00:00:00 --- log: started forth/03.09.18 00:05:12 --- quit: Grant_P (Read error: 110 (Connection timed out)) 00:36:29 --- join: gilbertbsd (~gilbert@fl-nked-ubr2-c3a-37.miamfl.adelphia.net) joined #forth 00:36:51 A B program is implemented as a reverse Polish threaded code interpreter: 00:37:01 that sounds very much like a language I've heard of ... 00:37:07 just can't put my finger on it :| 00:37:13 http://cm.bell-labs.com/cm/cs/who/dmr/kbman.html 02:59:35 --- log: started forth/03.09.18 02:59:35 --- join: clog (nef@bespin.org) joined #forth 02:59:35 --- topic: 'Welcome to the most versatile and extensible and l33t programming language (which incidentally started out as a prank) ---- Forth forum: forum.forth.bespin.org ---- Talk to us about all of your ANS/Color/whatever Forth needs .... and don't forget to say Moo.' 02:59:35 --- topic: set by arke on [Tue Sep 16 22:34:22 2003] 02:59:35 --- names: list (clog Grant_P skylan onetom _arke_ njd rpc ianP I440r-wrk @ChanServ Robert TreyB) 03:37:52 --- join: Serg_Penguin (Serg_Pengu@212.34.52.140) joined #forth 03:38:06 --- quit: Serg_Penguin (Client Quit) 04:51:36 --- join: Serg_Penguin (Serg_Pengu@212.34.52.140) joined #forth 04:52:48 --- join: Grant1 (~pyromania@202.59.99.234) joined #forth 05:08:26 --- join: hovil (~matt@CommSecureAustPtyLtd.sb1.optus.net.au) joined #forth 05:11:03 --- quit: Grant_P (Read error: 110 (Connection timed out)) 05:12:43 --- quit: Serg_Penguin () 05:13:55 --- join: Serg_Penguin (Serg_Pengu@212.34.52.140) joined #forth 05:14:03 --- quit: Serg_Penguin (Client Quit) 05:14:57 --- quit: Grant1 (Read error: 104 (Connection reset by peer)) 05:17:24 --- join: SDO (~SDO@co-trinidad1a-42.clspco.adelphia.net) joined #forth 05:39:58 --- quit: hovil ("No windows for this server") 05:46:55 --- join: mur (~mur@h31n2fls31o965.telia.com) joined #forth 05:52:02 Terve, mur 05:52:14 Today mur's IRCing from Sweden ;) 05:56:53 today i was feeling shittyu 05:57:03 i shoudlnt have started working in pairs 05:57:06 i want to change subject 05:57:10 Hm? 05:57:10 of the "poster" 05:57:22 school work 05:57:27 we are doing poster 05:57:31 Oh.. 05:57:34 ended 1/2 of the part 05:57:49 next perioid of the course we'll be working on computer 05:58:02 the person (telling about others in school is a bit hostile environment :) 05:58:47 anyways he's maintaining hardly resolving / unresolving array of thoughts in poster and he cant put them down to some decent amount of .. which woudl be required by the poster 05:59:03 Ehm, OK. Poor little mur having to work with such a person 05:59:06 ad it's hard to work, i want to do some political 05:59:10 yes 05:59:13 sad world 05:59:27 i coudl ask to do someting else next part of it 05:59:31 if it's possible 05:59:31 :P 05:59:34 Hehe. 06:00:31 visited btw new restaurant (where you get cheap student meals also) nearby. it's very nice place. i will visit it again. tomorrow they'd have has vegetarian (not sure if those are for real) some paprika mushroom pizza. i dont like to eat mushrooms. 06:00:47 cleared screen. in school anyways. 06:00:49 ;) 06:01:56 --- quit: mur ("Changing server") 06:03:06 --- join: mur (~mur@h31n2fls31o965.telia.com) joined #forth 06:04:59 wb 06:05:13 * Robert likes to eat mushrooms. 06:05:18 Don't you even like kantareller? :) 06:05:27 not any type 06:05:41 i coudl draw them, they look nice, but i dont like they feel in mouth 06:05:41 :( 06:06:51 hehe to the topic 06:07:30 Bleh. 06:07:34 tomorrow we have freehand drawing 06:07:49 practising style 06:07:54 i took nice digiphotos 06:07:54 I'm not that good at drawing stuff. :/ 06:07:59 me neither 06:08:22 next week we have superbooked living human drawing 06:08:24 :P 06:08:40 starts 8.30 and 30 min dinner break and ends 16.30 06:08:43 Heh. 06:08:49 Have you hired a naked lady? 06:08:53 => 1.5 hours more than normally 06:09:05 minus the coffee etc breaks 06:09:15 <_arke_> mur: terve. 06:09:22 terve rk 06:09:38 --- nick: _arke_ -> arke 06:09:42 i think Rk might mean riffle in finnish military abbreviations 06:09:56 not sure,but iirc 06:10:21 rifle? 06:10:23 hehe 06:10:33 yes, i've got a big rifle. 06:10:43 well military one 06:10:53 something like rk m-86 what ever 06:11:08 hehe 06:11:23 i dont remember models and dont care to 06:11:34 * arke eyes Robert 06:11:44 Robert: i just remembered... 06:11:49 m-86 implies a mauser 06:11:55 not sure what teh rk would be 06:12:02 M8 6 06:12:16 might have some interesting ideograms if you think teen way 06:12:49 Robert: i wanted to tell you that your parsing list is 18 bytes, without loops, and mine is 16 bytes, WITH loops 06:12:53 the mil officers (all around world) have so unimaginary way to name stuff with M representing word model and year number 06:13:10 so m03 riffle woudl be this year model of civil killer 06:15:43 Robert: also, my init code is smaller 06:16:42 Ehm. 06:16:51 What init code? 06:17:22 the code that every compiled proggie that you produce has at the beginning 06:17:36 OK. What does your look like? 06:18:07 mov bx, 0x9090 ; actual adress is filled in later 06:20:55 Robert: but either way, what is that extra byte in your token list? 06:22:57 oh, nevermind. 06:23:00 i found it :) 06:25:25 Eh. 06:25:52 Can you paste your full initialization code, please? 06:25:53 hm 06:25:58 For the brainfuck program. 06:26:50 this is the init code for the compiled program, not the actual compiler 06:27:08 I know. 06:27:11 Paste it. 06:27:18 (For the compiled program= 06:27:19 ) 06:27:24 mov bx, 0x9090 ; actual adress is filled in later 06:27:26 xor ax, ax 06:27:46 And then? 06:27:52 mov cx, 0xffff ; any better way to do that? 06:28:06 rep stosw 06:28:08 ack! 06:28:10 wtf 06:28:14 That'll crash everything. 06:28:15 its mov cx, 0x00ff 06:28:18 Oh. 06:28:24 :D 06:28:54 erm, damn, i forgot something 06:28:59 cx = 0xff 06:29:00 mov di, bx 06:29:04 When the program starts. 06:29:11 And ax is usually 0. 06:29:22 So those instructions aren't needed. 06:29:36 oh fun! 06:29:44 bx is 0 at the beginning. 06:29:48 dx is undefined. 06:29:54 di = 0xfffe 06:29:55 well so you'd think! 06:30:01 si = 0x0100 06:30:11 but wait until the processor / os developers have something to say 06:30:20 bp = 0x09xx (where xx is undefined). 06:30:40 wow, i can cut out major bytes that way! 06:31:33 * mur bytes arke a bit 06:31:41 7 bits left 06:31:47 :) 06:33:00 :) 06:33:07 and ax, ax = 2 bytes, no? 06:33:18 test ax = 1 byte, but does the same, no? 06:33:24 Eh. 06:33:28 You can't do that. 06:33:33 What are you trying to do? 06:33:36 damn. 06:33:48 and ax, ax; jz .eof 06:34:08 i count 5 bytes there. must be less. 06:34:11 That's a good way, and so is "or ax,ax" which I prefer to use. 06:34:15 I count 4 bytes. 06:34:20 or ax,ax ; 2 bytes 06:34:21 and ax, ax = 2 06:34:23 jz .eof ; 2 bytes 06:34:26 jz = 1 06:34:28 .eof = 2 06:34:30 No. 06:34:35 8-bit relative offset. 06:34:40 In the jump. 06:34:50 oh. 06:34:52 sw33t 06:35:00 4 bytes, then. 06:35:06 but the and part is still 2 bytes 06:35:22 isnt test ax the same as cmp ax, 0????????? 06:35:27 i can't bite you 6 more bytes, how'd you manage with one bit?! 06:35:34 You can't do "test ax". 06:35:43 erm 06:35:47 what does test do? 06:35:53 test 06:36:00 i think i migght have something else in mind here .... 06:36:03 arke: It's just like AND, but without storing the result. 06:36:21 erm 06:36:23 so 06:36:23 arke: So "and ax,bx" and "test ax,bx" sets the same flags, but test doesn't modify ax. 06:36:37 test ax, bx = 2 bytes tho, right? 06:37:26 Yes. 06:46:04 damn. 06:46:17 any one-byte way to do that? 06:49:13 gotta go to school, later 06:49:33 go school later 06:49:42 excuse with forth script 06:50:03 like perl's haxors /24sfepamfe/#@idfn24 06:50:10 (#)(%/)#(/%)"(&"=)&#(=)#(%=)#(%=)#(%ASFS 06:50:29 then you can say, "but it is valid code, how can you blame about it?" 06:56:05 bad joke :P 06:56:13 Robert: school's over? 06:58:06 --- quit: arke (Read error: 60 (Operation timed out)) 06:58:14 Yes. 07:05:55 --- quit: mur ("cd-rw todo") 09:09:41 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth 09:15:05 Hi Herkamire 09:32:45 hi Robert :) 09:33:57 I am soon to celebrate my 500th commit to my personal repository 09:34:03 I'm at 499 09:37:03 Heh, congratulations :) 09:40:43 and how old is your project ???? heheh 09:44:32 almost 8 months I think. 09:44:51 that's commits for the whole repository though. 09:45:03 but most are my forth project 09:46:43 the repository has been running since: 12 Jul 2002 09:48:09 that's still averaging more than a commit per day :) 09:51:33 Heh. 09:52:05 That's pretty much, more than my yearly backup of the projects/ dir ;) 09:52:59 yearly? 09:53:13 I try to backup every week 09:53:37 I'm so lazy. 09:53:56 And "backuping" usually means putting it on the web server, which is on the same HD. 09:54:39 same HD isn't backing up, unless you are just trying to protect against deleting stuff by accident 09:55:13 my major concern is HD corruption or lightning or something. 09:55:54 Yeah... I just need to find somewhere to backup to. 09:56:03 Or, well.. I do have a couple of shells. 09:56:08 They probably don't mind. 09:56:55 I really like having my own server :) 09:57:07 This computer is my server, heh. 09:57:21 On a semi-stable 0.5mbit line.. hooray. 09:58:34 eventually I will have two servers that automagically backup onto eachother. (probably hourly) 09:58:48 but for now it's manual, and just between the server and my home computer 10:46:12 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 11:25:02 --- join: squire (vikas@61.1.138.83) joined #forth 11:27:49 --- quit: squire ("Client Exiting") 12:22:02 --- join: bashfire (~bashfire@pD95458DE.dip.t-dialin.net) joined #forth 12:25:56 terve 12:33:09 Hyvää iltää 12:33:32 ää? 12:34:47 btw is there still a forthofftopic#? 12:35:08 No idea. 12:35:22 I think off-topic discussions are allowed in here again, when thin is away. :) 13:07:02 --- quit: tathi (""rebooting"") 13:15:41 --- join: thin (thin@bespin.org) joined #forth 13:17:54 --- topic: set to 'forth did not start out as a prank, are you thinking of a different language?' by thin 13:18:04 --- quit: thin (Client Quit) 13:31:59 Heh. 13:38:18 --- quit: Herkamire ("life, liberty and the persuit of happiness") 13:42:08 Robert: hey what? 13:43:49 "heh", i said. 13:44:03 I was just suprised by thin's appearance. 13:45:09 why surprised? 14:22:05 --- join: wossname (wossname@HSE-QuebecCity-ppp80661.qc.sympatico.ca) joined #forth 14:34:38 Yay. 14:34:40 I love Forth. 14:34:49 Spent 2 hours debugging 3 lines of code. 14:34:55 @$#! 14:34:58 Hi wossname 14:35:10 hi robert 14:35:19 i hope your lines of code had many symbols 14:35:53 lol 14:35:56 you moron 14:36:05 :P 14:36:24 is @$#! a valid word ??? "_ 14:36:28 what does it do ??? heh 14:36:30 I440r-wrk: Worst thing of all, I ripped those 3 lines off from isForth!! 14:36:35 lol 14:36:36 what 3 lines 14:36:50 I440r-wrk: ;code does> ,call 14:37:01 thats more than 3 lines 14:37:07 ;code is tricky :) 14:37:11 does> uses ;code 14:37:15 Nah. 14:37:16 do you have them working now ? 14:37:19 Yep :)) 14:37:30 you need ;uses too 14:37:36 i think ;uses is more efficient than does> 14:37:58 : (;code) r> last 2+ @ 1+ dup>r 2+ - r> ! ; 14:38:35 : does> (compile) (;code) (,call) (does) ; immediate 14:38:45 (does) is dodoes btw 14:38:52 figured that heh 14:39:03 Which is ported from isforth and converted to 16 bits. 14:39:13 Now things work fine. 14:39:13 thers a reason they are called "do" words - dovariable, doconstant etc 14:39:48 Well, I just put parens around do-words... looks nicer :P 14:39:52 Anyway 14:40:04 parens implies primatives to a word 14:40:17 which isnt wrong, just isnt as right here as it could be :) 14:40:23 like (lit) 14:40:28 which some people call DOLIT 14:40:29 lol 14:41:12 I really got to love even more Forth now, when does> etc., for/next, begin/until/repeat, could be implemented in like 8 lines. 14:41:25 * Robert would like to se the gcc developers do that. 14:42:27 its begin/until and begin/while/repeat and begin/again 14:42:47 they could do it in 8 lines 14:43:01 but they would have 20489652904886290475 lines underneath those 8 lines :) 14:43:33 Hehe. 14:43:59 Here I just have branch/?branch/?nbranch to support those definitions. 14:44:12 And >r r> for for/next 14:44:23 ?nbranch is 0branch 14:44:26 not ?nbranch 14:44:32 Bleh. Here it is! 14:44:39 0branch is the same as ?branch but with the opposite sense 14:45:12 I made up my own names, heh. But I happend to re-invent "param" which I later saw existed in isForth. 14:45:28 With the same function even. First time that happend ;) 14:46:16 same name ? 14:46:27 Yeah, and same function. 14:46:41 Same function but different name happens often ;) 14:46:42 i invented it too but MrReach told me at the time that there already existed a word to do that with a different name but i didnt like the name 14:46:46 so i stuck with param :) 14:46:53 Heh. 14:46:56 ill te you why 14:47:14 ytou have seen my code and when you need a function you "invent" it from memory without knowing it 14:47:25 im posative that alot of isforth was "invented" by me in that way 14:48:27 Heh 14:48:43 I had never worked with those parts of the code. 14:48:52 High-level words with params. 14:49:08 But sure, it might be possible 14:51:17 yup - but all it takes is for your eye to scan over the code 14:51:30 you can be scrolling at full speed and your brain DID register everything on that page 14:53:39 Hah, no explanation is too unlikely for you to claim copyright to my ideas :P 14:56:10 --- quit: skylan (Read error: 113 (No route to host)) 15:13:58 --- join: randolm (wossname@HSE-Sherbrooke-ppp79557.qc.sympatico.ca) joined #forth 15:17:54 --- quit: wossname (Read error: 104 (Connection reset by peer)) 15:18:01 --- nick: randolm -> wossname 15:22:00 --- join: randolm (wossname@64.229.237.190) joined #forth 15:32:24 --- join: arke (~rk@ca-cmrilo-docsis-cmtsj-b-36.vnnyca.adelphia.net) joined #forth 15:36:44 --- quit: wossname (Read error: 110 (Connection timed out)) 15:47:45 --- quit: randolm (":l^__^--") 16:08:44 --- join: kc5tja (~kc5tja@66-91-231-74.san.rr.com) joined #forth 16:08:44 --- mode: ChanServ set +o kc5tja 16:18:48 Hi 16:18:57 Howdy 16:19:08 * kc5tja is doing some design work on FS/Forth, now that I have the chance. 16:19:11 It's late here, 1:20, and school starts 8.30 tomorrow. Damn. 16:19:21 * kc5tja knows that feeling well. 16:19:22 Too much Forth coding, heh. 16:19:59 * kc5tja thinks his target compiler will be dreadfully simplified over its previous incarnation for DOS. 16:20:59 * kc5tja just wrote what I believe to be the core of the colon compiler, which is only three blocks in size: one for the colon compiler's main loop which also includes the code to scan the Forth wordlist, one to support compiler-words in the host's dictionary, and one to properly handle number conversion for compiling literals. 16:22:00 s/target compiler/Linux's FS/Forth target compiler/ 16:24:04 I make extensive use of return stack manipulation -- helps me to effect control-flow without a whole lot of IF statements and the like. 16:24:33 But I'm not sure I like it yet; manipulating the R-stack explicitly in code can be useful, but I'm wondering if what I did was an abuse. 16:25:06 So I'm not going to actually code it up yet. I'm going to sleep on it for a few days, and try designing several different versions so that I can compare them against each other. 16:25:53 y0 16:26:03 But, the problem with IF statements is they take up space in the source code, for not only the keywords but also the code to place sentinels on the stack. And the existance of those sentinels complicates stack management somewhat. 16:26:28 kc5tja: hows the stack like inside an if block, anyway? 16:26:33 So, exactly which is "simpler" is a matter of perspective, I guess. But I'm going to try both ways, then decide on a case-by-case basis. 16:27:19 arke: Like it is anywhere else; IF consumes its argument in my version of Forth, leaving whatever was beneath accessible. 16:27:38 aah 16:27:40 so 16:27:53 ( blah blah condition -- blah blah ) 16:27:57 Yes 16:28:43 aah fun :) 16:29:08 But in my code, I have constructs like this (note the declarative nature of it): 16:29:40 : forth? word literal ; 16:29:58 : word defined $e8 c, a, r> drop ; 16:30:20 : defined last -found? r> drop ; 16:30:37 : -found? begin dup table1 < if exit then check 4- again ; 16:30:44 : check length name ; 16:31:02 : length dup h@ here @ <> if r> drop exit then ; 16:31:05 and so on. 16:31:36 you use the return stack lots 16:31:58 The R> DROP statements effect the conditional control-flow I need to write what I want to say: : word defined $e8 c, a, r> drop means "a word is defined as $E8 followed by its address. Skip literal number interpretation." 16:32:20 No, actually, I don't. But I'm beginning to suspect I am not using it enough in my general-purpose Forth coding. 16:32:43 I'm using it a lot here only because the control flow would be damn complicated if I used regular IF statements and sentinels. 16:33:06 r> = psp to rsp, right? 16:33:30 Pop from R-stack -- so it moves the top of R-stack to D-stack. 16:33:41 aah 16:33:47 :) 16:33:57 : rdrop r> drop ; 16:33:58 :) 16:34:02 Heh. 16:34:21 You'd want to make that a macro instead ;) 16:34:39 : rdrop compile r> compile drop ; immediate 16:34:41 Something like that. 16:34:45 arke: Not quite, but the same idea. 16:34:47 Otherwise things will go wrong. 16:34:56 Anyway, I have to go to bed. 16:35:02 5 hours left to sleep... :) 16:35:09 Robert: Get some rest. 16:35:13 whats the compile thing? 16:35:14 I will, night. 16:35:17 * kc5tja knows what it's like to get 3 hours sleep before class. 16:35:19 It's not fun. 16:35:22 laters. :) 16:35:26 night 16:35:38 yeah, whats compile and immediate? 16:36:15 arke: Anyway, manipulating the R-stack as a first-class object like that allows me to produce some moderately English/declarative-style source code, and the core of the compiler loop plus the Forth dictionary search code all fits on 14 lines -- easily within a single source block. 16:37:23 arke: compile is a word that takes the next word in the input stream, and causes it to be compiled into the current definition instead of executed, even if the word is immediate (which R> always is, and DROP sometimes is). 16:41:04 arke: A word marked immediate has a flag set that tells the compiler to, unless otherwise directed (with POSTPONE or COMPILE, as shown above), execute the word rather than compile it in place. 16:42:16 IF, ELSE, and THEN are immediate words, for example -- instead of compiling a call to IF (which doesn't make much sense), the code for IF is executed "immediately," which usually causes the appropriate machine language instructions for condition testing and conditional branching to be emitted into the generated machine language. 16:43:57 Well, I gotta cruise. Aikido time, and I have to take a shower first. 16:44:04 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 16:53:47 --- join: TheBlueWizard (TheBlueWiz@pc1bdn1d.ppp.FCC.NET) joined #forth 16:53:48 --- mode: ChanServ set +o TheBlueWizard 16:53:58 hiya all 17:22:31 y0 17:24:14 why would : rdrop compile r> compile drop ; immediate be better than : rdrop r> drop ; 17:25:42 arke: cuz when the rdrop is *called*, the return address is pushed on the R stack, and thus perform the undesired behavior 17:30:07 --- join: I440r (~I440r@sdn-ap-010txhousP0477.dialsprint.net) joined #forth 17:31:13 hiya I440r!!!! :))) 17:32:22 hi :) 17:34:29 http://www.tripointglobal.com/Controls/default.htm 17:34:43 will look at it in a bit 17:35:08 :) 17:41:44 I440r: y0 17:42:09 hi 17:42:12 TheBlueWizard: oh, that makes sense. thanks! 17:42:53 I440r: how does isforth handle compiled and immediate code? 17:44:09 find returns a flag 17:44:19 and an tx 17:44:21 erm xt 17:44:26 hm? 17:44:31 the flag is -1 for non immediate words and 1 for immediate words 17:44:45 interpret will either compile or execute based on the current state and that flag 17:44:49 it xor's the flag with state 17:45:02 and the result is either zero or non zerp 17:45:11 meaning execute or compile 17:45:32 erm 17:45:45 i meant, how does it compile/execute, and how does it interpret? 17:46:44 an interpret is an execute 17:47:04 erm 17:47:05 i meant 17:47:08 find returns an execution token and we jump to that address if were in interpret mode or if the word is immediate 17:47:50 and calls otherwise? 17:48:06 comma 17:48:45 ? 17:49:12 , <-- 17:49:21 hem??? 17:49:22 if were in compile mode we COMMA the xt 17:49:29 store it? 17:49:49 no 17:49:51 NOT ! 17:49:53 , 17:49:55 comma 17:50:05 erm 17:50:15 make a new place in memry for it, and then store it? 17:50:45 : , here 1 cells allot ! ; 17:50:48 or something 17:51:03 what do all of those do? :P 17:51:20 allot = ( num -- address ), i know that 17:51:29 cells = 4 17:51:37 dont you know what here is ? 17:51:39 (in 32 bit :)) 17:51:41 nope. 17:51:55 i never use cells but everyone else does these days 17:52:01 its not even in isforth 17:52:27 portability :) 17:52:29 what does here do? 17:52:54 erm - here is like $ in assembler 17:53:04 its teh next address were going to compile into 17:53:28 so 17:53:57 blah $ blah $ would give me, lets say, 0x1111 and 0x1119? 17:54:03 or 17:54:03 no 17:54:04 heh 17:54:11 erm 17:54:37 bah, i dont get it :P 17:56:05 you dont understand fopr sure heh 17:56:09 do you know what 17:56:12 jmp $ 17:56:15 does in asm ? 17:56:19 infinite loop 17:56:27 what does $ mean tho 17:56:36 its the address currently being compiled into 17:56:38 $ is the address of the current line 17:56:42 thats what 'here' is in forth 17:56:47 but 17:56:52 how does that apply here? 17:56:57 as in 17:57:04 hm ... wait 17:57:07 , <-- comma IS forths compiler 17:57:25 it stores the top of stack at here and allots the space 17:57:46 ooh 17:57:48 hrm 17:57:48 so 17:57:57