00:00:00 --- log: started forth/04.09.05 00:07:13 --- quit: Sonarman (Read error: 104 (Connection reset by peer)) 00:15:50 --- quit: Herkamire ("good niiight") 00:33:19 --- quit: kc5tja ("THX QSO ES 73 KC5TJA/6 CL ES QRT AR SK") 03:19:02 --- join: crc (crc@38-pool1.ras11.nynyc-t.alerondial.net) joined #forth 04:25:18 --- join: FlamingRain (Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 04:37:14 --- quit: FlamingRain (Read error: 104 (Connection reset by peer)) 04:37:25 --- join: FlamingRain (Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 04:39:03 --- quit: FlamingRain (Client Quit) 05:12:14 --- join: FlamingRain (Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 05:14:44 --- quit: crc ("Time for bed... Goodnight!") 05:15:48 --- join: Topaz (~top@spc1-horn1-6-0-cust117.cosh.broadband.ntl.com) joined #forth 06:59:21 --- quit: slava (Read error: 110 (Connection timed out)) 08:33:26 grr slava's gone and his server is down. lol 08:33:34 1: slava took his machine down 08:33:41 2: someone took slavas machine down 08:33:49 3: slavas machine took itself down 08:33:51 lol 08:38:41 --- quit: FlamingRain (Connection timed out) 09:02:32 --- join: kc5tja (~kc5tja@66-74-218-202.san.rr.com) joined #forth 09:02:33 --- mode: ChanServ set +o kc5tja 09:20:03 `hah i just had a primer go kerbanga on me while seating it lol 09:20:12 last 2 that went off just went perfizzle 09:20:20 this one went BANG in a big way! 09:49:23 grrr slava wake uo and bring your box back online man 09:49:24 doh! 11:26:33 --- join: cccr (wossname@HSE-MTL-ppp77358.qc.sympatico.ca) joined #forth 11:27:11 --- nick: cccr -> wossname 12:08:34 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 12:10:59 ft slava lol 12:11:06 ft what happned to your server heh 12:12:51 --- quit: wossname ("what") 12:23:45 hi 12:23:52 I440r, its back up now 12:24:01 cool 12:24:05 i need to know something tho 12:24:19 but i cant remember the name of the equate 12:24:20 lol 12:24:41 the wha? 12:25:32 #define lol 12:25:38 erm hang on 12:26:09 TIOCGWINSZ 12:26:15 i think thats right 12:26:20 i need to know its valye 12:26:54 grep foo /usr/include/* 12:27:06 or find /usr/include -name \*.h -exec grep -l TIOCGWINSZ {} \; 12:27:07 it uses the _IOR macro 12:27:08 grr 12:28:30 that fscking macro does NOTHING except obfuscate the shit 12:29:09 i agree 12:29:31 write a C prog that prints its value 12:29:50 grr i hate that lol 12:29:54 im crap at c 12:30:03 how hard can it be 12:30:06 there has to be an easier way 12:30:06 #include 12:30:13 main() { printf("%d\n",TIOCGWINSZ); } 12:30:13 its not in stdio is it ? 12:30:20 #include 12:30:23 lol 12:32:13 sys/ttycom.h:#define 12:33:21 0x40087468 12:33:27 ok 12:33:34 you should've guessed that 12:33:39 lol 12:33:44 laughing rl here 12:34:16 i wonder how hard it would be to port your ncurses code to factor 12:34:34 oh btw 12:34:41 when i exit isforth, it doesn't reset the colors back to the default 12:34:44 ok. copy isforth out again 12:34:47 then extend 12:34:49 and ./isforth 12:35:32 yay it wokez 12:35:34 works* 12:35:35 woohoo!!!!!!!!!!!!!!!!!! 12:35:36 do this 12:35:43 ./isforth -fload src/examples/windwow.f 12:35:56 do THAT with ncurses :) 12:35:58 no such file 12:36:09 oh i'm writing -float 12:36:10 hehe 12:36:11 src/examples/window.f 12:36:21 ./isforth -fload src/examples/window.f 12:36:34 i think my terminal has bugs 12:36:38 let me try in console 12:36:53 use tab completion 12:37:22 its kinda messed up 12:37:36 make your terminal 100 by 36 and it isnt 12:37:43 lol 12:37:43 oh 12:37:53 i keep meaning to UN-hardcode taht 12:38:20 that some crazy 3d shit 12:38:31 3d ? 12:38:36 oh lol 12:38:37 yea 12:38:37 well how one window is in front then it goes behind 12:38:40 hehehe 12:38:42 the windows can be in front or behind 12:38:43 yeah, neat 12:39:13 now i'm going to type a command that begins with pkg_deinstall gf guess the rest of the line 12:39:32 hrm.... freebsd ??? :) 12:39:37 gforth :) 12:39:39 nah i might leave it there 12:39:44 but isforth will be the default :) 12:39:51 heh 12:39:55 can you fix the thing on exit tho 12:39:58 gforth can still do things i cant do yet 12:40:00 can it reset to previous colors in some ways? 12:40:09 resetting terminal to its default ? 12:40:21 i keep fergettin to do that too lol 12:40:29 look in src/terminal/* 12:41:02 actually, to do that ill need to add something to default 12:41:05 and to "atexit" 12:41:08 not a big deal 12:41:13 if i can figure out what lol 12:41:32 what is pluck? 12:41:40 copy third item to top 12:41:46 oh 12:41:48 i call that pick :) 12:42:03 err no 12:42:05 i don't have a pick ( n -- nth ) i think that's lame 12:42:07 wait lol 12:42:12 i have both pick AND pluck 12:42:18 i ferget which is which lol 12:43:02 i'm gonna be studying your ncurses code 12:43:06 it should be easy to port 12:43:33 i learned 99% of what i needed to know for that by reading the man page for term 12:43:41 oh really 12:43:52 there was only one thing i didnt get from the man page 12:44:00 so i got the termcap and terminfo book 12:44:17 that was to do with the translation thats required when you use the ibm box char stuff 12:46:22 took about 4 hours to code the initial run :) 12:46:32 many weeks to iron out the bugs tho and there are still some in there 12:46:38 try running isforth in a screen 12:46:47 im 50/50 as to weather thats an isforth bug or a screen bug 12:47:01 a LOT of apps break with screen 12:47:06 try running mc in screen 12:47:35 yea well i never use screen 12:48:23 i still have to fix the time shit 12:48:27 yea 12:48:41 which means i should parse timezone stuff 12:49:53 I use screen extensively. 12:50:01 I'm using it right now to yak on IRC, for example. 12:50:07 hi kc5tja 12:50:09 :) 12:50:16 Though, I have to admit, I have never, ever, ever, ever, ever, ever liked 'mc'. Ever. 12:50:23 i like mc 12:50:44 mc is just such an utterly *pale* imitation of a true dual-pane file manager, and frankly, it has actually never really worked for me. It doesn't at all map to how I think about file management. 12:51:04 RTFMing hasn't helped either. 12:51:19 I consider mc to be hostile and destructive to productivity as a result. For me at least. 12:54:57 man this is soooo freekin cool :) 12:55:07 isforth now runs 100% nativly under freebsd hehe 12:55:13 only ONE thing left to fix i think 12:55:20 thats a gettimeofday call 13:02:04 --- join: wossname (wossname@HSE-MTL-ppp66224.qc.sympatico.ca) joined #forth 13:02:32 what... no /proc/cpuinfo ??? lol 13:03:38 hehehe 13:03:42 you have to use syscalls 13:04:20 lol 13:04:25 how fast is your cpu 13:04:32 1.8ghz on the machine you're at 13:04:34 2.8ghz on mine 13:05:16 do a time ./extend 13:05:17 :) 13:05:30 0.131u 0.011s 0:00.16 87.5% 56+1348k 0+0io 0pf+0w 13:06:09 real 0m0.133s 13:06:09 user 0m0.117s 13:06:09 sys 0m0.014s 13:06:12 is what i got 14:00:01 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 14:01:23 --- join: ADD-101 (XINU@12-222-128-22.client.insightBB.com) joined #forth 14:01:31 tathi did you have to modify gettimeofday ? 14:01:36 in the extensions i mean 14:01:44 thats the ONLY thing wrong with the freebsd version now 14:02:00 oh and ive no idea why my svn server is offline. i have to put a monitor on the damned thing 14:16:07 I440r: Because you're running LInux instead of FreeBSD. :) 14:16:34 --- join: randolm (wossname@HSE-MTL-ppp73481.qc.sympatico.ca) joined #forth 14:16:41 I440r: I haven't done anything to gettimeofday 14:18:05 tathi can you compile the extensions ? 14:18:09 prolly not 14:18:16 you didnt do fsave yet lol 14:18:18 fergot about that 14:19:13 hang on :) 14:19:35 im fixin my time stuff first then ill put the fbsd version where youc an get it 14:20:07 fsave works fine on PPC, I think. 14:20:37 tathi: You're running a PowerPC version of isForth? 14:20:50 kc5tja: I've been half-heartedly porting it. 14:20:58 I have it fairly close to working. 14:20:59 Ahh. :) Cool. 14:21:09 --- join: FlamingRain (~Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 14:21:10 * kc5tja will have to find someone to port FTS/Forth for Linux when it's done then. >:) 14:22:40 :) 14:23:22 heh 14:23:43 stop being half heared grrr :P 14:23:47 * I440r thwaps tathi 14:24:07 I was wondering if you were going to notice that :) 14:24:18 lol 14:25:36 Well, today is turning out to be significantly more busy than I would like. 14:25:57 I really do want to spend more time coding on FTS/Forth than solving st00p1d customer support issues. 14:30:46 --- quit: wossname (Connection timed out) 14:33:27 --- quit: randolm (Read error: 104 (Connection reset by peer)) 14:33:51 --- join: wossname (wossname@HSE-MTL-ppp73481.qc.sympatico.ca) joined #forth 14:34:55 slava your gettimeofday is always returning ZERO in the tv structure 14:35:03 0 tv 14:35:07 returns 0 == success 14:35:13 and tv has the result of ZERO 14:37:27 gettimeofday(): returns the number of seconds elapsed since the current time. Any value other than zero indicates severe space-time discongruities. Please contact your system administrator right away if this happens. 14:37:52 :) 14:42:14 nono number of seconds since the epoch 14:42:15 lol 14:42:27 and the syscall is working cuz it doesnt return an error 14:43:45 You missed the joke, I guess. 14:43:52 At least, I think. 14:44:30 --- quit: wossname ("mr. push up ;x;x") 14:46:54 no i saw it lol 14:46:57 i SAID lol 14:48:24 Whatever -- it's hard to tell on IRC. 14:49:36 --- quit: qFox ("meh ircii's netsplit detection appearantly sux ^^") 14:52:42 Let me completely re-iterate my utter and total hatred for Perl. 14:53:41 lol 14:54:38 How, in Perl, do I get the current date and time expressed as a string? Anyone know? 14:54:41 --- quit: FlamingRain ("Leaving") 15:05:09 --- quit: ADD-101 (Remote closed the connection) 15:05:21 Got it. 15:05:24 localtime(time()) 15:06:52 sup kc 15:07:06 thats about how you do it in C too 15:25:00 --- quit: tathi ("leaving") 15:30:03 --- quit: Topaz (burroughs.freenode.net irc.freenode.net) 15:30:03 --- quit: warpzero (burroughs.freenode.net irc.freenode.net) 15:30:03 --- quit: madwork (burroughs.freenode.net irc.freenode.net) 15:30:03 --- quit: madgarden (burroughs.freenode.net irc.freenode.net) 15:33:11 --- join: Topaz (~top@spc1-horn1-6-0-cust117.cosh.broadband.ntl.com) joined #forth 15:33:11 --- join: warpzero (~warpzero@dsl.103.mt.onewest.net) joined #forth 15:33:11 --- join: madwork (~madgarden@derby.metrics.com) joined #forth 15:33:11 --- join: madgarden (~madgarden@Toronto-HSE-ppp3707020.sympatico.ca) joined #forth 15:33:38 kc5tja, ping 15:39:13 slava 15:39:18 yes 15:39:25 gettimeofday works MUCH better when you use the correct syscall number 15:39:37 edit your src/ext/timer.f and change the 74 to 116 15:39:40 and run ./isforth 15:39:44 oh cool :) 15:39:47 the only thing wrong now is the timezone offset 15:44:25 syscalls have numbers? 15:45:08 lol 15:45:20 syscall 0 is sys_syscall 15:45:25 syscall 1 is exit ? 15:45:28 i wonder how you call syscalls from C 15:45:49 well thats what libc does hehe 15:45:58 well what if you don't want to use libc 15:46:09 what do you mean "what if" 15:46:10 lol 15:46:17 there is no reason not to use libc unless you're I440r 15:46:22 yeah there is 15:46:27 libc sux 15:46:29 nuff sed 15:46:30 :) 15:46:37 if you're writing something that isn't c 15:46:42 libc is a library for c 15:46:52 now i gotta go find out how to read timezone info grrr 15:46:58 warpzero, its just called that for historical reasons 15:47:01 warpzero, libc is the unix api 15:47:15 i'll api your unix 15:47:17 slava no. libc is the C library 15:47:20 how do you call syscalls from c 15:47:41 yes its also the unix api but unix people have their head up their ass. they ASSUME your going to do everytyhing with c 15:47:42 well, compare the effort I440r had to go through to port his forth to freebsd, with 'porting' something using libc ;-P 15:47:46 or with something that uses c 15:47:50 warpzero, inline asm 15:47:56 really 15:48:11 hey I440r 15:48:17 --- join: TheBlueWizard (TheBlueWiz@modem-066.nyc-tc03a.FCC.NET) joined #forth 15:48:17 --- mode: ChanServ set +o TheBlueWizard 15:48:34 how hard would it be to use everything from isforth EXCEPT the forth parts 15:48:46 like the syscall interfaces and library functions 15:49:07 i'm going to figure out how to do FFI 15:49:17 are there any decent wrapper libs or do i have to use dlopen() etc? 15:49:27 i want to get libX11 and libGL going 15:54:14 fuck 15:54:20 i need to get my language implemented 15:54:23 but 15:54:54 but what :0 15:54:57 if i want to implement it quickly i'd have to use c/glibc 15:55:07 nothing wrong with libc 15:55:11 don't listen to I440r :) 15:55:12 yeah there is 15:55:15 what? 15:55:21 its for c 15:55:33 that's a philosophical point 15:55:48 lisp implementations usually use it too 15:55:51 as does python 15:55:52 and perl 15:55:53 etc 15:56:06 this language needs go fast 15:56:15 what makes you think libc is slow? 15:56:18 its not a scripting language 15:56:26 do C programs go slow? 15:56:28 i mean assembly fast dude 15:56:39 you can generate all the asm you want in your compiler 15:56:52 calling libc's open() -vs- open syscall will save, what, 1 nanosecond? 15:56:55 yeah but i also need a significant RTE 15:57:00 what is RTE? 15:57:03 runtime environment? 15:57:06 yeah 15:57:14 well, look at the open source lisps 15:57:16 they use libc 15:57:25 but in sbcl and cmucl's case, the entire lisp is metacompiled 15:57:27 yeah 15:57:31 only the garbage collector is in C i think 15:57:35 the rest is in lisp 15:57:36 define metacompiled 15:57:37 and they're FAST 15:57:40 like all in lisp 15:57:45 and you use another native compiling lisp to bootstrap it 15:58:14 my language isn't self-bootstrapable 15:58:29 i'd implement it firstr 15:58:35 then worry about performance once you've identified the bottlenecks 15:58:41 premature optimization is the root of all evil 15:58:48 yeah you're probably right 15:59:14 ugh 15:59:19 this is going to be an implementation nightmare 16:00:44 you guys with your forths have it easy 16:01:08 slava whats the date and time there ? 16:01:14 > date 16:01:14 Sun Sep 5 19:00:14 EDT 2004 16:02:01 well okay slava ill do it in C/libc 16:02:03 grr why is it reporting sunday 6 sept 2004 23:00:33 16:02:10 but i'm in no way going to use autoconf 16:02:28 warpzero, me neither 16:02:57 warpzero, my C code compiles and runs WITHOUT #IFDEFS on linux, freebsd, sharp zaurus, solaris (64-bit mode), macos x, windows/cygwin... 16:03:05 warpzero, because I stick to sane, portable APIs 16:03:18 warpzero, without doing stupid tricks to increase performance by 5% at the expense of all other OSes 16:03:47 show me how to do that 16:03:54 read man pages 16:03:59 note the 'PORTABILITY' section :) 16:04:04 lol 16:04:05 if it says POSIX standard, you're all good :) 16:04:22 actually, i think my FFI will only work on freebsd, at the start anyway 16:04:50 okay hmm 16:04:58 slava thats not quite true 16:05:01 you- DO use ifdef's 16:05:08 in the heder files you include 16:05:16 I440r, yes but the headers are standard 16:05:20 i don't care what's in the headers 16:05:27 as long as they provide the handful of standard functions that I need 16:05:27 i suppose i should start with the parser 16:06:28 should i use a lex/yacc parser 16:06:39 --- join: FlamingRain (~Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 16:06:42 i wrote my own parser in the language itself 16:06:47 the C core doesn't parse anything 16:06:52 i can't do that? 16:06:54 it just loads an image, maps it into memory, and begins executing 16:06:58 ooh i think i know why this time and date are wrong 16:07:11 is the Linux epoch on a different date to the freebsd epoch ? 16:07:33 afaik no 16:07:37 but i could be wrong 16:08:17 slava: if i don't write a parser i won't be able to write a parser in the language either 16:08:30 argh 16:08:38 warpzero, hehe 16:08:38 isforth in linux is displaying it wrong too 16:08:50 warpzero, i wrote a java implementation first 16:08:58 EW 16:09:01 gross dude 16:09:02 warpzero, then 'meta-compiled' the first image for the C implementation 16:09:16 warpzero, now the C implementation is 'self-hosting' in a sense 16:09:21 i use quotes since the iamge doesn't contain machine code 16:09:23 but rather bytecode 16:09:27 why does my time calc think its sunday the SIXTH 16:09:33 yeah i have a bytecode too 16:09:42 doh 16:09:44 its actually nibble-width 16:09:48 warpzero, the java implementation was actually quite fast 16:09:56 warpzero, it would generate java bytecode,and the jvm would generate machien code from that 16:09:56 ew java 16:10:02 so in some cases it beats the C implementation in performance 16:10:17 only thing i don't like about java is the memory consumption 16:10:20 Lunch break. brb 16:10:23 but i have 512mb of ram *shrug* 16:10:51 i could just use a hex editor i guess 16:10:59 lol 16:11:38 kc5tja, what's a good way of doing FFI 16:11:39 slava can you take a look at src/ext/timer.f and tell me why im a day off ? 16:11:50 i mean day of the week is right but the day is off by 1 16:12:24 I440r, shit how would i know? 16:12:44 slava codes forth yes ??? 16:12:47 yes 16:12:49 but 16:12:54 well thers a bug in mine lol 16:12:58 its hard to tell what the problem is in your code 16:13:09 is 116 the right syscall #? 16:13:16 are you sure the claling convention matches libc's 16:13:36 yes and yes - the syscall is right now 16:13:44 well you're doing tv @ 16:13:49 you realize tv has two fields right :) 16:13:50 yup 16:13:54 yup 16:14:03 so you're ignoring tv_usec? 16:14:22 yes 16:14:33 i dont care about MICROSECONDS lol 16:14:41 do i ? 16:14:52 well i'm not sure if gettiemofday returns GMT or user's tz 16:15:29 gmt i believe 16:15:48 thats why the TZ (execept thats not used in fbsd) 16:15:53 its ignored 16:16:24 what is ':' 16:16:25 Sun, 06 Sep 2004 23:15:14 16:16:34 a character literal 16:16:43 oh 16:16:50 'x' is interpreted by number as being a number equal to the ascii code for x 16:16:59 it works with everything exceppt space 16:17:11 ' ' will return the code field address of the word tick :) 16:17:20 that's a bit messy :) 16:17:26 lol 16:17:31 u should be using bl anyway :P 16:17:40 ans forths use 'x 16:17:48 which is freekin fugly 16:17:49 i don't understand, does .elapsed print the current time? 16:17:51 what is stime? 16:17:54 no 16:17:57 do ./isforth 16:17:58 oh since timer-reset 16:18:00 yes 16:18:01 do timer-reset 16:18:03 right heeh 16:18:12 its similar to time foo from shell 16:18:20 timer-reset do-something .elapsed 16:18:27 yes 16:18:36 good for useless benchmarks :) hehe 16:19:52 what is your gettimeofday right now 16:19:56 3504694794 16:20:03 where ? 16:20:07 on a linux box 16:20:19 if its GMT and same epoch, they should be equal 16:20:28 time@ . gives 1094426473 16:20:29 oh that's milliseconds sorry 16:20:38 do time@ h. 16:20:46 413B9F81 16:20:50 hex is easier to read 16:21:10 413B9F2A 16:21:19 were close 16:21:28 so i say its MY bug 16:21:54 i just notiecd factor's usage of gettimeofday is wrong 16:22:45