqr(x). for some, it's y = x 15:55:57 cool... 15:56:07 I know C and python... 15:56:21 and a whole bunch of others (a little bit) 15:56:34 then i dont think you'll have much of a problem with it 15:56:36 is forth a good language to code with daily? 15:57:24 can you implement big projects in forth? 15:57:41 yep :) 15:57:54 start reading the tuto i pasted 15:58:13 its the one i started with, mainly 15:59:22 oh... the preface here makes forth soud heavenly... 15:59:36 :) 15:59:41 matter of interpretation ;) 16:00:58