00:00:00 --- log: started forth/01.11.04 00:09:58 * plbofh is back (gone 07:25:32) 00:25:37 * plbofh is away: zzzzz.... 05:57:05 --- quit: nate37 (zelazny.openprojects.net benford.openprojects.net) 05:57:09 --- join: nate37 (nate@cx83983-d.irvn1.occa.home.com) joined #forth 05:57:09 --- mode: benford.openprojects.net set mode: +o nate37 05:57:12 --- topic: set to '' by ChanServ 05:57:50 --- quit: nate37 (farmer.openprojects.net benford.openprojects.net) 05:58:07 --- join: nate37 (nate@cx83983-d.irvn1.occa.home.com) joined #forth 05:58:32 --- mode: ChanServ set mode: -o MrReach 08:50:11 --- join: speuler (l@passionsfrucht.icafe.spacenet.de) joined #forth 08:50:25 g'day 08:50:40 hihi! *YAWN* 09:16:38 hi mrreach 09:16:51 * MrReach nods. 09:17:38 how do you like your new subnotebook ? 09:17:57 it's ok so far ... have yet to use it extensively 09:18:13 you're right about the WinCE interface being awkward 09:18:26 you run v2 or v3 ? 09:18:32 v3 09:18:46 hmm - that's supposed to be a huge improvement 09:18:47 was hella hard to find where that info was 09:20:05 graphics draw is pretty snappy 09:20:31 the way it uses compactflash is pretty clunky, though 09:20:51 in two or three years, that kind of machine will probably outperform current desktop machines 09:21:00 found a forth for it, but can't get the console yet 09:21:26 runs in window ? 09:21:28 heh, I bought a 128MB flash card, so my PDA has 8 MB more ram than my desktop 09:21:36 yes, it does 09:21:59 when i upgraded from 32 to 64 meg, i found it 2b a big difference. when going to 128 mb, i was still surprised 09:22:12 the forth WANTS to parse an input file and send output to another file 09:22:30 recently i have upgraded from 128 to 512 meg, and it is still more of a change than i expected 09:22:37 really? 09:22:46 harddisk fell quiet 09:22:54 no more head rattling 09:22:54 I'd have to look pretty hard to find a decent price for a 512MB flash card 09:23:09 oh, your desktop machine? 09:23:14 indeed. pricy stuff 09:23:18 desktop 09:23:29 RAM 09:23:35 ok, mem for the desktop is pretty cheap 09:23:47 i thought you were comparing against RAM on your desktop 09:24:17 heh, I evaluate a lot of software ... much of it installs modules that are not removed when the software is uninstalled 09:24:29 so it boots slower and slower over the years 09:24:35 (still running a 400 mhz cpu on desktop) 09:24:51 yeah, mines a k6-2/350 09:24:58 w/ 128mb 09:25:17 it's a good machine, still performs well 09:25:48 mine K6 too. no problems. 09:26:09 heh, is your a Compaq Presario? 09:26:14 K6/400 are not very popular with windows users, as this seems to be a problem-prone cpu with windows 09:26:25 but linux runs just fine with it 09:26:30 (prob too much to ask for @:^) 09:26:59 no compaq. machine has been assembled and upgraded through the years 09:27:07 TerraComm's servers were k6-2/400 128mb ram 09:27:07 the case is still original 09:27:17 no ATX yet :) 09:27:21 heh 09:27:30 started as 386/33 09:27:36 yikes! 09:27:46 seen 4 motherboards since 09:28:03 have to do any hacking at the case? 09:28:53 slightly, used a saw to provide space for another drive 09:29:22 there are 5 scsi and 1 ide drives in 09:29:38 nice 09:29:39 could add another drive :) 09:29:55 i kept the case because it is a quite nice one 09:30:13 between big-tower and midi-tower, but somewhat wider 09:30:20 makes it look quite bully 09:31:36 heh 09:31:58 I've been thinking of scrapping my server and going to new technology 09:32:14 it's a pentium/133mhz/48mb 09:33:19 48 mb is not a lot 09:33:26 not anymore ... 09:33:35 start with a full-size tower case ... can get one with 2 extra case fans for $100, ATZ 09:33:38 ATX 09:33:45 it's running linux 09:33:55 so 48 serves "ok" 09:34:34 a new one would start at 128mb, and dual channel scsi-160 09:35:05 scsi is a good thing to go for 09:35:15 but not the cheapest decision 09:35:28 no, about 1/2 again as much 09:35:42 but doesn't quit, and performs so much better 09:36:17 TerraComm had major probs with the IDE CDRW 09:36:20 the longer you postpone the decision to change to scsi, as more expensive it gets replacing the devices you bought already 09:36:39 huh? 09:36:46 you're saying don't wait? 09:36:47 the non-scsi, i mean 09:37:02 oh, ok 09:37:21 i think one ide device per ide channel is ok 09:37:35 the old ide will run just fine in parallel, but use for slower data 09:37:39 but don't start hooking up master + slave 09:38:11 yes, my policy is one fast and one slow device per channel. 09:40:07 right now, ANY drive is far faster than my dinky server 09:41:14 you can make win people panick if you burn a cd for them, and continue to use the system heavily 09:41:32 like, compressing several mp3s while compiling a kernel :) 09:41:54 try that with an ide writer under windows :) 09:41:57 HAHA! 09:42:27 we did ... didn't work at all 09:42:36 they should build a red LED into the windows key, saying "don't touch me" 09:42:52 had to be careful with the same IDE setup on Linux 09:43:03 better, build them into any key 09:43:25 the modern ide writers should be much better 09:43:35 * MrReach nods. 09:43:38 but i don't speak from experience 09:44:38 --- join: futhin (thin@h24-66-209-114.cg.shawcable.net) joined #forth 09:44:45 greets, futhin 09:44:54 greetings all 09:45:42 is plbofh a bot or something? 09:46:00 talkitive bot, if he is 09:46:21 Speuler: public class HW { public static void main( String [] args ) { System.out.println("Hello world"); } } 09:48:12 hi futh 09:48:18 hi futhin 09:48:36 so what's up ? 09:48:50 we should host Forth CON and have lots of booze! 09:49:08 ForthCon 09:49:18 forthbooze ? 09:49:25 and we'll have t-shirts that say "i went to ForthCon, and all i got was this lousy t-shirt!" 09:50:07 begin standing while booze integrate again 09:50:17 rpeat ... 09:50:24 we'll have the ForthCon in calgary ;) 09:50:55 or maybe in las vegas (cheaper in flights) 09:51:28 there will be EuroForth soon. rumours say charles moore is going to attend 09:52:34 hmm 09:53:00 in less then 3 weeks it should start 09:53:05 bah 09:53:11 i guess i'm gonna miss it :( 09:53:25 takes place in schloss dagstuhl 09:53:34 are you going? 09:54:00 not sure. quite busy, can't afford to miss a week 09:54:14 but i'll probably be mentioned there :) 09:54:43 heh.. why will you be mentioned? 09:55:06 because of some recent work which will be presented 09:55:14 in embedded stuff or? 09:56:04 http://www.cornu.purespace.de/forth/ 09:56:25 http://www.cornu.purespace.de/forth/oodc.html 09:56:48 A different implementation of the same interface (OODC) was made with strong contributions from Lothar Schmidt. It will be presented on 2001 euroForth conference. 09:57:37 OOP in Forth should be prefix or something 09:57:55 and it shouldn't look so ugly heh 09:59:16 hmm 09:59:22 OODC looks ok i guess 09:59:39 Odissey should be spelt Odessey 09:59:45 hmmmm 10:00:05 er 10:00:07 odyseey 10:00:08 blah 10:00:11 Odyssey 10:00:40 spelt is somethimes spelt spelled 10:01:08 *grin* 10:01:10 spelled 10:01:11 hmm 10:01:17 i dunno where i come up with spelt :P 10:02:01 Pronunciation: 'spelt 10:02:01 chiefly British past and past participle ofSPELL 10:02:15 so it's a british word or something 10:04:55 Odessey is the name of a greek ship, if I recall, written about by Homer 10:05:05 or was that just the name of the story? 10:06:40 nowadays: oddisey for a serching quest 10:08:38 oddysee 10:09:04 --- nick: MrReach -> MrGone 10:11:07 no.. it's spelled odyssey.. a searching quest 10:11:11 or journey 10:15:27 i want to talk forth 10:15:38 but i can't quite think of any topics to focus on 10:18:57 * speuler cleans up his hard disk 10:37:56 delete your 0-day pr0n 10:38:07 hmm 10:38:16 nevermind 10:56:44 --- quit: speuler (Ping timeout for speuler[passionsfrucht.icafe.spacenet.de]) 10:58:56 --- quit: futhin () 12:03:09 --- join: speuler (l@passionsfrucht.icafe.spacenet.de) joined #forth 12:03:34 --- nick: speuler -> Speuler 13:10:16 --- join: I440r (mark4@A010-0100.BLMG.splitrock.net) joined #forth 13:10:16 --- mode: ChanServ set mode: +o I440r 13:10:29 bongo! 13:10:31 --- mode: I440r set mode: +o clog 13:10:34 --- mode: I440r set mode: +o MrGone 13:10:39 --- mode: I440r set mode: +o nate37 13:10:46 --- mode: I440r set mode: +o plbofh 13:10:50 --- mode: I440r set mode: +o Speuler 13:48:39 g'day 13:48:50 thanks you 13:49:07 i440r ! 13:49:15 you know nasm pretty well, don't you ? 13:50:25 is it right that i can't do & for bitwise and with $ assembly location 13:50:49 hi :) 13:51:01 hmmm 13:51:11 i believe that is correct 13:51:26 i THINK 13:51:39 the problem is that the & is done bya pre processor 13:51:40 how to you align then ? 13:51:44 and $ is unknown to it 13:51:55 with the even directive ? 13:52:12 i think 13:52:20 it respects the size set with BITS ? 13:52:32 erm 13:52:34 i dont know 13:52:51 times (4- ($ & 3)) & 3 db 0 13:52:56 this win't work ... 13:52:56 actually there are alot of issues i have with nasm = mostly due to the pre processor - its a fucked up design 13:53:01 totally fucked up 13:55:30 --- join: qless (qless@clgr000977.hs.telusplanet.net) joined #forth 13:55:35 wont ? 13:55:42 what are you trying to do ? 13:57:49 writing a macro with puts a count byte of string argument first, then the string, and then aligns 13:57:55 which ... 13:59:49 i wanted that :) 14:00:06 i think it can be done with 14:00:28 db wordname_count,"word name goes here" 14:00:40 wait 14:00:44 wordname: 14:01:02 db wordname_count,"blah blah" 14:01:18 wordname_count = $-wordname 14:01:38 you pass the label and the name string to the macro and it creates the lable 14:01:54 the count byte and the body of the header 14:01:55 BUT 14:02:05 you have problems when you want to make words immediate etc 14:02:42 i think tcn figured a way to do it in nasm - i can send you his sources for isforth with his macros in it 14:02:51 ive not looked at it in ages tho 14:05:35 i got that part working, with the help of eks 14:05:41 now i just need to align somehow 14:05:49 why ? 14:05:51 heh 14:06:07 its either the even driective or align n directive 14:06:09 i forget which 14:06:14 align 2 14:06:16 align 4 14:06:17 etc 14:06:19 word immediate is just setting an additional bit 14:06:25 yes 14:06:28 usually bit 7 14:06:33 can be done by adding a constant, or oring it 14:06:36 but you have immediate and smudge bits 14:06:46 imm equ 64 or something like that 14:06:49 should be easy 14:06:50 yea 14:06:54 hopefullz 14:07:02 ur rite 0x40 14:07:05 but string don't need to be immediate :) 14:07:07 0x80 is smudge 14:07:21 depends on implementation 14:07:21 --- mode: I440r set mode: +o qless 14:07:27 yea 14:07:32 it is the implementors choice how he uses bits in header 14:07:35 some forths dont smudge 14:07:40 or whether they are in the header at all 14:07:44 xactly 14:08:00 i used additional header bits too several times 14:08:11 and the header structure is going to be different. 14:08:17 seperated from bodies 14:08:24 in different address space 14:08:29 so i can move them around 14:08:37 or throw away any 14:08:54 will have a pointer from header to body 14:08:59 not the other way around 14:09:23 need a name ? i'd have to search headers until i found a match with body pointer 14:09:26 i like seperate headers 14:09:29 easy to turnkey 14:09:30 hardly ever used 14:09:42 except for disassembly, and the like 14:09:50 no problem if name lookup is slow 14:10:09 seperate headers makes searching easier because you dont need any links at all 14:10:12 as long i can get quickly from header to body 14:10:18 header to header links i mean 14:10:22 xactly 14:10:34 count byte can lead to next header 14:10:39 no link field 14:10:45 no traverse 14:10:55 thats what i was trying to do with isforth but tcn ran into problems with seperate head space and had to interleave headers and code 14:10:58 and freedom of header structure 14:11:05 exactly 14:11:11 and that was another problem with nasm 14:11:13 you cant do 14:11:15 i've done that several times 14:11:20 db "xyzzy"+0x80 14:11:36 because of need to organize headers hi-to-low addresses ? 14:11:39 becayse "xyzzy" is type string and 0x80 is type integer 14:12:00 every fucking assembler i ever used would add 0x80 to the last byte 14:12:09 nasm is a piece of shit basically 14:12:16 but its the best available 14:12:23 countbytelocation: db %%countbyte 14:12:26 db %1 14:12:35 i tried to get the author of a86 to do a linux version but he refused :P 14:12:40 he lives 15 miles from me :) 14:12:46 %%countbyte equ $-countbytelocation+1+imm 14:13:18 that part seems to work 14:13:29 i'm just fiddling with alignment after string 14:13:54 wanted to do it myself, but doing bitwise operation on $ wasn't successful 14:14:07 correct 14:14:08 bleh 14:14:14 try x = $ 14:14:21 x &=1 heh 14:14:23 er no 14:14:24 you cant 14:14:26 align 4 works 14:14:37 strings work 14:14:38 you need xset x=$ BUT 14:14:54 you cant use set x = $ 14:15:03 becayse $ is UNKNOWN to the preprocessor 14:15:46 xset x $ 14:15:57 xset x x & 1 14:16:00 etc 14:16:02 oops. can put 1 to 3 chars in string. bombs on 4 or more ... 14:16:09 yup 14:16:13 doesnt supprise me 14:16:20 i had problems like that 14:16:56 could be a problem with stringlit 14:17:08 hope so 14:17:32 %macro header 2 14:17:33 ; section .data 14:17:33 %2_link: ;create a symbol for hew header location 14:17:33 %%link: 14:17:33 dd link ;LINK to previous word in dictionary 14:17:33 %xdefine link %%link ;remember current link symbol name 14:17:35 db (%%name-$-1)+080h+imm ;LENGTH + flags 14:17:37 db %1 ;NAME 14:17:39 %%name: 14:17:41 %xdefine imm 0 14:17:43 %endmacro 14:17:45 thats tcn's version of my header macro 14:18:59 you you have to run the immediate macro every time you want to define an immediate word tho 14:19:10 immediate macro sets the imm equate 14:19:22 which gets cleared every header 14:22:57 what's the state of isforth btw 14:23:59 i consider throwing out immediate words, and adopt charles moore's idea of having a compiler vocabulary 14:24:23 i also wanted a compiler vocabulary 14:24:24 but 14:24:27 a voc which is only searched during compilation, but before the other vocs 14:24:32 i consider it more of a "creating word" vocabulary 14:24:50 might call it "immediate" 14:24:53 because that whole vocab is DEAD WOOD in a turnkeyd app 14:25:04 isolate it, and prune it 14:25:09 correct 14:25:19 i had this planned for isforth right from the start 14:25:33 did you implement it ? 14:26:01 yes and no 14:26:11 but 14:26:33 --- nick: MrGone -> MrReach 14:26:33 im using nasm still and its very difficult to have macros that link to more than one vicabulary 14:27:04 Speuler: I didn't know you were currently in the process of writing your forth ... are you looking for participants? 14:27:09 i was trying to modify my header macros to link on the CURRENT voc at assemble time 14:27:12 greetings, I440r 14:27:22 hi dood :) 14:27:32 I've only got a few seconds here, but will read the log carefully 14:27:46 it assembles ok right now, and executes a some primitive and high-level words 14:27:47 not much said :) 14:28:16 it says "123* xdfsdfsdsdsfdf ok 14:28:24 I disagree, I see much said in my backscroll ... interesting to me, anyway 14:28:36 :) 14:28:40 but it is in a very early state 14:28:46 well -alot more than im usually here for hehe 14:28:52 about 400 lines 14:29:00 with lots of empty space among 14:29:09 Speuler: I didn't know you were currently in the process of writing your forth ... are you looking for participants? 14:29:26 it does branches allright right now 14:29:39 there may be a bug in stringliteral 14:29:46 or maybe it is a nasm bug 14:29:54 or not complete understanding of nasm macros 14:30:28 it is 4168 bytes now 14:30:43 that's a lot 14:30:56 after all, i had a working forth in half that space before 14:31:09 not on this size of machine 14:31:12 and this is merely the first signs of coming alive 14:31:22 on these machines ... 14:31:37 size is much less an issue, and organization a much greater issue 14:31:45 there's no inner interpreter, no vocabulary, no string compare in it right now 14:31:50 no numbers 14:31:59 just some primitives and hi-level words 14:32:01 Speuler: do you want help? 14:32:07 glued together for testing 14:32:25 you'd like to help i make from it ? 14:32:33 yes 14:32:34 it is a fun project, beware 14:32:40 will not argue about threading method 14:32:50 and may get experimental here and there 14:32:53 but will ask that it meta-compile itself later 14:33:00 good plan 14:33:03 like it 14:33:16 keep experimental parts as load-in modules 14:33:23 where can i put the sources ? 14:33:30 heh, figuring out ELF format pretty complicated 14:33:38 well, the source, that is 14:33:43 sourceforge, or can e-mail diffs back and forth 14:33:53 (single file, totally undoc'd 14:33:54 ) 14:33:56 I'll look around a bit, if you like 14:34:10 it caches TOS 14:34:13 I must go now, date with wife 14:34:25 and does a very efficient 0<> 14:34:27 yes, I'd like to participate ... I like your style 14:34:32 neg eax sbb eax,eax 14:34:34 --- nick: MrReach -> MrGone 14:34:45 (will read logs tomorrow) 14:35:00 what is your mail address ? 14:35:10 i don't know where to put it right now 14:35:15 so i'll send it 14:35:43 :) 14:35:47 oh, i try to implement predicatable timing where possible ... 14:35:50 have phun on date! 14:36:04 predictable ... 14:36:12 just an old habit :) 14:36:26 from real-time apps 14:36:52 that is, avoid branches if you can use logic instead 14:37:12 :) 14:37:16 no such thing like < if onething else otherthing then 14:37:27 i did that with my line drawing code heh 14:37:43 but have the same code for both sides of the logic 14:38:35 ehh 14:40:46 can't send mail right now as my notebook still tries to use my firewall as relay :) 14:41:07 hehe 14:41:09 doh! 14:41:18 i have hooked up my notebook to the internet cafe network :) 14:41:39 pulled out a network cable, and stuck it into the notebook 14:42:14 much better diagnostic tools on this machine :)) 14:42:16 hows the conversion to debian going ? 14:42:53 i only converted one machine (as dual boot), for bootserver. bootserver runs. bootclient causes problems. 14:43:07 have a nic with bootrom, but fails to broadcast 14:43:40 heh 14:43:51 otoh, using bootpc or bootptest shows the server responding all right 14:44:37 so i ot 3 choices: 14:44:40 ive never used bootp or dhcp 14:44:50 i got a q for you 14:44:58 figure out why the bootrom fails to send the bootp request 14:45:02 when i dial in with my current shitty dial in isp 14:45:06 use a floppy based boot client 14:45:14 i get my default gateway IP set equal to my own 14:45:21 MY ip is the default gateway 14:45:21 wait for the first real machine (diskless work station) 14:45:25 i am my own default gateway 14:45:28