00:00:00 --- log: started forth/02.03.29 00:03:11 --- join: futhin (~thin@h24-64-174-2.cg.shawcable.net) joined #forth 00:03:43 all fear! i have shamelessly promoted this channel on two forth wikis! 00:03:54 http://www.forth-ev.de/wiki/html/Forth/ForthWiki.htm and http://www.c2.com/cgi/wiki?ForthLanguage 00:56:14 --- join: davidw (~davidw@ppp-25-8.25-151.libero.it) joined #forth 01:09:05 --- join: Soap` (flop@210-54-78-9.dialup.xtra.co.nz) joined #forth 01:28:12 --- join: davidw_ (~davidw@ppp-179-19.25-151.libero.it) joined #forth 01:35:02 davidw_: i was disparaging about the statements you made re: "hitting your nuts with a hammer" because that was just mean :P 01:36:04 programming in forth is very slow going 01:36:08 it makes you focus on the details of how you implement something, rather than the algorithm 01:36:46 yeah 01:36:50 there's a steep learning curve 01:37:09 but after that, it's permanent, if you go and code in some other language, you'll still code forthish ;) 01:37:26 beyond the learning curve, so much stuff is just focusing on how to implement it 01:37:27 which isn't bad, for didactic purposes 01:37:58 but for real work, something higher level just makes more sense 01:38:45 --- join: aum (~david@l76-155.world-net.co.nz) joined #forth 01:39:19 after the learning curve, and some practice coding, you can do real work in forth, and probably end up with better code (more compact solution, etc) 01:39:30 true 01:39:46 i need a good infrastructure for high-level abstraction 01:39:46 I don't know, I'm not convinced 01:40:01 I can see that with a language like scheme, but not with forth 01:40:41 * aum is considering doing a forth with 8-byte cells, being 01:40:45 my idea of abstraction does not include twiddling bits on a stack 01:41:54 i developed an app in forth that's now got a user base, took me far longer to do it in forth than it would have in other languages - forth is candy-coated assembler 01:42:08 --- nick: aum -> aum|dinner 01:42:18 yep 01:42:30 which is great for some things 01:42:46 i agree that forth needs to be extended and supported better 01:42:46 but not 'high-level' nor 'abstract' 01:42:57 futhin: why? it does what it's good at well 01:43:15 one layer between the computer & user 01:43:17 the only thing I might add would be registers or something 01:43:19 simpler, better 01:43:29 but I don't know it well enough to say, really 01:43:29 registers? 01:43:38 something to put local variables into 01:43:41 it's not really an assembler 01:43:54 local variables can go on the stack 01:44:03 the trick to coding in forth 01:44:06 is factoring up your code 01:44:10 having small words 01:44:27 pass minimum 2 parameters to a word 01:44:49 --- quit: davidw (Read error: 110 (Connection timed out)) 01:45:03 yes, but then getting them on and off the stack is a pain in the ass 01:45:03 swap this, do that swap it back again, rot to get another one, and so on 01:45:04 I'd rather just put it in a register and say 'hold this one moment' 01:45:08 a word should only be about 1 or 2 lines of code 01:45:12 --- nick: davidw_ -> davidw 01:46:05 yeah, the code isn't set up well if you find yourself having to use ROT and PICK all the time.. OVER is sort of ok 01:47:22 it's hard to get comfortable making small words.. but after that, the coding is a lot more easier, it's easier to debug, and then ignore that word as you focus on the next, etc.. 01:47:31 and easier to change 01:48:47 forth kind of forces you to design the components and data flow 01:49:33 which sucks if you want to do 'high level' and 'abstract' 01:49:37 it might help if you read Starting Forth and/or Thinking Forth by Leo Brodie 01:49:48 you can buy Thinking Forth by Leo Brodie off amazon.com 01:49:53 yeah? hrm 01:49:59 i recommend the book, it's pretty good 01:50:13 Starting Forth discusses the language, coding forth, etc.. 01:50:23 Thinking Forth is more about the methodology 01:50:26 the design, etc 01:50:46 Thinking Forth is about software design 01:50:54 non-forth coders can read it 01:51:20 it is somewhat specific to forth, but still does a decent job of covering software design, etc 01:51:21 the language is pretty easy, I don't need much of a manual 01:51:43 I need to know how to useit 01:52:08 did you get the Forth Programmer's handbook.pdf ? 01:52:19 i haven't actually looked at the book 01:52:39 but i assume it does a good job of explaining forth & its use 01:52:50 hi 01:53:00 heya onetom :) 01:53:06 good morning and all :) 01:53:11 i'm about to go to bed soon 01:53:17 I couldn't extract the windows archive:-/ 01:53:24 it was an .exe, which I tried to unzip, but it didn't work 01:53:27 care to send it to me? 01:53:44 i'll put it on onetom's comp and give you the url :) 01:54:05 --- quit: davidw (Remote closed the connection) 01:54:06 --- join: davidw_ (~davidw@ppp-179-19.25-151.libero.it) joined #forth 01:54:57 --- nick: aum|dinner -> aum|partly-here 01:55:24 i should explain the OO forth concept as i see it 01:55:47 8-byte cells - each is class id then class instance 01:56:10 --- nick: davidw_ -> davidw 01:56:13 each word which operates on the stack needs to look up the class ids of the operands 01:56:19 and act accordingly 01:57:13 if I had the money, I would have absolutely a shitload of books 01:57:30 my other gripe with forth is that the only delimiters are whitespaces 01:57:42 I like that feature, actually 01:57:54 it makes for less readable code 01:58:21 i'd like to be able to have a 'delimiter' flag on words 01:58:26 I don't know... I think having all kinds of operators like ; @ ! and so forth doesn't help much either 01:58:57 a word, 'isdelimiter', could turn the last-defined word into a delimiter 01:59:38 and if there's any ambiguity, precedence could go to the longest delimiter 02:00:09 sounds hectic 02:00:14 just write clean code 02:00:18 so if '+' and '+!' are delimiters, the parser would see '+!' and know to return the word '+!' instead of '+' then '!' 02:00:24 use newlines where appropriate 02:00:46 group things logically...use indentation 02:00:52 i do that already 02:00:55 no, that's the least of my worries 02:01:06 but forth code tends to look like food put through a blender 02:01:10 yes, i was thinking about the problem of spaces a few days ago, but i forget my conclusions :P 02:01:50 it sucks.. i had a good idea.. 02:01:59 but i'm really warming to the idea of a forth where *everything* is an object 02:02:16 --- join: Stepan (~stepan@p5084778B.dip.t-dialin.net) joined #forth 02:02:21 moin 02:02:53 yous'es 02:02:58 good morning, we're discussing a forth where everything is an object 02:03:01 --- nick: aum|partly-here -> aum 02:03:18 aum's idea 02:03:37 well, it would take away the need for a lot of assembler-like byte-bashing 02:04:18 oh.. sounds interesting. 02:04:26 lie C++ is to C? 02:04:31 it would be the responsibility of all the operators to determine what class type the operands are, and to act accordingly 02:04:46 Stepan: yes - F++ :) 02:05:11 the word '+' would be an OO addition operator 02:05:31 there'd be a small secondary stack for raw ints 02:05:44 the word '_+' would add two numbers on this small stack 02:06:18 an int literal, eg '3', would push a cell containing 02:06:25 where contains the value 3 02:06:41 so you could do '4 5 +' and expect it to work as normal 02:06:54 also, the '.' operator would be able to output strings 02:07:36 as new classes are defined, extension methods could be added which extend the functionality of existing operators 02:08:20 for example, if there was a class 'mystruct', the '.' operator could get extended with code to dump out the contents of an instance of 'mystruct' 02:08:52 the dilemna is whether to build inheritance into the language 02:09:06 or to have just a flat set of classes 02:13:29 probably it should work 4 ya 02:13:41 oops 02:19:24 it would be slow, though 02:24:27 aum: i'm interested in extending forth to a higher-level, and gaining more support for forth 02:24:33 not sure if the OO model is the way to go 02:24:39 but being open minded 02:24:48 but its a possible way 02:24:55 exactly 02:25:17 so c++ and obj.pascal users can immediately use a 4th os 02:25:27 w this slight "extensiion"? 02:25:39 i really look to MUF as a great example of what forth can be like, and taken further.. an easy scripting language that anybody, including my grandma can learn! 02:26:00 muff? 02:26:01 MUF isn't _that_ easy.. but it's an example 02:26:28 MUF = multi-user forth. a forth-like language used as the online programming language for TinyMUCK (which is a MUD) 02:26:36 do you know what MUD is? :) 02:26:38 yeah, ive also took a glimpse on muf & i liked it 02:26:42 it's weird explaining programming to people who haven't ever done anything 02:26:47 thx 4 mentioning it, futhin 02:27:11 i showed you my muf code :) 02:27:16 I remember trying to explain to this guy what a 'return value' was 02:27:41 the help system and the debugger that came with MUF was _great_ 02:27:58 i picked up MUF in no time at all, it's got a really simple setup.. 02:28:30 like, picked it up in 15 minutes, and it was my first encounter with "forth" 02:28:58 i was coding in muf within 15 minutes of seeing it.. debugging my code effectively, etc.. 02:29:06 coding simple stuff, but coding regardless 02:29:45 i think that forth has the potential to be extended into the best scripting language :P 02:30:18 hm 02:30:22 i'm going to go to bed 02:30:24 good night all 02:30:29 * futhin sleeps 02:31:12 That's not a bad idea 02:31:15 --- nick: Soap` -> SoapSleep 02:44:49 --- join: rob_ert (~robert@h237n2fls31o965.telia.com) joined #forth 02:45:25 another idea in my forth++ concept - garbage collection 02:46:08 so if anything gets popped off the stack, its class instance is destroyed unless code specifies otherwise 02:47:35 onetom: so how would you write that 'shift' operation? 03:16:45 --- part: aum left #forth 06:14:25 --- join: tathi (~tathi@wsip68-15-54-54.ri.ri.cox.net) joined #forth 06:24:16 --- join: oxygene (~oxygene@xdsl-213-168-119-200.netcologne.de) joined #forth 06:50:00 --- quit: davidw (Read error: 113 (No route to host)) 07:02:59 --- join: davidw (~davidw@ppp-217-7.25-151.libero.it) joined #forth 09:15:29 --- join: slick_rid (~user7168@bgp413575bgs.medwln01.nj.comcast.net) joined #forth 09:15:39 ji 09:15:43 hi 09:15:47 ; 09:15:47 ; 09:15:47 ; 09:15:47 ; 09:15:48 ;; 09:15:49 ;;; 09:15:49 ;; 09:15:51 ;; 09:15:53 --- part: slick_rid left #forth 09:28:40 --- join: TheBlueWizard (~tbw@ip-216-25-202-35.vienna.va.fcc.net) joined #forth 09:28:40 --- mode: ChanServ set +o TheBlueWizard 09:28:55 hiya all...just stop by to see who's on, then ta-ta... 09:29:10 hmm...gotta go...bye all 09:29:13 --- part: TheBlueWizard left #forth 09:46:25 fuck this, I'm going to do the whole thing in C++ 09:46:31 --- part: davidw left #forth 10:33:05 goodmorn everyone 10:42:04 morning :) 10:42:49 what are you doing with forth these days? still coding a forth in c? 10:43:19 Who? Me? 10:43:33 yes 10:48:14 hi futhin 11:00:32 howdy onetom 11:00:53 brb, shower 11:03:16 ez melle 11:03:23 shit :))) 11:03:28 im still a bit sleepy 11:03:34 :) 11:03:53 i wanna tell: im still sleepy 11:04:04 but told on the bad channel 11:04:36 then i said oops in hungarian (ez melle) there 11:04:49 switched 2 here 11:05:08 pressed the up key and then enter... 11:05:35 well, yes, ive also proved it. im really sleepy yet :) 11:06:17 :) what time is it there? 11:07:25 its just 20:15 11:07:43 but i took a short (~2hrs) nap 11:08:05 ah 11:08:07 & im not totally awake yet 11:08:13 that will do it :) 11:09:40 but hows yr day, tathi 11:09:41 ? 11:10:14 what network card do u have? 11:12:44 my day is going pretty well... 11:13:47 I'm currently reading through the isforth sources 11:14:09 and, how do u like it? 11:14:23 r u also familiar w x86 instruction set? 11:14:29 somewhat 11:14:51 haven't done any x86 asm in probably 5 years though 11:15:32 well written though -- the most readable piece of code by someone else that I've seen in a while :) 11:16:24 oh, what a praise :) 11:17:40 but then, most of what I've been reading lately has been the linux kernel sources, which are fairly awful, if you ask me 11:18:26 & what about bsd's? 11:18:46 haven't looked at bsd at all... 11:18:49 bsds' 11:19:05 ppl say those r pretty nice 11:19:49 and netbsd should also run on yr box 11:20:03 coz it really runs on everything 11:20:29 yeah, I hadn't thought of that, I'll have to look into it. 11:20:33 what has @ least a handful of pins 11:21:02 u get a fully posix os w netbsd 11:21:22 and all the bells & whistles of 11:21:40 mostly the latest development of the linux world 11:22:03 so u must get gnome & kde w it 2 11:22:36 tho, its a bit handdriven, its much more clean 11:22:50 than any linux distros 11:23:14 I've been pretty much building my own system 11:24:12 how much papers have u read from chuck? 11:24:12 might install bsd at work though 11:24:37 not a lot... mainly just the stuff on his colorforth page 11:24:40 have u seen the videos about his lectures? 11:24:50 I did watch one or two of the videos 11:24:51 on www.ultratech.com? 11:24:57 yeah 11:25:43 thats great. 11:26:19 u have some advantage, then :) 11:26:45 but lets get back 2 the network card 11:26:58 when do u plan 2 implement networking code? 11:27:22 is there any tcp/ip reference code already in 4th? 11:28:06 don't know. OF must support it to some extent, as you can get an executable over TFTP to boot from 11:28:40 I'm really just starting to get into hardware stuff 11:29:17 Most of what I've done so far is deciding how much of Chuck's colorforth stuff I like and how I want the basic engine to work 11:31:05 networking is one of the first things I want to get working though 11:31:50 aha 11:32:10 has u already assembled a prj implementation plan? 11:32:56 haas?!!!?!! have 11:33:15 working on it...I keep getting distracted by finding new interesting unix libraries and stuff :) 11:33:17 i cant speak good :) 11:33:25 loll 11:33:45 k 11:34:03 no i gotta go do some work 11:34:18 k 11:34:21 but then i come back and finish 11:34:28 my kernel.awk 11:34:44 then i also implement it in pascal 11:35:37 or probably step fwd 1 & i create a bit more complete kernel.4th (v0.1) 11:36:14 sounds like a plan 11:44:47 yup 11:48:00 --- join: herkamire (~jason@wsip68-15-54-54.ri.ri.cox.net) joined #forth 11:49:25 but u still havent told me what is the chip of the network cards in G4s by default? 11:49:38 or what do u have actually 11:49:43 hi herkamire 11:50:41 onetom: The more I think about your idea (a standard forth environment on very platform on which to build an OS in forth and higher level languages) the more I like it 11:50:48 ah. Don't remember off the top of my head what chip it is. 11:51:10 my remaining nagging doubt is whether anybody will implement it 11:51:55 but I say start on one platform :) 11:52:39 herkamire: hm, happy to hear it :) 11:52:56 I still don't know about supporting wildly different hardware with the same code though. 11:53:09 herkamire: tho, i know it will b a very restrictive subset of the 4th 11:53:28 I really want my cells to be 32 bits, and a lot of processors (handhelds etc) don't support that. 11:53:48 herkamire: coz ppl r not good @ agreement, BUT! 11:54:10 they can code the extra bits they want, so long as you lay a solid foundation 11:54:26 herkamire: we will write a set of 4th implementations in 4th, so 11:54:51 herkamire: any of those could run on top of eachother 11:54:59 cool 11:55:36 coz i plan 2 implement many imaginable 4th engines i 4th 11:55:49 so ppl could pick up 1 and use it 11:55:56 cool 11:56:07 on top of any other 4th coz 11:56:12 onetom: I'm getting Vendor 106b, Device 0021 from the PCI 11:56:28 which linux's pci.ids file says is Apple, GMAC ethernet (Sun GEM) 11:56:37 i plan 2 implement those by using a very small subset of 4th words 11:56:46 sorry, UniNorth GMAC ethernet 11:56:54 tathi: ah, thx 11:57:25 well, it doesnt seem a very std 1 from the pc world 11:57:41 but it interfaces through pci @ least :) 11:57:58 it's Gigabit ethernet... 11:58:12 --- nick: SoapSleep -> Soap` 11:58:21 woooahohh 11:58:27 hehe 11:58:38 gigabit ethernet is another story :) 11:58:46 that's why macs are expensive :) they have good parts :) 11:59:04 oy c oy c 11:59:38 but linux have a driver 4 it 11:59:45 could u post it 2 me? 11:59:45 yes 11:59:53 the driver? 12:00:06 u must have the kernel srces installed, iguess 12:00:11 sure the driver 12:00:24 ok 12:00:38 nah, what did u think i thought about? the card itself? :) 12:07:05 I'm going outside for a while 12:07:25 * herkamire is away: [frisbee] 12:11:25 hm, the driver seems quite clean 12:11:36 tho, its a not a small piece of sw :-/ 12:11:57 and it also requires irqs & dmas inited well 12:12:49 yeah 12:13:19 i think we should start w a simpler thing, like 12:13:23 ne2k-pci 12:14:00 the basic ne driver 4 isa cards could also b a bit complicated... 12:18:13 so do you have some kind of implementation plan yet? 12:22:19 --- join: oxygene_ (~oxygene@xdsl-213-168-108-225.netcologne.de) joined #forth 12:23:05 --- quit: oxygene (Read error: 104 (Connection reset by peer)) 12:34:25 --- join: CaffeineJunkie (~icafe@195.30.184.53) joined #forth 12:34:50 g'd day 12:35:37 heya 12:38:55 --- nick: oxygene_ -> oxygene 12:49:59 --- nick: CaffeineJunkie -> Speuler_ 12:50:10 --- part: oxygene left #forth 12:53:55 --- part: Speuler_ left #forth 12:57:48 --- join: Soap- (flop@210-54-106-168.dialup.xtra.co.nz) joined #forth 13:07:14 --- quit: tathi ("Client Exiting") 13:12:42 * herkamire is back (gone 01:05:17) 13:17:12 --- quit: Soap` (Read error: 113 (No route to host)) 13:40:21 onetom & futhin: are you thinking of a 32 bit forth, or 16? (8??) 13:50:19 herkamire: im thinking in all of them actually 13:51:00 but they might b implemented separately i think 13:52:32 im writing code for a 8bit & a 32bit 4th simulteniously also @ the very moment 13:56:15 coz my building control program should b able 2 run on pic micros w 8bit alu 13:58:33 or probably later on larger pic w 16bit alu and w a writable prg mem 14:00:02 but momentarily a pc does the main control job 14:00:44 but my code just runs perfect on it 14:00:56 tho, its not optimal certainly 14:01:41 it does unnecessary dbl2single and singl2dbl conversions 14:02:39 but it doesnt matter. the pc has enough computation power 4 it 14:20:50 wow. 14:21:02 I'm off for a bit 14:21:05 --- quit: herkamire ("Client Exiting") 14:22:38 --- join: Etaoin (~david@ljk2-14.sat.net) joined #forth 14:23:31 helo-belo 14:24:08 hello 14:25:10 --- join: St3pan (~stepan@pD9E5328D.dip.t-dialin.net) joined #forth 14:27:39 hi 14:39:26 --- quit: Stepan (Read error: 110 (Connection timed out)) 15:07:02 --- join: I440r (~mark4@1Cust41.tnt3.bloomington.in.da.uu.net) joined #forth 15:07:23 hi ppl 15:07:28 hello 15:07:29 bongo :) 15:08:35 hi!! 15:08:46 been working on the n^2-1 puzzle solver 15:08:55 got about half of its brain coded so far heh 15:09:05 the next part is actually the easy part i think :) 15:13:28 half of its brain? 15:13:41 the left side or the right side? ;) 15:18:05 or half of both? 15:18:20 heh 15:18:23 not sure 15:18:51 it can build the search tree now, automatically prunes positions it has already encountered from the tree 15:19:15 theres one small part of the evaluation funciton to be coded yet 15:19:21 and then the tree search 15:19:35 aha 15:19:38 but just BUILDING the tree does half the work of searching it :) 15:20:34 do u make an in-deepness search or an in-wideness 15:21:26 (just a test wheather u know these terms >:) 15:22:08 (certainly i dont even know what is the right form of these things in english, so... :) 15:23:04 sort of half ahd half heh 15:23:16 i do an exhaustive search of the entire tree up to ply 6 from the root node 15:23:29 then i expand the best node to +6 ply 15:23:36 and repeat 15:40:10 --- join: herkamire (~jason@ip68-9-58-81.ri.ri.cox.net) joined #forth 15:42:58 hi 15:43:13 hey :) 16:30:44 --- join: MrReach (~mrreach@209.181.43.190) joined #forth 16:31:53 mrreach :) 16:32:00 hihi 16:32:01 dood - do you have a binary search word in forth ? 16:32:10 nope 16:32:22 grr i prolly have to write one then eventually heh 16:32:37 there should be several floating around out on the internet 16:32:38 i can get by with a sequential search because its a sequential search of hashed data :) 16:32:50 i jst started looking for one now :) 16:33:21 I'm pondering how objects should uniquely itentify themselves across machines 16:34:42 by having an identidy entry in their structure ? 16:34:50 got tcl/tk runing on the Jornada, finally, not that anyone here would care 16:34:58 yes, of course 16:35:35 but there's three machines involved, each running a different OS and each having different network connection characteristics 16:36:17 an object created on one machine must never conflict in id with an object created on another machine 16:38:04 large random numbers come to mind :) 16:38:17 use tunneling ??? :) 16:39:45 or assign each computer an ID, and then computers only make objects that satisfy objID mod MAX_COMPUTERS = computer_ID 16:40:35 or, more simply put, make the computer ID part of the object ID. 16:40:37 yeah, I think I'm going to do something like that 16:40:57 I think one of the machines will become "The Authoritative Server" 16:41:24 who will arbitrate that ? 16:41:27 the computers ? 16:42:23 naw 16:42:43 I will ... my most reliable machine ... which would be my linux box 16:43:33 whenever one of my software components connects to the server, it identifies itself with its unique machine ID 16:43:50 if it doesn't have a machine ID, then it is assigned one by the server 16:44:27 your network is all trusted computers that you set up? 16:44:44 if so you can assign them ids manually 16:44:51 then, objects are numbered sequentially from 0 and prefixed with the machine ID 16:45:16 everything this side of the router is trusted 16:45:41 but there WILL be an open port that will allow me to connect with the laptop from a dialup. 16:47:17 what are you building? 16:47:30 PDA software 16:48:00 there's a Linux box, a windows98 desktop, and a Linux network server 16:48:19 PDA as in palm? 16:48:21 the central repository will be on the linux box 16:48:40 address books and such? 16:48:42 and clients on the desktop and the WinCE Jornada 16:48:46 yes, exactly 16:48:53 cool 16:49:02 well i gtg eat - and code some :) 16:49:03 the software that came with it is *COMPLETELY* lame 16:49:08 I've been trying to think of something cool to do with my addressbook. 16:49:09 be well, I440r 16:49:13 i shud have the puzzle brains coded soon :) 16:49:24 herkamire: what is it? 16:49:30 cu l8er! 16:49:35 --- quit: I440r ("Reality Strikes Again") 16:49:40 I converted my addressbook to xml by hand, 16:49:46 but I'm loosing interest in XML 16:49:58 what is your address book? 16:50:28 an XML file with my friends and other contacts. email addresses directions, IM nicknames, notes etc... 16:50:28 are you talking portable computer? 16:50:35 no 16:50:38 ah! ok 16:51:08 it was originally from my palm (visor) but that broke, so I grabbed the data out of the Palm Desktop software 16:51:30 are you going to replace the palm? 16:51:36 do you know what with? 16:51:41 not now. maybe later 16:51:48 I don't have enough use for it at the moment 16:52:10 I would like to get a linux or forth handheld. 16:52:10 * MrReach nods. 16:52:22 (my schedule is so simple right now...) 16:52:29 I *COULD* install linux on this thing, I think 16:52:50 but that seems a very risky operation ... and this OS works pretty good 16:52:51