00:00:00 --- log: started forth/04.05.26 00:03:52 hi thin 00:16:32 goodnight everyone! 00:16:38 --- quit: crc ("ChatZilla 0.9.61 [Mozilla rv:1.7b/20040316]") 00:31:49 --- quit: Herkamire ("past my bedtime") 01:50:17 --- join: Serg (~z@212.34.52.140) joined #forth 01:58:50 --- quit: Robert (Remote closed the connection) 02:00:26 --- join: Robert (~snofs@c-bf5a71d5.17-1-64736c10.cust.bredbandsbolaget.se) joined #forth 04:53:32 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 04:53:32 --- quit: Serg (Read error: 54 (Connection reset by peer)) 05:32:19 --- join: Serg (~z@212.34.52.140) joined #forth 05:47:59 --- quit: Serg () 05:49:56 --- quit: I440r ("Leaving") 06:09:00 --- join: madwork_ (~madgarden@derby.metrics.com) joined #forth 06:48:17 hmm, another project that uses darcs 06:57:44 --- quit: warpzero (Read error: 60 (Operation timed out)) 07:05:08 --- quit: arke (Read error: 104 (Connection reset by peer)) 07:05:17 --- join: arke_ (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 07:08:12 --- join: warpzero (~warpzero@199.104.115.142) joined #forth 07:32:55 --- quit: arke_ (Read error: 110 (Connection timed out)) 07:34:39 --- quit: warpzero (Read error: 60 (Operation timed out)) 07:43:06 --- join: arke (~Chris@wbar8.lax1-4-11-100-108.dsl-verizon.net) joined #forth 07:47:37 --- join: warpzero (~warpzero@dsl.142.mt.onewest.net) joined #forth 09:07:04 --- join: ASau (~asau@158.250.48.204) joined #forth 09:07:16 Guten Abend! 09:09:19 Privet, ASau. 09:17:38 7w 5 09:17:44 ? 09:18:28 secret talk! 09:18:37 1 09:36:49 given the channel's high number of INTPs, there's no need to greet each other at all. just launch straight into discussion 09:37:13 if you greet, if you say "hi " then that means you have nothing to say 09:37:20 and if you have nothing to say, why are you here? 09:37:53 To mark you're not absent. 09:38:07 To state... 09:38:27 assume no one is absent 09:38:39 or just start talking right away 09:39:55 IRC is not USENET. 09:41:55 If it is not real-time, than what is its use? 09:43:54 killing the time 09:49:08 And one another thing. 09:49:56 "High number" does not mean that everyone here is INTP. 09:51:58 i know, but this channel is all about personalitism! 09:52:13 lalal 09:53:05 INTP? 09:53:17 introvert somethign someething something? 09:53:30 some personality test result? 09:54:28 i recall i was INTJ but might be wrong too 09:56:14 http://209.15.29.56/myersbriggs/personhome.htm 09:56:15 whee 09:56:16 Hi futhin 09:56:24 only 1% of INTJ, INTP of all 09:57:43 http://www.humanmetrics.com/cgi-win/JTypes2.asp 09:58:11 duh 09:58:17 test. NO, no, nononononono 09:58:23 duh, easy to fill in ;) 10:08:05 still INTJ 10:46:01 --- join: proteusguy (~proteusgu@69.79.24.31) joined #forth 10:48:44 i forgot what i was 10:48:48 something like that too 10:51:13 http://herkamire.com 10:51:35 INFP: qFox 10:51:52 (From my notes.) 10:52:45 oh yeah 11:19:59 i guess INTP is like debugger 11:20:03 and INTJ is like lazy programmer 11:20:10 debugger as person 11:35:58 hi all 11:36:29 i think im ENTP but i don't recall 11:36:55 i think that means i'm geeky *and* an arsehole 11:38:23 That's too common. :) 11:48:35 --- join: wossname (wossname@Toronto-HSE-ppp3699046.sympatico.ca) joined #forth 11:48:50 ENTP: slava 11:48:58 Everything is recorded. 11:49:38 "Big Brother watches you." 11:51:40 :D 11:52:00 asau must beINTJ and build diagram on how to organise the channel work more efectively 11:52:06 Don't forget about lofs. 11:52:09 ...logs. 11:52:14 ASau, what are you? 11:52:33 INTJ: ASau, crc, tathi 11:52:48 Sorry, outdated. 11:52:49 INTJ: ASau, crc, mur, tathi, yeoh 11:53:12 add me to INTP ~ 11:53:43 OK. 11:53:44 Done. 11:53:47 oh! 11:53:49 that explains 11:53:57 what'd i say :D 11:54:20 wossname, you'll have 5 people to start asking you to check their ideas and plans ;) 11:54:56 ;( 11:55:23 --- quit: skylan (Read error: 113 (No route to host)) 11:56:37 I like http://typelogic.com/intj.html assessed profiles. 11:58:04 Jane Austen and J.F.K. for to be fan of. 12:02:11 Jefferson, Gandalf, Darcy. (I don't know that they're on that particular list.) 12:02:35 And everyone knows God is an INTJ. 12:06:04 You're not on that list, it seems. 12:06:28 You're not so critical. 12:06:34 fffffft 12:06:37 ah 12:06:44 i dind't like any of them really at all 12:07:02 but i liked the text 12:07:03 :) 12:07:15 Esp. prof. Moriarty? 12:07:47 oh i was reading this one: http://209.15.29.56/myersbriggs/intj.htm 12:08:13 but it seems tha they are more and less each others plagiars 12:09:41 Yes, common basis is obvious. 12:09:47 uh 12:09:52 it's clear plagiarism! 12:10:04 word plagiarism, mosaic plagiarism and even 12:10:09 expressions are partly plagiarised 12:10:14 i dont know where from 12:10:16 but they are 12:10:18 Can you say who was the original author? 12:10:32 probably someone else than these 2 12:10:43 May be. 12:10:51 i guess they're from the orginal author's research 12:11:04 just not mentioned. not credited 12:11:04 But it's not necessarily so. 12:11:08 well 12:11:14 they have obviously same source 12:12:05 hmm.. for some reason i trust more the 209.IP than the other one 12:12:09 even the other one has names 12:12:23 the 209.IP has more text and contents 12:12:26 more deeply 12:12:33 but they still have both same text 12:12:56 either the telelogic is plagiariesd from 209.IP or they both are from soemwhere else 12:13:32 I take a minute for reading. 12:14:55 takes 10 12:15:55 Well. Easily you can see, that the first part of typelogic.com's is stripped down 209.*'s version. 12:16:05 Lists are identical. 12:16:56 Typelogic's second part, "Func. analysis," is less expressive. 12:18:26 I don't believe Heiss, but I can't say anything about Butt. 12:49:46 --- join: crc (~Charles_C@0-1pool88-119.nas48.philadelphia1.pa.us.da.qwest.net) joined #forth 13:01:51 Dobryjj vecher, crc. 13:02:04 Hello ASau 13:02:27 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 13:03:17 I'm going to extend the load/save words to handle actual files of various sizes 13:04:02 Something like " filename" size [load or save] 13:04:03 Dobryjj vecher, tathi. 13:04:37 Can you read forward? 13:04:46 ??? 13:04:57 It's easier to write LOAD file-name then " file-name" LOAD 13:05:10 I could do that 13:05:34 This avoids string processing in general. 13:05:35 Since load/save are primitives, it's easy to do forward parsing :-) 13:06:49 Hi ASau. 13:07:09 It seems to me that string handling is not a necessary in most tasks. 13:09:16 In fact I don't have literal strings of that " STRING" kind at all. 13:10:00 I have only words that parse ahead. 13:11:10 I don't use them as much now that RetroForth has supports s" and ." within words 13:11:19 but I think they can be useful at times 13:13:31 May be. 13:13:50 Though I haven't seen a proper situation yet. 13:14:50 Ok, load/save now take the filename by reading ahead 13:16:58 Should the length of the file be specified before or after word? 13:17:23 Why do you need it? 13:17:46 There's no easy way to determine the exact length of a file 13:17:51 ??? 13:17:54 man seek 13:17:58 Hm. 13:18:01 man lseek 13:18:05 Doesn't work under Windows 13:19:52 DOS API 42H 13:20:08 Should be somehow possible under WinAPI. 13:21:16 not in win32 mode 13:22:17 The documentation from microsoft says to read in the file in small pieces (4k IIRC) and check each for an EOF or a read error 13:22:30 But that gets complex, and doesn't always work properly 13:23:00 It's very weird. 13:23:06 Yes it is 13:23:27 * crc curses Microsoft for their stupidity in developing the Win32 API 13:23:32 That doesn't seem so complex to me... 13:24:04 It is when you're at the assembly level 13:24:05 It's more complex than classic lseek way. 13:24:27 I have the words set up to do: 13:24:40 LENGTH load/save FILENAME 13:24:44 at the moment 13:25:00 Maybe this can be done in Forth for Windows? 13:25:48 Especially for stupid Windows. 13:25:49 oh, I'm probably misunderstanding. What do LOAD/SAVE do? 13:26:08 load = load the contents of a file to a memory buffer 13:26:11 Load and save file in/from RAM. 13:26:18 save = save the contents of that buffer back to hard disk 13:26:50 Well. Tathi is rght. 13:26:50 I need a way to specify the buffer address too 13:26:55 oh, ok. 13:26:57 LOAD is reserved for Forth LOAD. 13:27:23 RetroForth is *not* a standard Forth by any means 13:27:24 This confuses. 13:27:29 Yeah. LOAD often means include. 13:27:42 crc. I know, but this confuses. 13:27:51 Nah, that's fine. 13:28:02 --- quit: proteusguy ("Leaving") 13:30:00 damn. With the new syntax, load/save don't work within words anymore 13:30:40 Hm. 13:30:42 if RetroForth had a word to include files, I'd call it INCLUDE 13:30:43 Of course. 13:31:11 But what do you need this for? 13:31:28 I mean load/save inside another words. 13:32:07 I want to be able to write words to copy files, etc 13:32:33 e.g. RetroForth will eventually replace BASH and Windows explorer as my computer's user interface 13:32:39 : COPY LOAD SAVE ; ( why not?) 13:33:16 Parses two file names ahead. 13:33:23 Not in RetroForth 13:33:30 The compiler chokes on that 13:34:18 I'd have to rewrite the compiler and interpreter, which would mean breaking a lot of other words 13:36:05 If I could work out a clean way to determine the length of a file, it might work 13:38:07 Should I rename "word" to "parse"? 13:39:07 I use "word" not "parse." 13:39:24 But "parse" seems to be a more logical choice 13:39:45 The purpose is to parse a string until the next occurance of a given character 13:40:01 * crc wonders why "word" is used for this 13:40:15 * tathi too 13:40:31 I call it parse. 13:40:42 and do : word bl parse ; 13:41:14 :-) 13:41:27 --- join: tgunr (~davec@vsat-148-63-4-106.c001.g4.mrt.starband.net) joined #forth 13:41:36 It's functional style: "word" results in word parsed. 13:42:18 what do you guys think of dynamic -vs- lexical scope? 13:42:56 Yeah, but I usually think of a word as being delimited by whitespace, and WORD doesn't necessarily give you that. 13:44:08 slava: in Forth, or in general? 13:45:43 tathi, 16 bit "word" is not a word too. 13:46:04 ASau: ? 13:46:30 And algorithmists also have their own view on what is word. 13:47:02 ASau: I know, I was just saying that _I_, personally, don't like the name "word" for that functionality. 13:47:24 For me, "parse" seems more descriptive. 13:47:24 Sorry, nothing personal. 13:47:51 * crc is going to rename "word" to "parse" in RetroForth 6.31 13:50:02 parseword 13:50:12 ???? 13:50:43 well, the word parses a "x"-delimited word from the input buffer, dont it? 13:51:00 It is not just for finding words 13:51:27 oh 13:52:02 Yeah, it's kind of how you do anything with the text input stream in forth. 13:53:10 slava: actually, I don't think I have enough experience with dynamically scoped languages to have a good opinon on the issue anyway... 13:53:14 I-do-not-like-long-names-of-this-kind. 13:53:48 i'm fine with them as long as they are not intercaps case sensetive 13:54:09 just use short, lowercase word names :-) 13:54:41 well i'm fine with that, but WORD is a bit ambigious 13:54:49 parse... well i gues 13:54:50 s 13:55:07 qFox, you should try to work without word completion. 13:57:04 ASau, why? 13:57:06 word completion? 13:57:45 Just try it to feel how good it works. 13:58:23 uhm, what do you mean with word completion? :\ 13:58:55 like tabbing to finish a word while typing? 13:59:34 Yes. 13:59:48 ah right, yeah thats a great feature :) 14:00:09 It's a complex feature 14:01:00 crc, if reading file piece by piece and waiting for end is complex, word completion is not complex. 14:01:05 hmmmm, not neccesarily (euh spelling?) 14:01:35 word completetion is more complex than reading a file in piece by piece 14:01:41 A lot more complex 14:02:47 hm, read the last word of the inputbuffer, look it up alphabeticly, remember what word is last found if any, take the next word. i wouldnt say its complicated, just alot of searching (unless the dictionary has a alphabetical index...) 14:02:53 an. 14:03:41 --- quit: wossname () 14:03:47 It requires tracking words closely, and tight bindings to the editing environment 14:04:21 hmm, still dont think it has to be complicated 14:04:31 --- join: wossname (wossname@Toronto-HSE-ppp3699046.sympatico.ca) joined #forth 14:07:07 it's probably not bad in forth. 14:07:26 not so fun in asm though. 14:07:31 :) 14:10:05 I'll have word completion in my visual forth when I actually get around to writing it 14:10:35 wow - 2x speed improvement in inner interpreter 14:11:09 slava: good 14:23:24 --- join: lalalim_ (~lalalim@p508AAF1C.dip.t-dialin.net) joined #forth 14:35:39 * crc is leaving. I'll be back tomorrow... 14:35:42 --- quit: crc ("ChatZilla 0.9.61 [Mozilla rv:1.7b/20040316]") 14:46:52 --- quit: lalalim (Read error: 110 (Connection timed out)) 14:56:58 --- join: Herkamire (stjohns@h000094d30ba2.ne.client2.attbi.com) joined #forth 14:57:22 Dobryjj vecher, Herkamire! 14:58:54 hi ASau 15:05:31 --- quit: tgunr (Excess Flood) 15:07:30 --- join: tgunr (~davec@vsat-148-63-4-106.c001.g4.mrt.starband.net) joined #forth 15:31:53 --- join: SolarFire[ (SolarFire@pD9EE1C0D.dip.t-dialin.net) joined #forth 15:33:41 --- join: DrMartens (~DrMartens@81-86-121-165.dsl.pipex.com) joined #forth 15:35:01 Good morning, DrMartens! 15:41:22 hi 15:44:49 Dobryjj vecher. 15:45:44 guten tag\ 15:45:52 Guten Abend, arke. 15:46:39 Wie geht's, Robert? 15:46:41 --- quit: DrMartens ("Leaving") 15:50:16 --- quit: Solarfire (Read error: 110 (Connection timed out)) 16:07:26 Hm. 16:07:40 Link for Linuxoid Forthers: http://www.muppetlabs.com/~breadbox/software/tiny/ 16:17:44 tathi: hi 16:18:04 hi arke 16:26:23 arke: Gut, danke.. Mit dir, auch gut? 16:26:43 Robert: jaah, eigentlich schon. 16:27:07 Ich glaube ich werde spaeter ins linux booten und dann dort am OS anfangen. 16:27:22 Ist halt scheisse dass ich dann nicht online gehen kann. 16:27:45 --- join: Sonarman (~matt@adsl-64-160-165-237.dsl.snfc21.pacbell.net) joined #forth 16:28:31 Hehe... the Forth OS? 16:29:46 yeah 16:30:35 hrm 16:30:40 maybe I can convince cygwin 16:34:51 --- quit: wossname ("ideas") 16:39:29 --- quit: tgunr ("Quit") 16:41:22 Robert: Would you like to help, not including your indirect contribution with F? 16:41:44 With? 16:42:33 my OS :) 16:42:41 Oh. What will it be like? 16:42:45 * arke might actually begin with a mere reqrite of F 16:43:01 Robert: basically, ColorForth with less colors and much, much, MUCH more crack 16:43:15 and the option of punctuation 16:43:27 so that you can even use it interchangeably with color 16:44:24 crack? 16:45:44 Colombian stuff, slava. 16:45:56 I'm not sure... 16:46:05 i'll stick with beer, or gin&juice :) 16:46:40 * Robert prefers his water without cocaine, alcohol or.. well.. orange juice is OK. 16:48:22 :) 16:48:25 Robert: would you like to? 16:48:38 I'm gonna try cygwin soon, and then I'll tell you my success 16:48:47 Robert: btw, when is F due? 16:49:30 Was finished a number of months ago. 16:50:12 awesome. 16:50:18 Do you plan on continuing it? 16:50:22 Not really.. could have been better. 16:50:29 heh 16:50:37 Maybe if I get bored enough. x86 OS design feels so pointless. 16:50:41 may I continue, with your hosting and help? :P 16:50:47 But then I look at the current systems. 16:50:51 And I see the point again. 16:51:05 And then I look at the hardware of modern PCs, and the point is lost. 16:57:01 lol 16:57:09 hardware is a problem, definetely. 16:57:38 A problem? 16:57:40 Robert: how much of F can be rewritten in Forth, if any? 16:57:44 There's too much of it. :) 16:57:50 Yes. Current hardware sucks. No exceptions. 16:57:59 A simple Forth system on a computer with 1M times the resources I need. 16:58:05 Robert: EXACTLY. Either there are no standards, or the standards are too restrictive 16:58:17 You misunderstand me. :D 16:58:28 But of course, that's another problem. 16:58:32 Yes. 16:58:35 :) 16:58:54 Robert: answer my other question though -- how much of F can be rewritten in F from asm? 16:59:10 Oh, didn't see it. 16:59:38 --- join: skylan (~sjh@vickesh01-4582.tbaytel.net) joined #forth 16:59:44 Hmm... Not really sure, probably quite a bit of it. 16:59:48 --- join: skylan_ (~sjh@vickesh01-4582.tbaytel.net) joined #forth 16:59:48 --- join: tgunr (~davec@vsat-148-63-4-106.c001.g4.mrt.starband.net) joined #forth 17:00:30 --- quit: skylan_ (Client Quit) 17:02:40 --- quit: Callidus ("Abero") 17:03:01 DTC or STC? 17:03:10 oh yeah 17:03:10 DTC 17:03:11 :( 17:03:29 THATS why I wanted to rewrite F, now I remember 17:03:35 oh well 17:03:39 it doesn't need a rewrite 17:03:50 I'm sure the compiler is simple to adapt to STC 17:03:56 Hehe. 17:04:13 You'll ONLY have to rewrite a few kB of library routines and so on... 17:04:28 I'd rewrite it from scratch if I were you. 17:06:23 :( 17:06:34 Of course you can steal some parts 17:06:46 Will you help? 17:06:49 Like I did from IsForth (did I mention that in the F credits? Do I have F credits?) 17:07:08 :) 17:08:34 Hmm.. 17:08:42 You know, I feel kind of tempted to write the compiler. 17:08:49 What language will you use for it? 17:08:50 awesome. 17:09:00 asm+forth 17:09:06 no exceptions :) 17:10:02 What Forth? 17:10:12 Hmm.. 17:10:25 er 17:10:30 well 17:10:30 If I write the original compiler in asm, someone else could port it to Forth in case you want a metacompiler. 17:10:38 kinda self-made 17:10:39 :) 17:10:50 Robert: well, I actually want the least amount possible in asm 17:11:02 :) 17:11:05 but that would work. 17:11:30 what is DTC and what is STC 17:11:49 warpzero: DTC means Direct Threaded Code, STC means Subroutine Threaded Code 17:11:51 warpzero: http://www.zetetics.com/bj/papers/moving1.htm 17:12:03 warpzero: I suggest you read "Moving Forth" 17:12:11 Sonarman: agh fuck you 17:12:12 Sonarman: :P 17:12:15 oh that 17:12:16 sorry 17:12:23 i just didn't recognize the acronyms 17:12:24 arke hehe 17:12:26 oh 17:12:27 lol 17:12:30 Sonarman: hate you :) 17:15:36 arke: OK, the connection is really unstable over here. But sooner or later I might write a 32-bit x86 Forth compiler in asm. If I ever make it, I'll help to make you port it to your system.... 17:16:26 Robert: cool. I'll send you an email in a moment, ok? 17:17:00 Sure. 17:17:11 We'll see when home.se feels like delivering it. 17:17:34 ok 17:17:39 home.se? 17:17:48 i thought your email was somewhere else.. 17:19:07 oh 17:19:08 nevermind 17:19:08 lol 17:19:13 Ok, check in a few minutes 17:19:41 Heh, I got a mail, but it was just some TI spam. 17:20:15 --- quit: tathi ("leaving") 17:21:24 :) 17:23:05 Nothing yet. 17:23:22 haven't snet it yet :P 17:39:32 Oh. 17:41:29 ok, sent 17:41:40 Got it 17:42:08 Yay. 17:42:43 --- join: blockhead (default@dialin-248-tnt.nyc.bestweb.net) joined #forth 17:42:50 HERE in EBP doesn't sound like an ideal solution. 17:45:11 I was thinking that if I (we? :)) don't do a dual-register stack, then EDI will have HERE 17:45:55 which makes sense --- , will be a stosd 17:50:25 I was thinking about a multi-register stack cache. 17:50:34 would ESI be used if , were stosd? 17:50:41 Sonarman: no 17:50:51 So at least the top 3 elements or so could be cached in registers. 17:50:52 Robert: bleh, isn't that a little hard to implement in x86? 17:51:04 That way you could optimize quite a lot, I think. 17:51:05 Robert: since you constantly have to exchange? 17:51:07 Everything is. 17:51:15 Robert: lol 17:51:22 arke: oops nevermind sorry :) 17:51:26 Sonarman: man i like that new song lol 17:51:29 Well, I implemented an experimental compiler for the PIC, using that design. 17:51:36 Actually, just a piece of it. 17:51:36 arke: the Dimmu Borgir? 17:51:43 And it's not all that complicated. 17:51:43 Sonarman: rotting christ 17:52:03 Robert: well, is the top of stack always in a certain register, or does it move around? 17:52:10 arke: yeah, it's great! :) 17:52:16 Sonarman: :) 17:52:35 arke: Move around. 17:52:51 Robert: how do you know, then, where the current TOS is? 17:52:54 arke: So "swap" and "drop" would generate no machine code. 17:53:00 Robert: ooh 17:53:02 Robert: OOH 17:53:04 Robert: yeah 17:53:04 You store that in a table during compile-time. 17:53:08 Robert: I remember you telling me about this 17:53:13 Robert: nevermind :) 17:53:18 And dup would be one single mov. 17:53:23 As would over. 17:53:30 "+" would be a single add 17:53:36 Between registers. 17:53:42 And so on.. 17:53:50 Problem comes with branches and stuff. 17:54:34 yeah, because you have to restore the registers to a special position 17:55:11 Right, the function that's called can't know where your current TOS is. 17:55:23 So you'll have to move it to its TOS register. 17:55:34 and all the others too 17:55:43 thats probably the most expensive part of the whole thing 17:55:57 doesn't sound any more expensive than using a stack frame 17:56:28