00:00:00 --- log: started forth/02.03.28 00:00:12 --- join: davidw (~davidw@ppp-74-10.25-151.libero.it) joined #forth 00:04:13 --- join: MrGone (~mrreach@209.181.43.190) joined #forth 00:04:13 --- quit: _MrReach_ (Read error: 104 (Connection reset by peer)) 01:22:54 --- quit: davidw (Read error: 104 (Connection reset by peer)) 01:22:56 --- join: davidw (~davidw@ppp-74-10.25-151.libero.it) joined #forth 03:12:12 --- join: tom (tom@adsl52235.vnet.hu) joined #forth 03:17:59 --- nick: tom -> onetom 06:21:13 --- quit: davidw (Read error: 110 (Connection timed out)) 06:40:51 --- join: Fare (fare@samaris.tunes.org) joined #forth 06:51:38 hi 06:52:17 is there any1 here who already has managed 2 boot into forthos (www.forthos.org) 06:59:48 no. I suppose it needs djgpp and/or vsta as a metasystem to compile it 06:59:58 I feel that retro has more stuff and less C. 07:00:45 heh 07:02:24 djgpp?? aaajaj, thats sad 07:17:08 has anybody have some experience w structures? 07:17:32 which structure model is the most convenient? 07:17:47 heh 07:17:55 --- nick: MrGone -> MrReach 07:18:30 well, you need byte, word, long word, and long long word 07:18:59 and mesc length fields ... like 8.3 filenames 07:19:25 I wrote my own structs, actually 07:19:46 use it almost everywhere ... but that's the sotry of most forthers 07:19:48 and how do u avoide name conflicts? 07:20:03 well ... 07:20:06 w naming conventions? 07:20:10 it depends how you mean 07:20:13 w vocabularies? 07:20:14 ok ... 07:20:36 each TYPE of structure gets its own prefix 07:20:54 WAV.Freq might be a field, for example 07:21:09 or WAV_Freq 07:21:47 and inside a var, it look like tmp WAV.Freq @ blah blah 07:22:07 "tmp WAV.Freq @ blah blah" 07:22:34 the word WAV.Freq adds its own offset 07:22:45 and how do u nest structs, then? 07:22:56 s1 struct 07:23:10 field s1.f1 07:23:11 well, right now I have a word "sizeof" 07:23:15 field s1.f2 07:23:17 endstruct 07:23:24 s2 struct 07:23:30 field s2.f1 07:23:47 field s2.f2 (this is type s1) 07:23:49 endstruct 07:23:56 then u say 07:24:04 "somestruct sizeof FIELD large_subfield" 07:24:18 s2.var s2.f2 s1.f1 @ 07:26:02 onetom: how would *YOU* want a structures package to work? 07:26:18 or should i note s2.f2 is an s1 type? 07:26:32 I'd want struct union1 substruct @ 07:26:34 so i could name it s2.s1.f2? 07:26:41 I'd want "struct union1 substruct @" 07:27:04 well, yes, Win32forth does that 07:27:13 but I prefer not, personally 07:27:36 I prefer "s2 s1 f2 @" 07:27:48 do you follow? 07:27:50 id like 2 c a "more HL" implementation 07:27:57 well, not really 07:28:08 let me meditate on it a bit 07:28:14 ok, lets pretend we have a simple struct ... 07:28:24 struct: t1 07:28:33 1 bytes f1 07:28:46 1 words f2 07:29:00 1 cells f3 07:29:11 ;struct 07:29:17 (follow?) 07:29:21 yes 07:29:28 ok, then we have ... 07:29:40 struct: t2 07:29:55 1 bytes f4 07:30:12 t1 sizeof f5 07:30:30 1 cells f6 07:30:36 ;struct 07:30:42 (follow?) 07:30:44 Q: 07:31:03 ok 07:31:04 should the fieldnames b unique, then? 07:31:15 yes 07:31:23 eermmm 07:31:24 or more extenuous 07:31:34 so now I would have ... 07:31:37 and how would u name them 07:31:45 yes, tell me.. 07:32:06 "tmp f5 f2 w@" 07:32:11 (follow?) 07:32:23 ....thinkin... 07:32:37 well, that depends entire on the structure 07:33:08 right, up to now, but 07:33:20 "f2" is a word allocation, so can only W@ or w! to it 07:33:50 how would u prefix the field names so they could b unique? 07:34:00 but f2 is a subfiled of f5, a type of t1 07:34:14 you wouldn't 07:34:34 if you want to "!" and "@" to f2, you could 07:34:43 forth is unique that way 07:35:03 .... 07:35:10 the programmer *HAS* to keep track of sizes 07:35:22 i think i should find a real world problem... 07:35:31 nooo, theres nothing wrong w the sizes 07:35:44 the problem is only around the names 07:36:07 that's discipline on the programmer's part 07:36:47 lets consider a "list" and a gemoetrical object, say, a vector 07:37:00 nono, one or the other 07:37:05 both could have a field names "length" 07:37:10 a two-part vector? 07:37:18 k, let me show it 07:37:27 vectors don't have variable lengths 07:37:28 what other lang do u know? 07:37:38 langs 07:37:40 just about any except perl 07:37:45 pascal, C ? 07:37:58 only "fluent" in quick basic 07:38:00 k, then let me write it in pascal, k? 07:38:04 yes, either 07:38:08 sure 07:38:19 "records" 07:38:20 me fluent in pascal :) 07:38:22 yup 07:38:26 type 07:38:36 TList = record 07:39:01 data: ~array of integer; 07:39:08 s/~/^/ 07:39:19 length: integer 07:39:30 end; 07:39:40 TVector = record 07:39:43 ok, I read that as ... 07:39:59 one pointer to an array, and a count ... right? 07:40:03 origin: TPoint; 07:40:14 length: real; 07:40:17 end; 07:40:22 TPoint = record 07:40:29 x, y: real; 07:40:30 end; 07:40:40 lets play w this system 07:40:45 ok 07:40:53 & yes, u werer right 07:41:04 so the problem is the 07:41:12 clash between the two length fields 07:41:31 not really, but the compiler doesn't check for you 07:41:42 follow? 07:41:51 he? 07:42:05 ok ... 07:42:35 let's pretend we have a POINT that consists of x and y, but 32b uints 07:42:38 follow? 07:42:50 let's pretend we have a POINT that consists of x and y, BOTH 32b uints 07:42:58 yes 07:43:24 now what happens if you "c!" to a POINT? 07:44:16 whatever it ends up being, it's likely to be garbage ... do you follow? 07:45:28 now, the nameing conventions for a point is pretty obvious ... 07:46:06 given a point in a VALUE tmp ... there's "tmp .x" and "tmp .y" 07:46:30 both of which can be @ and ! ... follow? 07:46:32 hoohohoho 07:46:57 forget that theres no auto type chking, thats not a problem 07:47:04 heh, is "hoohohoho" a good or bad thing? 07:47:12 what words should i define 07:47:13 that's right 07:47:24 if id like 2 define the above struts in 4th 07:47:25 ? 07:47:35 who should i represent 07:47:39 what is the most primative? 07:47:46 the following references in 4th: 07:47:49 --- part: Fare left #forth 07:47:49 var 07:47:54 a point of two reals ??? 07:48:02 AList: TList; 07:48:11 AVector: TVector; 07:48:25 AList.length = 111; 07:48:27 nope, TVector is more prim 07:48:31 AVector.length = 222; 07:49:21 think interms of "bytes allocated" and how to calculated 07:49:39 (and dont bother w that TPoint, it doesnt count @ the moment) 07:50:02 (and its just an ad-hoc example, thats why its a bit non-sense) 07:50:38 but it STARTS with the TPOINT ... even though you might change its size later 07:51:14 forth is an EMPICAL language 07:51:38 it doesn't go back later and try to fugure out what you *REALLY* meant 07:52:19 do you follow? 07:52:23 sec 07:52:33 * onetom is on the phone 07:52:37 ok 07:53:07 if you knew how drunk I was, we prob wouldn't be having this conversation at all 07:53:08 * onetom is back 07:53:15 heh, wb 07:53:19 :)) 07:53:56 yes, theres no type checking, but how would u translate 07:54:03 those pascal examples into 4th? 07:54:10 imean, the 07:54:13 AList.length = 111; 07:54:15 AVector.length = 222; 07:54:18 ? 07:55:11 not just try 2 tell me & tell me - a'la matrix ;) 07:55:34 ok, there's two issues ... 1. the syntax of the structures package and 2. the fields themselves 07:56:08 I think Matrix one of the lames scripts ever written 07:57:01 ok, let's assume a data type that consumes 111 bytes in length 07:57:42 ok? 07:58:22 (that's a HUGE real, btw) 08:02:48 --- join: davidw (~davidw@ppp-144-9.25-151.libero.it) joined #forth 08:02:55 nevermind 08:03:23 I've got other things to think about, right now. 08:03:27 --- part: MrReach left #forth 08:03:40 --- quit: rob_ert (Remote closed the connection) 08:16:44 eh, we r talkin beside eachother, i think :-/ 08:36:00 --- join: uiver (~segher@a43195.upc-a.chello.nl) joined #forth 08:49:41 --- join: herkamire (~jason@68.15.54.54) joined #forth 08:54:44 anybody here written a forth engine? 08:55:37 yes. 08:56:15 do you guys think forth would be good as an apache module? (to replace PHP or whatever) 08:56:53 should I write my own forth engine just to get better with forth? (even if I'm planning on using someone elses) 08:57:24 herkamire: no, it would be pure torture 08:57:33 (apache + forth, I mean) 08:58:06 davidw: getting it to work? or doing scripting for web pages? 08:58:36 doing scripting for web pages 08:58:44 use a languages that's good at string handling... 08:59:20 like Tcl 08:59:28 for web scripts, you want a language that's good at databases and/or strings, i.e., perl. 09:03:35 hmmmm. I haven't done enough programming in forth to know if I can make it good at string handling. 09:04:10 good luck;-) 09:04:23 hehe 09:05:47 I hooked gforth up as a CGI and was playing around with this a little 09:06:22 I whipped up a print<< sounds like as much fun as hitting yourself in the nuts with a hammer 09:08:18 I never thought seriously about doing databases with it though :) 09:09:39 I was thinking that I could get it set up so it would be really easy to create pages (invent my own syntax). I could then run it with forth, and it could output as html or plaintext or ps or whatever. 09:10:26 or you could just run your head into the wall a lot 09:10:43 and I was thinking of writing a forth browser that could take the code and do any of those on the client side :) :) needless to say I haven't bothered with this stuff :) 09:13:20 well I gotta go 09:13:22 --- quit: herkamire ("Client Exiting") 09:45:11 --- join: rob_ert (~robert@h237n2fls31o965.telia.com) joined #forth 10:17:02 --- join: Speuler (~l@195.30.184.51) joined #forth 10:17:13 'morning 10:18:29 --- join: herkamire (~jason@wsip68-15-54-54.ri.ri.cox.net) joined #forth 10:29:39 Speuler: Merry christmas. 10:31:30 I see you have 'date' written in forth 10:36:51 davidw: what do you do with forth? 10:37:47 herkamire: hahoo 10:38:31 ? 10:38:45 herkamire: denigrate it, mostly 10:38:49 your idea cgi-4th idea is absolutely usable 10:40:32 hehe 10:40:53 I'm trying to use it for genetic programming 10:41:00 it's perfect for the programlets 10:41:10 unfortunately, it's a halacious pain in the ass for everything else;-) 10:42:11 davidw: wow, genetic algo is a very nice thing 10:42:25 I don't care much for complex html. I was thinking of something like texinfo, plus the ability to interact with servers 10:42:31 no, genetic programming, not algorithms 10:42:33 lol 10:42:38 but its not too complicated anyway 10:42:48 u just have 2 operate on arrays 10:43:09 u mean, genes? 10:43:10 davidw: did you write your own forth engine for the genetic stuff? 10:43:16 fuck no 10:43:27 what engine are you using? 10:43:32 gforth 10:43:43 you think I'm a glutton for punishment or what!? 10:44:00 onetom: basically, I generate a lot of forth words 10:44:03 then I run them 10:44:07 then I see how that compares 10:44:16 you ARE using forth ;) 10:44:17 then... I still haven't really got that far 10:44:58 herkamire: there is a nice 4th-html app on a site... let me look 4 it 10:45:16 cool 10:45:40 http://www.ultratechnology.com/rmvideo.htm 10:45:51 near the bottom of the page 10:46:09 there is "Neil Bawd's actual Ugly Home Page" 10:46:28 those pages r actual 4th programs 10:46:35 oh, isn't there a wiki written in forth? 10:46:41 but if u filter them through a simple 4th program 10:46:56 it generates html pages from them 10:47:32 so those sources can also b read in a nice html representaion 10:47:58 and in its native source form 10:48:01 their 10:48:35 there is a wiki 10:48:44 where is that wiki? 10:51:22 yeah, I was thinking it would be fun to make a simple parser like that to make web pages quick. 10:51:28 cant find @the moment... 10:53:03 herkamire: use a language that's meant for it... tcl, perl, python... 10:54:19 I am. I'm using PHP. it's just kinda annoying that's all :) Seems kludgy 10:54:50 oops. my mistake. it seems to be a wiki abouth forth, not a wiki written in forth. 10:59:06 * onetom is surfing and finds *very* exciting pages about 4th chips & 4th philosophy 11:00:59 html forms work, but they are pretty limited as far as being interactive. 11:02:12 I was thinking of making it so you can write scripts in for to do interactive stuff and if clients support it you can send them the little forth scripts, and if someone connects with a standard www browser than the scripts can be run by the server to create something usable through http/html 11:02:25 scripts in forth 11:03:36 you've seen Chuck Moore's 25x chip right? http://www.colorforth.com/25x.html 11:09:48 this is a cool page about fourth values: http://www.c2.com/cgi/wiki?ForthValues 11:12:42 yeah, but forth needs registers or something 11:12:54 twiddling stuff on the stack is a pain in the ass, and error prone, IMO 11:14:11 it just seems 2 b 1 11:14:33 if ur programming style is apropriat 11:14:34 e 11:14:51 stack is not a pain, its a gain 11:15:09 well, errors.... 11:15:25 if u write so small words 11:15:40 its also harder 2 make mistakes 11:16:39 * onetom still cant finds that valuable wiki about 4th.. 11:16:45 find 11:17:35 http://www2.tunes.org/cgi-bin/TunesWiki?ForthOS 11:17:47 has been found finally 11:20:42 oh. sorry. this is what I found: http://www.forth-ev.de/wiki/html/Forth/ForthWiki.htm I haven't looked at it much though. 11:22:09 I have not done a lot of forth programming, but usually (if not always) when I was doing a lot of stack twiddling I eventually figured out a quicker simpler way to do it 11:23:46 also, if I'm doing a lot of stack twiddling I figure that I probably didn't get the data on there in the right order, and I look at my earlier code. 11:24:33 davidw: what are you trying to get your genetic programs to do? 11:26:29 work 11:27:06 davidw: it probably wouldn't be that hard to write a little forth engine for your evolution thing. that could make it go magnitudes faster, which may be required if you want them to evolve far enough to do anything useful. 11:28:00 you probably wouldn't even have to deal with text 11:28:16 if I had to deal with text, there is no way I would be fooling with forth 11:28:28 herkamire: how much experience do u have in 4th programming? 11:28:48 herkamire: (i dont have much personally :-/ ) 11:28:54 not much really 11:29:17 davidw: its would b nice if u could tell us the 11:29:23 I write something that translated to dvorak as you typed and is almost a text editor 11:29:30 davidw: specification of your problem... 11:30:03 herkamire: could u explain it a bit more? 11:30:21 herkamire: we r about 2 write a multiuser editor w futhin 11:31:00 onetom: manage an array of structures, which each contain a number, and an array or list of execution tokens 11:31:11 herkamire: (futhin is also a #forth member, and my mate in planning a 4th OS) 11:31:25 mr moore already did a forth OS 11:31:27 you could have your evolver thing spit out 8 or 16 bit codes for the instructions. the forth engine could look them up in a table (hash table?) and run them or spit out an error (stack under/overflow etc) 11:31:40 although he doesn't seem to believe in swapping or threads or anything like that 11:31:40 davidw: "list of execution tokens" r words infact... 11:32:30 davidw: yes, he did some kind of 4th os, but thats pretty basic 11:32:30 chuck is only doing one thing at a time :) 11:32:49 chuck sais he uses his os constantly I believe 11:32:58 davidw: there r more feature rich 1s like enth/flux 11:33:00 my roommate and I are making a forth OS 11:33:17 he just wrote the forth engine in asm for ppc 11:33:26 it works under linux. 11:33:32 ppc? neat 11:33:35 I use a tibook 11:33:41 --- join: tathi (~tathi@wsip68-15-54-54.ri.ri.cox.net) joined #forth 11:33:54 well, im writing a 4th kernel in EveryLanguage :) 11:34:08 we've gotten the machine to boot and everything, but we haven't gotten keyboard input hooked to the forth engine yet 11:34:23 impressive 11:34:34 I've just been helping with testing. 11:34:48 (if i wanna sound like that auld-english person in dramas, named Everyman) 11:34:54 :) there he is. tathi wrote the asm 11:35:27 what's everylanguage? 11:35:33 actually, 2 days b4 i was experimenting w kernel.awk 11:35:47 a simple threaded-code execution engine in awk 11:35:53 but it has some bugss 11:37:17 where is this fucker going wrong...grrrr 11:38:39 hermantom@yahoo.com 13:04:21 --- quit: Soap- (Read error: 104 (Connection reset by peer)) 13:24:20 --- join: Etaoin (~david@ljk2-14.sat.net) joined #forth 13:29:51 ale`... made the evil piece of shit work:-) 13:30:16 yummm :) 13:30:28 and what was the problem? 13:30:57 wrong sized entry 13:31:12 well, on to the next problem 13:31:13 hrm 13:31:56 & do u still think forth is baaaad? 13:32:24 yep 13:32:32 it needs registers 13:33:25 could we have a look @ yr code? 13:33:53 if you want 13:33:55 id b happy if i could show u a more 4this solution what doesnt starves 4 regs 13:33:59 --- join: I440r (~mark4@1Cust78.tnt2.bloomington.in.da.uu.net) joined #forth 13:34:17 davidw: ithink so 13:34:19 not many ppl here :P 13:34:21 heh 13:34:30 oh, there r 13:34:33 onetom: where shall I send it? 13:34:41 even new 1s! 13:34:46 davidw: hermantom@yahoo.conm 13:34:47 :) 13:34:49 davidw: hermantom@yahoo.com 13:35:32 it's not commented much/at all 13:35:49 davidw: nevermind, sendit! 13:35:56 sent 13:36:04 davidw: i will comment it 4 u ;) 13:37:59 dinner time 13:38:13 davidw: good appetite 13:38:46 I'd have to sort of agree about the registers... 13:39:16 tathi: hi maan! :) 13:39:25 not that I've ever run into a situation where I felt cramped, but I just feel like forth is kind of inefficient on RISC CPUs with all these registers 13:39:30 (now u can write in my name ;) 13:39:36 onetom: hi :) 13:39:41 ? 13:39:45 oopps :)) 13:40:16 tathi: herkamire is watching the same screen as me 13:40:41 so I see 13:40:45 and he (or she?) can also type in the middle of my sentences 13:41:14 that BitchX? 13:41:22 EPIC4 13:41:44 and ive not even showed my further screens! 13:41:46 look 13:42:05 im also running a light control system inside it 13:43:12 nah, do u c it? 13:44:08 sure, coz 13:44:09 inside emic or screen? 13:44:26 the word main captures all the keys 13:44:30 except Q 13:44:40 but plz dont stop it 13:45:13 unless u want my father not able 2 switch his lamps on or off :) 13:45:31 and now, what u c is an icq client 13:45:45 do u c it? 13:45:52 yeah 13:46:16 ok, then end of show, right? 13:46:22 ok :) 13:46:37 is that a good icq client? 13:47:09 well... it works, but not really good 4 serious chating 13:47:34 u can easy send msgs 2 a wrong destination 13:48:04 ahh :) 13:48:08 its 1-char-command user interface is really awkward 13:48:32 and dont even has a separate command line... 13:48:39 so.. yuck :) 13:48:46 :) I'll stick with gaim :) 13:48:49 irc is much more better 13:49:13 does gaim has a console interface? no. ergo, it sux :P ;) 13:49:22 hehe :) 13:49:28 I haven't gotten that extreme about it. 13:49:49 coz u can run it inside a screen, uknow :) 13:49:53 --- join: qless (~cerberus@clgr000977.hs.telusplanet.net) joined #forth 13:50:35 tathi: we should discuss this 4th os thing later if u like 13:50:36 I have X running almost all the time anyway... 13:51:02 --- part: qless left #forth 13:51:06 just like me, but i love to access my virtual workplace from eveywhere 13:51:27 from my girlfrineds, eg 13:51:35 netcaffes 13:51:45 from the unviersities 13:51:49 and so on 13:53:00 but lets work a bit now 13:53:11 b back l8r 13:53:53 and now back to your irregularly scheduled programming 13:54:13 ööö, what? 13:55:40 herk gaim is a bit crappy - but then so are most other chat clients in linux (and windows hehe) 13:56:51 right, but vicq is selfcontained @ least. its implemented in 1 big perlscript 13:57:03 hehe 13:57:27 tathi: what do U think about the multiuser editor? 13:57:38 what multiuser editor? 13:57:53 herkamire didnt told it 2 u? 13:58:05 no, I've been pretending to do actual work... :) 13:58:08