00:00:00 --- log: started forth/03.06.10 00:00:35 --- quit: fridge (asimov.freenode.net irc.freenode.net) 00:00:35 --- quit: kc5tja (asimov.freenode.net irc.freenode.net) 00:00:35 --- quit: XeF4_ (asimov.freenode.net irc.freenode.net) 00:00:36 --- quit: Serg_Penguin (asimov.freenode.net irc.freenode.net) 00:00:36 --- quit: a7r (asimov.freenode.net irc.freenode.net) 00:00:36 --- quit: onetom (asimov.freenode.net irc.freenode.net) 00:00:37 --- quit: paxl (asimov.freenode.net irc.freenode.net) 00:00:37 --- quit: cleverdra (asimov.freenode.net irc.freenode.net) 00:00:37 --- quit: ianni (asimov.freenode.net irc.freenode.net) 00:00:38 --- quit: ChanServ (asimov.freenode.net irc.freenode.net) 00:00:38 --- quit: TreyB (asimov.freenode.net irc.freenode.net) 00:01:04 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 00:01:04 --- join: Serg_Penguin (Serg_Pengu@212.34.52.142) joined #forth 00:01:04 --- join: fridge (~matt@dsl-203-33-160-107.NSW.netspace.net.au) joined #forth 00:01:04 --- join: TreyB (~Trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 00:01:04 --- join: cleverdra (nailuj@ACC6186F.ipt.aol.com) joined #forth 00:01:04 --- join: paxl (paxl@modemcable110.168-130-66.que.mc.videotron.ca) joined #forth 00:01:04 --- join: ianni (ian@inpuj.net) joined #forth 00:01:04 --- join: ChanServ (ChanServ@services.) joined #forth 00:01:04 --- join: XeF4_ (Vatteluttu@12-245-108-157.client.attbi.com) joined #forth 00:01:04 --- join: onetom (~tom@novtan.bio.u-szeged.hu) joined #forth 00:01:04 --- mode: asimov.freenode.net set +oo kc5tja ChanServ 00:01:08 Bah 00:01:13 Damn splits 00:01:27 * kc5tja nods 00:01:29 Well, OK, I see what you mean I suppose 00:01:29 Hardly a signifigant change, though 00:01:29 Surely you're not anti ANS because of one simple word that doesn't fit into your forth's design at the moment... ? 00:01:40 No. 00:01:50 I'm actually not all that anti-ANS. 00:02:13 But having tried to implement an ANS-Forth before, I felt that the result didn't justify the amount of investment I put into it. 00:02:14 Well that's a pleasant change, actually. :) 00:02:22 So now I'm writing what I want, when I want it. 00:03:19 I don't think writing a forth system designed for other people's use should be an easy task. 00:03:22 And concentrating on making the system as a whole as small as practicable, as safe as practicable (it won't crash nearly as often as ColorForth, for example!) while not being overly burdensome, and as flexible as practicable. 00:03:28 Well... It shouldn't be a trivial task, in any case 00:03:50 Er, what I mean to say is... 00:04:03 I think interoperability and portability is worth putting effort into 00:04:18 OK, I have to laugh when people say this. 00:04:35 As I indicated, so far, I've had very, very little difficulty getting ANSI Forth code to work under cmForth/PygmyForth. 00:04:40 Yet, they're very different environments. 00:05:00 I spend, maybe, a couple of hours writing a shimmy between the ANS program and what PygmyForth defines. 00:05:30 I quite often find I have to do exactly the same thing with a so-called "portable" language like C. 00:05:39 Well pygmyforth is relatively close to the ANS standard (It's F83 compliant, no?) 00:05:47 When porting C source from, say, AmigaOS to Linux, I have to go through, and fix up all the compiler errors, etc. 00:05:53 No 00:05:55 It's cmForth 00:06:01 It's actually more closely related to ColorForth. 00:06:11 (though distantly) 00:06:14 C is not a 'portable language' 00:06:20 Neither is Forth. 00:06:26 neither is java 00:06:42 And combined with an overwhelming variety of non-portable libraries, C can be very difficult to port 00:06:44 This is why I don't bother with standards compliance. There's really no point to it. 00:07:22 What's really needed, more than anything else, is an unconditional understanding of the program being ported. 00:07:42 What do all these words actually DO? Don't just tell me HOW they do it. 00:07:51 If necessary, even tell me WHY. 00:09:19 This is one of the biggest complaints I have against the comp.lang.forth crowd. 00:09:36 People get all happy about their latest bit of slick program code, and they post it to the newsgroup for all to gawk at. 00:09:49 ALL fo these code snips suffer from several critical problems: 00:10:05 1. They usually have very hard to decipher word names, with usually inadequate comments explaining what they do. 00:10:24 2. Each word is written as if they were programming in C: long word definitions with deeply nested control-flow structures. 00:10:40 Well I feel that the only reason why portability is feasible *at all* is because of at least rudimentary standards in place either formally or informally. POSIX is a good example of a formalized standard that works. Porting between POSIX environments is not a problem. The BSD socket interface was an example of an informal standard that worked: while there were machine differences etc on different machines, all socket AP 00:11:22 The FORTH Inc implementation and the subsequent FIG implementations were good example of standards that work 00:11:38 Your first paragraph got chopped off: ". . ., all socket AP" 00:12:17 ANS forth is, in my opinion, the only practical way of making sure that forth as we know it doesn't severly fragment and become completely unusable... 00:12:22 Er 00:12:26 all socket APIs acted in fundamentally the same manner. 00:13:18 I don't agree. 00:13:18 There's little evidence to support it. 00:13:33 Sure there is. We have a perfect example available for us: LISP 00:14:04 Forth, Inc's and FIG's standards evolved to become quite popular because both disclosed how their systems worked. 00:14:20 ANS is just another standard -- one of so very many to support. 00:14:32 What about Lisp? 00:14:43 The fact that it's covered by an 800+ page standards document? 00:15:00 There are still tons of "underground" Lisp systems -- Emacs Lisp is the most popular. 00:15:10 People still port code back and forth all the time. 00:15:42 No, the fact that LISP has been impossible for serious industry/academic use until recently. 00:15:44 And then there's Scheme. 00:16:29 Languages are what people make them out to be. 00:17:05 LISP, a technically superior system to C/unix in most respects, was not popularly adopted because of it's fragmented disorganized nature. C was dictated fairly absolutely and it, of course, thrived 00:18:06 Unix as well... Even though there were a large number of different unix implementations, they all followed the same fundamental (informally standard) model 00:18:10 C thrived because it was distributed with a free operating system called Unix, which AT&T donated to some of the world's most highly attended and well-respected universities, **LONG** before ANSI/ISO standardized it. 00:18:36 There were at least as many different variants of C back then as there are of Forth today. 00:19:13 Lisp was a freely distributed language used to implement OSs too and was also popular in academic fields as well, though 00:19:32 Not as a general purpose computing language, though. 00:19:38 There were at least as many different variants of C back then as there are of Forth today. 00:19:39 ^ such as? 00:20:11 VAX C, OS/370's C, C on the PCs (when they came out), MacOS ThinkC, et. al. 00:20:48 Each of these C implementations were all different and did things just differently enough to warrent HUGE quantities of conditional compilation directives just to get fairly trivial applications to compile across different platforms. 00:21:00 kc5tja : Nonsense... LISP was percieved to be more of a general purpose, more flexible language than the (supposedly OS specific) C language... 00:21:37 Fractal: Nonsense. I don't see any shred of evidence to support that claim, but there are tons of documentation that highlight Lisp's superior support for symbolic computing. 00:21:40 They all were programmable according to the 'founding C scriptures though' 00:21:55 And in such diverse environments, things such as those are unavoidable 00:21:57 Fractal: And they were all different enough to create huge incompatibilities. 00:22:25 Just as all Forths are built on the same fundamental principles: two stacks, colon definitions, and an accessible compiler/parser. 00:22:48 kc5tja : Yes, exactly 00:23:50 Which is why I find it alarming that people such as chuck moore are trying to radically change the parsing/compilation syntaxes of forth 00:24:03 Don't get me wrong - I have no problem with language experimentation 00:24:42 I really don't see it as being all that different. 00:25:01 But the way colourforth is being advertised and promoted seems to be going nowhere other than fragmenting an already sparse community 00:25:10 But whether punctuation or colors are used to define words isn't the point of the argument. 00:25:20 well. but the way forth has STATE and fiddles with different versions of "compile,", "[compile]", "," etc is majorly stupid 00:25:46 kc5tja : Well perhaps not so much in the colour aspect, but these forths are actively anti the established standards 00:25:47 Oh, yeah, that's a good point. FS/Forth doesn't have STATE. :) 00:26:07 "actively"? 00:26:28 they include a virus deleting other forths from hd 00:27:02 Actively meaning they go out of their way to not comply with the ANS standard 00:27:29 Well, or any standard other than the whim of the implementor 00:27:50 Chuck doesn't give two shits about ANS Forth. He codes his Forth according to his own standards. Whatever happens to line up with ANS standards, well, that's just frosting on the cake. 00:28:11 But not caring about an existing standard and being "actively anti-ANS" are two radically different concepts. 00:28:23 If I wanted to be radically anti-ANS, I'd swap the meanings of @ and !. 00:28:25 But CM *is* actively anti-ANS 00:28:41 His opinion is that the ANS standard is too big. 00:28:49 That the CORE wordset contains words it shouldn't. 00:29:21 And I have to agree with him, to a limited extent. 00:29:27 But he also raises another issue: 00:29:32 we need a publication standard. 00:29:54 We need some standard means of communicating with each other regarding our code, because right now, simply exchanging source is wholesale inadequate. 00:30:18 Remember that he WAS originally on the first ANS committee. 00:30:35 He left after his input was soundly rejected almost unilaterally. 00:30:51 I think he has a right to be a bit upset. 00:30:53 I know I would be. 00:33:42 At any rate, FS/Forth is a classical Forth (I do have plans on experimenting with my own ColorForth in the future, but it'll go by another name). It is a cmForth and PygmyForth derivative. It is built with a minimalist philosophy. 00:34:15 Well I've had this argument too many times. You do put the argument more elequently than most (and you're american you say? ;) ), but my opinion still holds. I feel the fact that forth is still being actively pursued and developed today is due to ANS, and I feel that CM's efforts are dangerously close to reducing forth to "novelty status" 00:34:16 It is built with the assumption that portability IS a myth, and that portability can only be achieved through rigorous factoring and programmer intellect. 00:34:30 But, I've got to get going... 00:34:53 I'm not convinced of that argument. 00:35:04 So we'll just have to agree to disagree on that. 00:35:15 Heh, sounds good to me. :) 00:35:23 Even so, nothing Chuck is doing is stopping anyone from developing new classical forths -- like me, for example. 00:35:28 Yes, I'm American. :) 00:35:39 Eerie, some of us actually know how to communicate. 00:35:48 Though, we're a shrinking minority. :( 00:35:49 Heh 00:35:55 Yeah... 00:36:02 Well anyways thanks for the conversation. 00:36:06 'Night 00:36:08 Some days, though, I get too easily upset. 00:36:13 And that interferes with my writing. 00:36:14 Oh well. 00:36:16 OK, goodnight. :) 00:36:20 :) 00:45:47 OK, I need to hit the sack myself. 00:45:56 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 01:39:48 --- quit: Serg_Penguin () 03:49:29 --- join: Serg_Penguin (Serg_Pengu@212.34.52.142) joined #forth 03:49:45 --- quit: Serg_Penguin (Client Quit) 04:33:15 --- join: mur (murr@baana-62-165-185-171.phnet.fi) joined #forth 05:04:50 Morning all :) 05:05:49 moi 05:16:25 --- join: Serg_Penguin (Serg_Pengu@212.34.52.142) joined #forth 05:16:44 hi 05:17:25 privet! 05:18:42 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 05:25:31 mur: whadda u think, what is minimal interesting ADOM-like game ? 05:25:42 min = min rules and features 05:26:09 probably one room where there is one character standing still and you can only move towards it and use axe :o) 05:26:32 there were many players in school who played adom after school days 05:26:36 in computing class 05:27:36 it was not nice that computing class was shut down because some smart person had tried to override bios by installing cih (he had changed bios language to finnish :o) 05:27:59 that's almost as wise as to steal mouse ball 05:28:26 at my school, stupids throwed rat ball at one another 05:28:41 well we did leave the "hackers" nice surprise when we left... 05:29:00 some fresh fish.. the room was out of use for half year (that was common room, not computing room) 05:29:50 some fish left to rot ? 05:30:18 * Serg_Penguin would find it w/ aid of a dog 05:31:15 they didn't like that much .. but neither did we that the computing class was closed so we couldn't spend time there after school and in breaks 05:31:31 during breaks the whole cs room was always full 05:32:03 * mur ircced there sometimes until 6 pm :o) 05:32:11 it was like cs labs in universities 05:32:18 somethign entirely different from normal schools 05:32:26 people were there til late evening sometimes 05:32:40 spending time at school. until they shut it down because of the smart ones 05:33:07 i heard that sometimes people did leave the place 9 pm :] 05:33:23 i found one rat ball under table and had my own ;))) 05:33:44 ;) 05:34:15 well practically there were 4 types or people in cs room after 4 pm 05:35:27 1. leechers and warez downloaders 2. real person shooting games (3rd person?) who played also rolegames and 3. irccers and 4. random www surfers 05:38:14 i heard what in some RU towns public phones are destroyed by vandals 05:38:46 so authorities decided let's anyone carries his own portable phone, and left only outlets in cabins ;))) 05:38:56 the same i had w/ mouse at school 05:39:29 so everyone have to have own ball to use mouse (unless you want to use fingers inside mouse or keyb only) ? 05:39:55 no one neede rat coz it was 286 05:40:00 no Windoze 05:40:14 aha 05:40:31 only i needed it, 4 DOS gfx editor and writing games in Qbasic 05:40:44 :) 05:40:54 and no Iet ;) 05:40:57 Inet 05:41:05 you shoudl have showed those programs to others and they might have stopped throwing them around :o) 05:42:03 anyway, all balls were lost or stolen by the time 05:42:21 heh 05:42:27 ppl even played SimCity w/ kbd only 05:42:45 wow 05:43:05 and comps were xtremely hangy/glitchy/badblocky 05:43:44 so we, w/ physics teacher, started to make one half-live out of 2 half-dead 05:43:53 this is how i learned hardware 05:44:27 :) 05:45:11 we had Novell network, but it gone down soon 05:45:25 heh, we brought our own hardware 05:45:32 hard disks i mean 05:45:42 and used schools internet connection :) 05:45:50 it was a bit faster that the ordinary modems :o) 05:46:03 traffic cost ? 05:46:12 especially as the whole city network was at school upstairs .. 05:46:19 school network 05:46:24 center / sevrers 05:46:47 i think it was upgraded to 2 MB/s when i was leaving 05:46:53 maybe year ago 05:47:09 sounds quite few 05:47:21 at least the internal network is 12 MB/s 05:47:25 between the schools 05:47:30 (done with wlan) 05:48:37 here schools are home of poverty and crime 05:48:59 it's quite sad 05:49:18 i'll go now 05:49:20 bbl 05:49:23 our school was very lucky to have any computers at all 05:49:38 well not exactly all finnish schools have computers 05:49:56 but i was in big city's school in centre:) 05:50:05 there were too few spaces, how ever 05:50:06 kids of rich are going to 'good' schools for money 05:50:57 we do have some elite schools but all schools are free 05:51:21 elite schools like ressu which students average is about A with english/usa scale 05:51:49 but i think its more about background 05:51:51 in my grade, all boys but me and one other comp geek was heavy alkohol drinkers ;(( 05:52:06 :/ 05:52:39 at graduate party they made outrageous scandal and fight 05:53:07 drunk like pigs 05:53:55 uh 05:54:19 i was wiser - grabbed my attestat (graduate document) few hours be4 and BUG OUT 05:54:33 :) 05:54:36 i'll have to go 05:54:40 i'll be back soon 05:54:44 ok, cul8r 05:54:44 30 - 90 min 05:54:58 cya½ 05:55:00 ! 06:04:36 --- quit: Serg_Penguin () 06:17:13 back 06:31:17 --- quit: tathi ("leaving") 06:48:55 --- join: Serg_Penguin (Serg_Pengu@212.34.52.142) joined #forth 06:49:06 re 06:50:04 re! 06:52:02 how do you think, is it practical to write minimal skeleton of a proggie, and then let it grow wildly out of imagination ? 06:52:10 or you vote for plan ? 06:54:14 well 06:54:29 i like to program creatively (from imagination) 06:54:36 but plan makes it faster 06:54:41 (maybe more dull) 06:56:48 Serg: apparently it is possible because it happens that way time and again and that way is taught in universities 06:57:13 what way ? plan or wild ? 06:57:37 top-down, i.e. skeleton->all-devouring monster 06:58:12 * mur is for creative programming 06:58:24 XeF4_ were you at readme (programming art festivals)? 06:58:46 no 06:59:29 when was it? 07:01:07 mur: have you any ideas for something rogue-like ? 07:01:50 shoudl it have dungeons and dragons and middleages? 07:01:51 guh 07:02:00 what's that time whent here were castles and such? 07:02:13 midages? 07:02:48 not really :/ 07:02:50 middle ages/medieval 07:02:58 i haven't played them so much 07:03:05 i had a good plot: 07:03:23 but you coudl perhaps have a bit another kind figures, like trolls and bugs and such 07:03:24 you and friends come to comp club to play air combat over net 07:03:57 you see a distant target - let's go kick ass ! 07:04:08 mur: when was readme held? 07:04:17 before 1.6. 07:04:29 at uiah houses, lume probably 07:04:36 oops - one mage on a gragon, few on flying wererats 07:04:40 30 - 31.5. maybe 07:05:23 your wing takes them out, but all your friends are down, and you bail out then fuel ends 07:06:09 hehe, you coudl have little humans running inside house and avoiding rats, suddnely the players just shrinks because of unknown reason and they try to find out how to become normal again. (then you coudl make 5x10 sized computer too that clearly looks like one :) 07:06:10 so player is alone in a hostile magic land, and enemy patrols are going to the battle place.... 07:07:17 i also had another idea : 07:07:30 * mur has been doing ascii arts some times 07:07:36 for irc mud kind game :) 07:07:43 newer coded it though :) 07:07:52 go inside your mind and take out anyone who ever brainwashed\controlled you 07:08:09 * XeF4_ hasn't done ascii art for ages 07:08:45 spalsh a teacher w/ fireball, pierce TV man by crossbow bolt 07:11:02 fry a priest by magic lightning 07:11:19 and parents are final boss monsters 07:11:29 .oOOOOq' <- a bug:) 07:11:41 defeat them all and your mind is free ;) 07:12:13 what about a moral thinking of what have been done in the end? ;) 07:12:25 and then the person dies in suicide 07:12:29 how positive ! :D 07:12:40 why suicide ? 07:12:51 heh.. 07:12:53 clearing any brainwash and becoming sane ! 07:13:24 * mur thinks some picassoish ascii art woudl suit such game very well 07:13:28 something abstract 07:13:58 woudl it be top grid or then 3d or left view (like in platformers) ? 07:14:35 3D in ascii ??? heh... 07:15:14 well 07:15:20 i'll make example :) 07:16:28 better use 3Dmax and ascii converter 07:16:35 no no not like that 07:16:48 do you have fixed height font? 07:17:23 * Serg_Penguin has good old console NortonCommander clone 07:17:40 ( ) ___ 07:17:40 o ( ) /___\ 07:17:41 ( ) || | ___ | 07:17:41 /'\ |_| |_| 07:17:59 like that (umm it is not that bad after all in var height font ... ) :) 07:20:00 looks like man gonna pee or he has a gun instead of pee and gonna fire 07:21:30 well that you could move all diagonals and it was a bit like left view 07:22:06 no, better nethack like 07:22:11 @ is a man etc... 07:25:21 even fixed width fonts are VERY different 07:25:48 then i hit Alt-Enter (window/fullscreen), i see very different man ;( 07:26:23 and colours, will you use them? or allow? 07:30:41 i think more of 320x200, VGA 13H 07:30:56 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 07:30:57 --- mode: ChanServ set +o kc5tja 07:31:12 Hi 07:31:18 Howdy 07:31:19 re 07:31:23 hi 07:32:19 * Serg_Penguin goes home 07:33:30 --- quit: Serg_Penguin () 07:53:14 --- nick: mur -> mur_will_beback 08:16:27 --- nick: mur_will_beback -> mur 08:48:24 --- quit: XeF4_ ("POIS") 09:15:10 --- join: remote (~remote@Toronto-HSE-ppp3714296.sympatico.ca) joined #forth 09:25:47 --- join: crc (~crc@ACA7665B.ipt.aol.com) joined #forth 09:26:02 --- quit: crc (Client Quit) 10:41:01 --- part: remote left #forth 11:29:15 doo doo 12:05:58 ewww . . . 12:06:03 not here, please. :) 12:08:02 ;) 12:08:06 --- nick: mur -> mur_brb 12:22:32 --- nick: mur_brb -> mur 12:23:04 wb 12:23:57 thanks! 13:04:00 --- join: a7r (~a7r@206.72.82.135) joined #forth 13:05:30 hey 13:06:04 re 13:06:10 hey 13:07:05 forth is a tiny god 13:07:25 I wrote an assembler for these AVR chips in 15 minutes of pforth 13:07:43 and all my opcodes assembled right the first time. 13:09:45 Lucky you. :) 13:10:05 I wrote an 80286 assembler in PygmyForth for my Forth implementation. It took more than 15 minutes to write. :) 13:10:08 no way, I'm thinking Forth. 13:10:12 And it did NOT produce correct opcodes first time. 13:10:13 er thanking 13:10:41 kc5tja: I think it helps that the AVR is brain-dead simple. 13:10:47 Yeah. :) 13:10:57 it has 3-4 argument encoding formats 13:10:58 My assembler was still quite easy to write though. It's just that it took much longer than yours. :) 13:11:06 x86 has . . . over 50. :) 13:11:13 haha 13:11:30 The user usually uses only a few of those though. 13:11:37 Most of those '50' are for special purpose instructions. 13:11:45 nod 13:11:53 MOV is a huge exception to that rule though. >:( How the hell they could fsck up MOV so badly, I'll never know. 13:12:37 Many op-codes, one mnemonic. 13:12:40 haha, how'd they do it? 13:12:55 It's too complicated to even explain. 13:12:59 ouch 13:13:21 Moves to the CS, DS, ES, and SS registers take a different form than moves to the FS and GS registers. 13:13:29 Primarily because they ran out of bits to encode the extra segment registers. 13:13:49 damn 13:13:54 Moves to AL, AX, or EAX take different "optimal" forms than moves to other integer registers. 13:13:58 Those Intel guys always think ahead. 13:14:23 Moves to CRx registers are different from moves to DRx registers. 13:14:29 I can go on and on. 13:14:51 TreyB: Heheh :) 13:14:56 that's horrible. 13:15:12 Like I said, they gave all of these different op-code the same mnemonic. 13:15:17 Oh, and then there's the TEST instruction. 13:15:22 That's my favorite pet peeve. 13:15:48 ADC, ADD, AND, OR, SUB, SBB, and XOR all have the same instruction encoding format. 13:16:05 TEST, however, not only has different encoding, but *slightly* different modr/m byte encoding. Lovely! 13:16:16 --- join: XeF4 (~XeF4@in-vitro242.gprs.suomen2g.fi) joined #forth 13:16:34 I read an article a while back explaining how the instruction set actually makes sense if you decode it in octal rather than hex. 13:16:54 which machine? 13:16:58 x86 13:17:00 TreyB: Yes, because the 4004 and 8008 were originally made before hexadecimal caught on big. 13:17:20 And, as we all know, the 8086 and its successors are all derivatives of that same basic architecture. 13:17:21 yeah because a lot of bitfields are 3 bits wide or at least fall on 3-bit boundaries 13:17:41 * kc5tja nods 13:17:46 nothing more magical about it in there 13:17:50 s/in there// 13:18:39 Still, say what you will about it, the x86 architecture is a pain in the ass to write an assembler for, octal or not. 13:19:27 : rtor 100 << swap dup 101 << 1000000000 and swap 1111 and or or ; 13:19:30 : /mov rtor 0010110000000000 or ; 13:19:42 kc5: z80 is almost as bad fwiw 13:19:46 that's the AVRs 13:19:55 which you certainly know anyway 13:21:15 XeF4: Worse. 13:21:55 * kc5tja doesn't know what a7r's code does. 13:22:05 What is rtor? 13:24:02 it does register to register encoding 13:24:13 so if you want to move the contents of r16 to r17, you execute: 13:24:20 16 17 /mov 13:24:45 Ahh, 13:52:21 --- quit: mur ("MURR!") 14:10:33 --- quit: a7r (Read error: 110 (Connection timed out)) 14:32:16 --- join: a7r (~a7r@206.72.82.135) joined #forth 14:32:44 --- join: deluxe (~deluxe@pD9EE1FC6.dip.t-dialin.net) joined #forth 14:33:50 elo 14:34:45 re 14:46:35 * deluxe was hoping to meet mur here 14:56:55 --- quit: XeF4 (Read error: 110 (Connection timed out)) 15:08:08 --- quit: paxl (asimov.freenode.net irc.freenode.net) 15:08:08 --- quit: cleverdra (asimov.freenode.net irc.freenode.net) 15:08:08 --- quit: ianni (asimov.freenode.net irc.freenode.net) 15:08:09 --- quit: ChanServ (asimov.freenode.net irc.freenode.net) 15:08:09 --- quit: TreyB (asimov.freenode.net irc.freenode.net) 15:08:09 --- quit: kc5tja (asimov.freenode.net irc.freenode.net) 15:08:23 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 15:08:23 --- join: ianni (ian@inpuj.net) joined #forth 15:08:23 --- join: paxl (paxl@modemcable110.168-130-66.que.mc.videotron.ca) joined #forth 15:08:23 --- join: cleverdra (nailuj@ACC6186F.ipt.aol.com) joined #forth 15:08:23 --- join: TreyB (~Trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 15:08:23 --- join: ChanServ (ChanServ@services.) joined #forth 15:08:23 --- mode: asimov.freenode.net set +oo kc5tja ChanServ 15:18:15 --- join: XeF4 (~XeF4@nucleus28.gprs.suomen2g.fi) joined #forth 15:25:47 --- quit: XeF4 ("pois") 16:11:26 --- quit: deluxe (Read error: 54 (Connection reset by peer)) 17:08:33 --- join: Herkamire (~jason@h0030657bb518.ne.client2.attbi.com) joined #forth 17:20:08 re Herkamire 17:20:14 haven't seen you here in a bit. 17:20:20 Hey Herkamire :) 17:21:16 hi guys :) 17:21:45 yeah, I've been doing a bunch of forth programming lately though :) 17:22:39 Of what nature? 17:22:40 * kc5tja has too. 17:22:49 I'm a few hours (translation: probably 8) away from having my forth bootstrap completely, and not need any of the PPC asm code I compiled with gas 17:22:55 Oh, nice. I continued working on an old Forth project today, but then I noticed I had forgot how to use that compiler I wrote like a month ago. 17:23:29 I have been picking/working at this code consistantly that I haven't forgotten how any of it works :) 17:23:32 FS/Forth for DOS interpretted its first set of commands mid-way last week, and by last Sunday, compiled its first colon definitions. 17:23:33 that's good :) 17:23:46 Herkamire: what PPC setup? 17:24:01 * kc5tja has sufficient documentation stored in shadow blocks that I've not forgotten how the code works. I started FS/Forth last year, about this time. 17:24:20 a7r: Apple hardware. Dual G4 450Mhz 17:24:56 Herkamire: sweet, what're you planning on doing with it? 17:25:01 Herkamire: Neat. Is your Forth environment going to run on bare hardware? If so, will it handle the two CPUs? 17:25:14 kc5tja: Complex solutions, bad memory and poor documentation don't go very well together :) 17:25:39 I'm planning on making it into an OS 17:25:54 * kc5tja nods 17:25:55 I've been looking for a Forth OS for my laptop. 17:26:02 * kc5tja has a ways to go before making FS/Forth an OS. 17:26:17 I've got an iBook, and a dual G4 17:26:25 a7r: Well, I'm trying to write one. I do need to write FS/Forth for Linux first though. It just became a commercial priority for me. 17:26:31 right now it works under linux only. I may develop it into a convenient linux development tool. 17:26:48 kc5tja: nod, I want one for my router as well. :) 17:26:51 I know how to make elf files, and I have plans of sticking the source and kernel in an ELF. 17:27:03 werd. 17:27:07 * kc5tja isn't going to do that. 17:27:21 kc5tja: I plann to use both CPUs eventually. 17:27:26 I'm going to have a C file which acts as a "boot loader" for the Forth environment under Linux. It'll read in a binary image from a block file. 17:27:44 Though I don't think I'm going to be swapping processes from one CPU to another. 17:28:00 Once the C boot loader reads in the executable image, it'll jump to it directly. At that point, the Forth image shouldn't have any further reliance on the bootloader. 17:28:10 Herkamire: if you want anyone to test, I'm interested. 17:28:11 (except to exit gracefully to the Linux prompt again) 17:28:40 I think I will use one CPU for background tasks (such as playing music, or rendering a raytraced movie) and one for interacting with me. 17:28:57 sounds good 17:29:27 Sounds like a waste of a perfectly good PowerPC to me. :) One is doing all the busy-work, while the other is wasting 99% of its time waiting for user I/O. :) 17:29:39 haha 17:29:47 depends on what you're doing. ;) 17:30:07 * kc5tja spent some more time in XcolorForth today. 17:30:28 It's actually more stable than I thought. It seems it crashes only if you edit blocks 0 through 17. 17:30:47 (Note: by edit, I mean you type 0 EDIT or whatever -- I didn't actually change anything) 17:31:10 I might go crazy and do full SMP, but I don't know. 17:31:14 Also, it'll crash when attempting to access various hardware registers directly. Of course, this is to be expected when running under X. 17:31:17 if I have a use for it... 17:32:12 Herkamire: I say not to bother with full SMP -- twin cooperative multitasking environments running concurrently ought to be PLENTY sufficient. Some synchronization will be required between the two CPUs, but it will be a lot simpler than a full-blown SMP. 17:34:29 that kind of stuff is miles away. I need to write an editor first, then get it to boot, then get the network card working, and write some apps (web browser, email) 17:34:48 that puts any sort of multitasking a year from now at least I think. 17:35:24 Depends. At least, SMP multitasking. 17:35:39 I'm not real clear on what SMP is actually. 17:35:47 ColorForth uses multitasking to implement its display management tasks as background tasks. Not sure how you're going to handle output. 17:36:14 I don't think it will be worth it to make it so I can swap a process from one CPU to the other. 17:36:20 Symmetric Multi-Processing -- where all microprocessors in the system are running at the same time, from the same memory store, each running different threads of execution. 17:36:49 It's "symmetric" because all CPUs are the same kind, and threads can migrate. 17:37:05 If threads cannot migrate, OR, if the CPUs are not of the same kind, it's called AMP -- Asymmetric Multi-Processing. 17:38:41 cool 17:40:16 I haven't figured out how the OS will deal with that. 17:41:16 anyway, that's miles off. 17:41:52 current task: bootstrap fully (have forth code that outputs a kernel (ELF file)) 17:50:32 I'll probably make an OS to see that I can, but it probably won't be terribly useful for at least 2 years. 17:51:49 --- quit: Herkamire ("dinner time") 17:57:59 --- join: Herkamire (~jason@h0030657bb518.ne.client2.attbi.com) joined #forth 17:58:51 yessss :) my girl frind is getting the food. that gives me 30+ minutes :) 17:59:47 werd. 18:00:28 my big question tonight is: do I right this microcontroller firmware in C, or in asm and Forth 18:02:18 forth forth1 18:02:36 I need to write my own Forth though, if I go that route. 18:06:49 cool :) 18:14:59 a7r: Is this for the AVR device? 18:25:18 kc5tja: yeah 18:26:19 * Robert is writing Forth code for his AVR. :) 18:26:38 Robert: you mind if I take a look at what you've got? 18:26:42 Soon I'll be coding Forth _on_ a microcontroller, through a serial console. 18:27:04 a7r: Of course not. But I think you'll laugh ;) 18:27:09 yeah, that's what I'm looking for, I want to be able to modify the firmware in the field, over a serial console. 18:28:15 Robert: I won't laugh, got a URL? 18:28:21 a7r: http://robert.zizi.org/pub/projects/mtv-030611.tgz is the latest snapshot. 18:28:25 killer. 18:28:52 hrm, got a 404 18:28:59 oops 18:29:05 a7r: http://robert.zizi.org/pub/projects/mtv/mtv-030611.tgz 18:29:19 great. 18:29:38 minmos.f is the start of a Forth system for the VM (called MTV). fc.f (in mtvfc/) is the Forth compiler I use to compile code for the VM. 18:29:48 Because of the change of emphasis (in particular, towards increasing minimalism) in my Forths, FS/Forth 2.0 will henceforth be called FS/MachineForth 1.0. 18:29:59 The forth/ dir contains obsolete code. 18:34:57 Robert: is there any weirdness on the AVR, w/ executing instructions out of SRAM? 18:36:03 Uhm.. what do you mean? 18:36:25 I've done that, with simple programs. Not really enough SRAM to use it for big ones. :) 18:36:45 yeah, I just mean calling into, and executing instructions, from SRAM.. 18:36:48 From the VM of course, since the AVR can't execute any code except from the built-in Flash :( 18:36:57 yeah, that's what I mean. 18:37:03 Then I've done it. 18:37:19 Both RAM, internal EEPROM and an external DataFlash chip. 18:37:34 but not AVR instructions right? you mean instructions for your VM 18:37:39 Yes. 18:37:43 okay. 18:37:47 It would also be possible to execute code from another computer, or any other device. 18:38:02 Just write one function to read and one function to write to your device. 18:38:02 so a direct threaded Forth isn't possible right? 18:38:26 In native AVR code? 18:38:30 yeah 18:38:31 Sounds like a byte-coded or packed-instruction VM is required to run Forth on that processor. 18:38:35 No, it isn't. 18:38:45 kc5tja: nod, that's what I'm thinking. 18:38:49 kc5tja: Yes, and that's why I wrote one. :) 18:39:10 I'm just trying to evaluate my options. 18:39:11 I believe that's how the PIC-Forths work too. 18:39:12 kc5tja: Such a neat little computer HAS to run a Forth! 18:39:36 I think the PIC instruction set is a little bit small. 18:39:53 Not suited for assembly programming, unlike the AVR. 18:39:57 Executing an enhanced Steamer-16 instruction set would be neat. I've been meaning to make a 16-bit homebrew CPU along those lines. 18:40:33 Robert: It's precisely because its instruction set is so small that you must implement a VM to get any kind of decent programming environment on it. 18:40:43 Hey, if they have BASIC for the damn things, it ought to be possible to run Forth! 18:41:02 8051 chips can execute code from external memory, right? 18:41:06 :D 18:41:08 Sure. 18:41:35 No, it can't. It's a Harvard-architecture device. 18:41:40 A *crappy* one, but one nonetheless. 18:41:41 :( 18:42:09 You know that it takes 12 clock cycles to complete a single bus cycle for that architecture? It's *horrible*. 18:42:40 I'm working with a VM that requires several times that for simple instructions ;) 18:42:45 To get 1 MIPS on an 8051, you need to give it a 12MHz clock. 18:43:03 I thought Motorola's 4x clock requirement was bad. 18:43:20 Hehe. 18:43:38 I think some of the third-party 8051-clones can run code in SRAM, but I believe there are limitations on how to do it. 18:43:46 PICs also use 4 cycles/instruction, but AVR is nice, most instructions take just one cycle. 18:43:48 But Intel's (original) 8051 didn't. 18:44:06 * kc5tja thought the AVR was an 8051 clone, actually. 18:44:23 I suppose you could have the Forth be able to flash the AVR directly. 18:44:24 Apparently it isn't. 18:44:32 but that gets 'interesting' 18:44:33 No, AVR is the AT90 series of microcontrollers, the AT89s are 8051 clones. 18:44:49 Ahh, I knew there was some association of AVR and 8051s. :) 18:45:27 The AT89s are very cheap... I bought some just to see what I can do. 18:45:52 See, the AT89s might be able to run code in external RAM. I don't know. 18:46:12 There's always SOME use for a small computer :) 18:46:37 Hey, someday, I fully intend on building my 65816-based PC. 18:46:55 I have 7 of the dang CPUs, I may as well use them for something. 18:46:57 * Robert would like to build a computer from scratch, too. 18:47:14 Generating video is a pretty major problem for me though. 18:47:18 Got a z80, so... 18:48:03 I guess so. Thought about what one could do with an AVR.. the big ones have like 1 kB SRAM, if you use that for ASCII video memory, you could have something like a 16x64 text mode :) 18:48:20 Just to give it a nice, primitive feeling. 18:48:38 * kc5tja is looking for a graphical system. 18:48:48 Unless of course you're happy with a couple of LED segments and a bunch of LEDs :D 18:49:18 Getting late here now, 4am, should go to bed... Night! 18:49:23 Heheh :) 18:49:28 'night Robert 18:59:02 How fast do AVRs run? 19:03:18 up to 16 MHz or so 19:03:25 hmmm.... 19:03:27 with most instructions taking 1 clock cycle. 19:03:52 These are 8-bit wide microcontrollers? 19:03:56 yeah 19:04:01 16-bit opcodes 19:04:50 And some are 32-bit 19:05:00 yeah 19:05:18 --- join: jstahuman (~justahuma@pcp053555pcs.brlngt01.nj.comcast.net) joined #forth 19:05:44 hmm 19:06:01 * kc5tja still doesn't think they'll work as a cheap video display device. 19:07:02 Not for my needs, at least. 19:07:06 What would you use? 19:07:07 Oh well. 19:07:17 I don't understand the question. 19:08:01 Nevermind then, I should go to bed for real instead ;) 19:10:12 Heh 19:10:21 Well, I'm on the way out myself. Going to get dinner, then we're going to see a movie. 19:10:30 we == myself and my roommate 19:19:36 --- nick: kc5tja -> kc-movie 19:43:44 --- quit: a7r (Read error: 60 (Operation timed out)) 20:31:29 --- join: a7r (~a7r@206.72.82.135) joined #forth 21:45:59 --- quit: jstahuman ("leaving") 21:56:49 --- quit: Herkamire ("bedtime") 22:16:25 --- join: Serg_Penguin (Serg_Pengu@212.34.52.142) joined #forth 22:16:43 hi ! 22:16:53 --- nick: TreyB -> TreySleeping 22:17:02 --- nick: TreySleeping -> TreyB 22:18:04 hey 22:53:29 --- join: wtail ([G3YocnIaI@panix1.panix.com) joined #forth 22:54:51 --- part: wtail left #forth 23:08:12 --- quit: Serg_Penguin () 23:36:02 --- nick: kc-movie -> kc5tja 23:36:15 back 23:58:16 hey 23:59:59 --- log: ended forth/03.06.10