00:00:00 --- log: started forth/04.08.12 00:00:02 hi :) 00:03:48 re 00:04:16 :) 00:46:01 --- nick: Tomasu -> TomasuDlrrp 00:46:06 * TomasuDlrrp is away: night 00:49:40 I ahve to get to bed. 00:49:43 Work tomorrow. 00:49:57 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 03:09:48 --- join: Serg_penguin (~z@212.34.52.140) joined #forth 03:09:59 hiya from work ;( 03:10:10 everything died, i was called to revive 03:14:05 --- part: imaginator left #forth 03:32:59 foof... kinda like hardw is OK 03:33:15 but TCP drivers of W2K Serv flew off 03:36:01 eek 03:52:03 * Serg_penguin raised one (un)dead, moving to another 04:08:25 --- join: crc (crc@0-1pool40-18.nas36.philadelphia1.pa.us.da.qwest.net) joined #forth 04:20:33 hi CRC ! 04:20:58 please remind me of J2me mail ;(( your link is at home, i'm at fat line at work 04:25:42 Good morning Serg_penguin :-) (link is http://mail4me.objectweb.org/) 04:26:24 mega TNX ! 04:26:38 put it on USB flash drive in links collection ;)) 04:28:06 now will try to download it right by cellfone 04:30:25 suuuuxxx !! 04:30:33 link won't fit in SMS ! 04:30:38 :-( 04:33:10 is it English or German ? 04:33:19 i see DE in filetree 04:35:01 I don't know, I haven't used it... 04:37:07 * Serg_penguin edited SMS at phone and trying to download 04:39:24 ICQ ? IRC ? WWW ? on j2me, of course... 04:42:05 --- join: Topaz (jonny@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 04:49:03 failure.. will try it thru cable at home 04:55:40 The only real j2me www browser I can find is a commercial one :-( 05:05:53 --- quit: slava (niven.freenode.net irc.freenode.net) 05:05:54 --- quit: TomasuDlrrp (niven.freenode.net irc.freenode.net) 05:08:01 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 05:09:25 --- join: Tomasu (~moose@S010600045a4c73cc.ed.shawcable.net) joined #forth 05:43:26 * Serg_penguin is trying to make ZYXEL Omni 56K Mini work on crappy line 05:44:15 * crc finally has an air conditioner -- makes the house much cooler :-) 05:44:27 It got colder here today 05:44:36 Now it's below 30C in my room :) 05:45:11 the weather here is fairly typical... raining on monday, crazily hot on tuesday, thunderstorms yesterday, hot today... 05:46:14 Hehe 05:51:42 --- quit: madgarden ("*frotz*") 05:52:35 --- join: madgarden (~madgarden@Kitchener-HSE-ppp3576090.sympatico.ca) joined #forth 06:26:32 * crc is attempting to write a small (512 byte) 'monitor' program at the moment 06:28:27 This would be much easier if I had 1k to work with... 06:44:09 --- quit: crc (Read error: 104 (Connection reset by peer)) 08:24:01 --- quit: Serg_penguin () 08:51:00 --- join: warpzero (~warpzero@mi179.dn177.umontana.edu) joined #forth 09:12:57 would you guys like to meet my brother, who's learning IRC? 09:24:00 learning irc? that's a new one on me. is he going to become a certified IRC technician? 09:24:46 :D 09:24:47 naah. 09:24:51 just teaching him /join etc. 09:24:54 and how to behave 09:25:00 that's the most important part. 09:25:24 not acting like an AOL n00b :) 09:25:43 or a skript kiddie. 09:25:53 yep 09:31:36 --- join: warp0b00 (~warpzero@mi179.dn177.umontana.edu) joined #forth 09:41:00 YAY 09:41:05 i still have hope for humanity 09:41:12 somebody found my wallet! :) 09:41:25 Heh. 09:41:30 That's good..:) 09:41:46 :) 09:52:15 --- quit: warpzero (Connection timed out) 10:21:54 still cant drive until it arrives 10:21:58 :/ 10:22:20 Woo, pixel art is fun! http://www3.sympatico.ca/ppridham/misc/project/deathfist/mockup/fight_test.gif 10:53:22 --- join: thefox (~fox@adsl-68-122-3-243.dsl.pltn13.pacbell.net) joined #forth 13:01:55 --- quit: thefox (Read error: 60 (Operation timed out)) 13:35:50 * Tomasu is back (gone 08:26:27) 13:40:42 --- part: Topaz left #forth 13:40:52 --- join: Topaz (jonny@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 14:22:59 --- join: TheBlueWizard (TheBlueWiz@ip-207-198-223-121.nyc.ny.fcc.net) joined #forth 14:22:59 --- mode: ChanServ set +o TheBlueWizard 14:28:38 --- quit: Topaz (Read error: 110 (Connection timed out)) 14:29:10 --- join: Topaz (jonny@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 14:30:45 --- join: doublec (~doublec@coretech.co.nz) joined #forth 14:34:24 hi doublec 14:34:50 hi slava! 14:35:58 doublec, let me commit something nifty 14:36:06 ok 14:36:55 ok, update cvs 14:37:03 stack overflow/underflow checks are now done with guard pages 14:37:09 i'm curious if this will compile on linux 14:37:18 nice, I'll try it now. 14:37:24 i use mmap, mprotect, and a PAGE_SIZE constant that fbsd defines -- the latter might not be defined in linux I fear. 14:37:59 'word of the day' cute :-) 14:38:08 haha i needed some way of testing the random number words. 14:38:35 does it work? ie, does 'drop' on an empty stack cause a bus error? :) 14:38:38 i haven't done a signal handler yet 14:39:02 * Tomasu is away: for a bit 14:39:02 Factor compiles but when run I get: 14:39:08 Loading factor.image... relocating...Fatal error: Not empty 54 14:39:23 make a new image 14:39:27 oh wait 14:39:34 are you sure you have the latest code from cvs? 14:39:48 Just did an update and build of the image then. 14:39:51 is there native/port.c and native/iomux.c files? 14:39:52 I'll do a clean and try again. 14:39:58 i just added those 14:40:06 yep, they're there. 14:40:31 are you sure it compiled w/o errors? 14:40:42 just checking now... 14:41:08 the string 'Not empty' is not present in the latest code 14:41:42 ahh, you're right. It didn't build. 14:41:51 native/memory.c: In function `alloc_guarded': 14:41:52 native/memory.c:7: error: `PAGE_SIZE' undeclared (first use in this function) 14:41:52 native/memory.c:7: error: (Each undeclared identifier is reported only once 14:41:52 native/memory.c:7: error: for each function it appears in.) 14:42:00 ok 14:42:17 let me think 14:42:37 are there #ifdefs i can check for linux/386? 14:44:58 doublec, can you just #define PAGE_SIZE 4096 in factor.h for now 14:45:00 --- quit: tgunr (Excess Flood) 14:45:14 yep 14:45:58 i started refactoring the i/o code into some kind of CPS style 14:46:07 Fatal error: Cannot allocate low guard page -1 14:46:10 cool 14:46:23 oops 14:46:33 here is the offending line: 14:46:47 --- join: tgunr (~davec@vsat-148-63-4-107.c001.g4.mrt.starband.net) joined #forth 14:46:50 mprotect(array,PAGE_SIZE,PROT_NONE) 14:46:53 --- quit: Topaz (Success) 14:47:16 --- join: Topaz (jonny@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 14:48:16 not if you want it to run on ppc 14:48:21 i think ppc page size = 16k 14:48:28 I440r_, yes i know 14:48:35 ok :) 14:48:44 doublec, can you add a printf to check the address returned by mmap 14:48:48 maybe its not 4kb aligned for some reason? 14:49:00 I440r_, want to take a quick look at the code/ 14:49:01 its guaranteed to be 14:49:12 note! - you cannot use mmaped buffers for dma 14:49:15 --- join: Nahuel (~cuco@200.45.247.112) joined #forth 14:49:19 --- part: Nahuel left #forth 14:49:25 sure 14:49:40 http://paste.lisp.org/display/2292 14:49:47 slava, do you want an account on a linux box to try it out? 14:50:15 doublec, i'm actually busy with some jedit bugs right now, i wanted to finish those first before doing factor work 14:50:30 no worries, I'll do some investigating. 14:50:33 doublec, but maybe later? 14:51:17 doublec, try this at the shell 14:51:20 yep that's fine 14:51:23 touch foo.h; cpp -dM foo.h 14:51:26 it will list all #defines 14:51:30 and pastebin it please 14:51:36 try mprotect it to something other than prot none to see if its disallowing none 14:51:43 mmap result: ffffffff 14:51:55 hmm 14:51:59 (that's from a printf of 'array') 14:52:09 right 14:52:38 the parameters to mmap must be confusing linux 14:52:48 I440r, does linux mmap allow an address of 0 to mean 'anywheere'? 14:53:40 according to the man page you must supply one of MAP_SHARED or MAP_PRIVATE 14:53:46 oh ok 14:53:51 try MAP_PRIVATE|MAP_ANON 14:53:53 you cant have anon shared 14:54:08 well... you can in 2.4 and up 14:54:19 but 2.2 doesnt support anonymous and shared at the same time 14:54:25 that fixed it! 14:54:28 we're not doing shared here ;) 14:54:30 :) 14:54:30 doublec, great 14:54:57 doublec, does it detect underflow? 14:55:06 and overflow? (try [ 1 2 3 ] forever) 14:55:19 erm are you traping pagefaults ? 14:55:22 not yet 14:55:23 segfault 14:55:30 at the moment it just dies 14:55:32 heh 14:55:35 doublec, bus error on fbsd :) 14:55:36 but that's great 14:55:42 doublec, this gives a 2x performance boost. 14:55:42 well dying is good them :) 14:56:04 doublec, can you pastebin the output of that cpp command please 14:56:44 ok, here are the changes I made to get things to compile tho: http://paste.lisp.org/display/2295 14:56:49 thanks 14:56:59 doublec, please don't specify #jedit ast he channel, use 'none', otherwise #jedit gets spammed ;) 14:57:06 oh, ok. 14:57:21 oh linux has PAGESIZE? 14:57:53 errr are you doing stack checks after each word executed or just when the user hits enter ? 14:58:10 : foo xxx yyy zzz ; 14:58:12 I440r_, i'm not doing stack checks 14:58:27 slava previously i mean 14:58:41 would execution of foo do a stack check after xxx, yyy and zzz ? 14:58:42 http://paste.lisp.org/display/2296 14:58:43 I440r, after each word executed 14:58:57 hrm ok :) 14:59:03 then this is worthwhile hehe 14:59:05 slava re PAGESIZE, apparently. The code I used there is the recommended from the manpage. 14:59:10 doublec, ok great 14:59:19 isforth only does a stack check when control is passed back to quit 15:00:54 doublec, i added this: 15:01:04 #ifndef PAGE_SIZE 15:01:04 #ifndef PAGESIZE 15:01:04 #define PAGESIZE 4096 15:01:04 #endif 15:01:04 #define PAGE_SIZE PAGESIZE 15:01:05 #endif 15:01:17 if this gets any more complex i'll move it to pagesize.h 15:01:20 cool 15:01:23 * slava can hear I440r scream at the #ifdef :) 15:02:36 slava, I wrote a 'parser combinator' library and to support that a lazy evaluation library. Is it ok to add it to 'contrib'? 15:02:50 * TheBlueWizard chuckles 15:03:36 --- join: FlamingRain (~Ecoder@c-24-129-95-254.se.client2.attbi.com) joined #forth 15:04:07 doublec, sure! 15:04:25 --- join: rpc (~rpc@S0106000d6151238b.gv.shawcable.net) joined #forth 15:04:52 lazy list of prime numbers: 1 lfrom [ prime? ] lsubset 15:05:15 The lazy stuff is pretty complete and I'm finishing off the combinators. 15:05:21 this is really cool 15:05:34 --- quit: Topaz (Connection timed out) 15:05:36 So far you can do pretty much any parsing with the combinators quite easily. It's pretty neat. 15:05:43 lol 15:05:54 no - i wont scream... not so you can hear it anyway :) 15:05:58 doublec, is it good enough to parse xml? 15:06:10 yep 15:06:22 In factor, the definitions would follow very similar to that listed in the XML spec. 15:06:34 i'm interested in an xml-rpc library 15:06:40 me too 15:06:41 --- join: Topaz (jonny@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 15:06:49 time to go home 15:06:51 :) 15:07:05 Code for generating a combinator that parser zero or more integers: 15:07:07 doublec, i think its time to add cont-html and cont-responder to library/httpd 15:07:16 digit <*> 15:07:29 With an optional minus sign: 15:07:35 "-" token digit <*> 15:07:36 what is <*>? 15:07:39 kleene star? 15:07:43 yep 15:08:14 I'm still working on nice words. I've used those from a haskell paper at the moment. 15:08:34 how hard was this to code? 15:09:21 not hard. Mostly manipluation of code blocks to do currying and stuff. 15:09:40 does it use continuatoins? 15:09:54 No. To do the backtracking it uses lazy evaluation. 15:10:03 I could have used continuations though. 15:10:10 doublec, i asked in #solaris, they said: 15:10:12 #ifdef sparc 15:10:16 #define PAGE_SIZE 8192 15:10:35 But I was basically working through a paper from a lazy functional language background so lazy evaluation was the easiest approach. 15:10:42 ok 15:11:58 --- quit: FlamingRain ("Leaving") 15:13:14 slava: you should check for linux - sparc linux would be the 4 one 15:13:19 (i believe) 15:13:42 arke, maybe autoconf already has a template for this? 15:13:43 ~4096 15:13:47 ~4096 15:13:49 4096 15:14:01 possibly 15:14:37 slava, the start of documentation for the lazy vocab: http://www.double.co.nz/factor/lazy.html 15:15:37 doublec, why lconcat and not lappend? 15:16:37 doublec, can you add this html file to the contrib directory too? 15:16:45 slava, yes it should be lappend. I'll be changing that. 15:16:57 slava, yep. I'll have it in the same dir as the code. 15:17:53 slava, should you maybe use getpagesize() for returning the page size? 15:18:00 is it portable? 15:18:09 HISTORY 15:18:09 The getpagesize() function appeared in 4.2BSD. 15:18:34 Linux has it but not sure when it first came in. 15:18:48 i wasn't aware it existed. its probably better to use it then 15:19:43 :) 15:19:51 why do you need the page size? 15:19:56 arke, mprotect() 15:20:00 * arke will check in a sec for sparc64/linux page size 15:20:06 hmm? 15:20:28 whats mprotect do? 15:20:37 changes memory protection of pages :) 15:23:04 --- quit: Topaz (Read error: 110 (Connection timed out)) 15:23:29 :) 15:24:35 arke, stack overflow/underflow detection 15:25:51 doublec, i'll probably have non blocking i/o done later today. 15:26:22 slava, nice! 15:26:34 native factor has come a long way in a short time. 15:27:42 i'm amazed that i haven't had to use autoconf yet ;) i checked, getpagesize exists on solaris and mac os x too 15:29:07 well "checked" means bugging people in #java 15:29:10 :) 15:29:38 heh :) 15:35:42 --- join: jDoctor (~rex@pcp08550359pcs.manass01.va.comcast.net) joined #forth 15:36:43 doublec, jDoctor is a guy from #java :) 15:37:28 cool. You're slowly pulling java people towards factor aye ;-) 15:37:29 must convert him then! 15:37:32 }:) 15:37:37 not bad of a programmer at all for a high school kid 15:38:11 i guess there's nothing to do in mansas on weekends for kids 15:38:15 when i was in high school i mostly just partied 15:39:03 ^ fuuunnt 15:39:07 funny. . 15:39:09 but you can't spell jDoctor 15:39:28 yah well. anyway, I at least gotta gf 15:40:04 we party in some way. 15:40:17 the kid next door is my age and plays Xbox all day 15:40:27 my Xbox doesnt even play games anymore. it has debian on it 15:40:40 haha 15:40:45 does it work well? 15:40:52 its just an x86 pc really right? 15:41:03 yah, actually. the xbox is a 733 mhz with 10 gigs. 15:41:13 what video hardware? 15:41:34 the most powerful thing on it is the nvidia with 64 MB vram, but I dont think I have drivers to use that 15:41:41 not the full potential anyway 15:41:44 can't you use the linux drivers 15:42:28 one month left before university starts again, i'm looking forward to it 15:42:30 I dont even know what the hell it is using. it's blurry on TV, so I leave the Xbox headless and do xdmcp to it 15:42:38 summer holidays are soooo boring 15:42:46 one month left for.. school. 15:42:51 dude, summer is freedom 15:43:01 well uni is less of a routine than hs 15:43:07 I learn more in summer than I do when there is all this pointless school work. 15:43:09 you don't spend ALL FRIGGIN DAY at uni 15:43:29 * Tomasu is away: again 15:43:37 is it on campus or you commute? 15:43:52 i live about 15 mins drive from downtown 15:44:06 the campus is in downtown 15:44:36 ah. slick. what college? 15:44:39 there's a bar across the road from the campus 15:44:50 in the first year, i'd always go there on my breaks to have pitchers with friends 15:44:57 jDoctor, www.uottawa.ca 15:45:00 universiyt of ottawa 15:45:07 check out their "optimized for IE" site 15:45:25 dang. that many people speak French there?! 15:45:35 oh yah. its canada 15:45:44 i can show you my dad's home page (he's a math prof) 15:45:48 let me find it 15:45:52 <-- doesnt know much outside his field 15:46:30 http://aix1.uottawa.ca/~vpest283/ 15:46:48 apache personal directory thingy ay? 15:46:52 I bet he gets ftp to it 15:47:05 yes 15:47:09 not much of a homepage eh? 15:47:49 fine for a math guy. man you're lucky 15:48:08 you can get a phd and still never leave a comfortible env. 15:48:47 I'd hang around a college alot just because I'd rather be surrounded by people who use their powerful minds than people who waist them. 15:49:02 well you can do commercial work and not waste your mind too you know ) 15:49:03 like bill gates who kinda only went to college to pull out useful people. 15:49:05 that should be :) 15:49:47 commercial work? 15:49:58 like enterprise programming 15:50:04 oh, no doubt 15:50:26 I'm just too liberal to do that for too long. I'd HAVE to if I wanted to go to berkeley 15:50:29 in california 15:50:39 but after so long, I'd have my own org. 15:51:32 ok i'm a dumbass 15:51:53 jDoctor, i did a JList.setFixedCellHeight(list.getFont().getHeight()) before a setFont() call 15:52:10 i'm wondering why its all squished 15:52:34 I thought java does that automatically 15:52:46 where are you useing jlists? 15:52:52 * Tomasu is back (gone 00:09:24) 15:52:55 no, if you don't set a fixed list height, it calculates the height for each row individually 15:53:02 since they might differ due to custom renderers 15:53:08 for large lists you always set a fixed height 15:53:47 ah for efficiency 15:54:42 ok, that's enough jedit for today. squished 3 bugs 15:56:02 how funny. kid next door (john) just called. mentioned that he finally beat final fantasy 10 and he will give me the game back, but I'm just genna sell it back at funco land. 15:57:50 I440r, did you ever make any progress with freebsd isforth? 16:00:21 man i'm getting confused between ascii chars and unicode chars 16:00:42 unicode is the superset, right? 16:01:19 ascii is actually a subset of unicode, iirc 16:01:31 i mean 8 bits vs 16 bits 16:02:06 ah 16:02:08 doublec, you'll notice right now the i/o primitives strip the high 8 bits of each character. eventually i want to have utf8 and utf16 encodings 16:02:18 doublec, i don't want to bother with using iconv or anything 16:02:30 or have 10000000 encodings like java 16:02:50 yah I saw that list in the api. 16:04:22 slava, sounds like a good idea. Otherwise you spend your life supporting weird character encoding bugs. 16:04:26 where you are, slava , cant you snowboard alot? 16:04:33 alotta .. snowy places? 16:04:40 yea, i never been tho 16:04:43 maybe this winter 16:04:59 ah man. I'm king in that field. 16:05:12 they have GNU boards, you know 16:05:20 its just a snowboarding company 16:05:25 slava i need to log into ur box again sometime 16:05:34 I440r, let me know when and i'll set you up. 16:05:52 --- nick: onetom_ -> onetom 16:05:54 ok 16:06:19 might be able to do some work on in in a min or 2 :) 16:06:21 I440r, how's tathi's ppc port (eer rewrite :)? 16:06:28 slava, native factor doesn't show the stack effect when you 'see' a word. Is there a way to get this information? 16:06:41 doublec, at the moment it discards doc comments, i plan on fixing this though 16:06:43 wekk. but tarhi has disappeared :) 16:07:02 I440r, i don't like your state smart words tho :) like -> 16:07:17 !> ? 16:07:22 for storing in a VAR? 16:07:26 nothing wrong with state smarts 16:07:34 yes there is. what happends if i do ' -> 16:07:35 yes or for modifying a constant :) 16:07:47 whats -> ??? 16:07:55 what is it named then? 16:08:05 !> 16:08:05 the one that stores into the variabled named by the following word 16:08:06 oh 16:08:07 store 16:08:07 to 16:08:12 heh 16:08:21 +!> 16:08:23 incr> 16:08:24 decr> 16:09:02 state smart is like GOTO. nothing wrong with it unless you OVER do it 16:11:15 slava, what happened to 'map'? Should I use 'inject' instead? 16:11:19 doublec, yes 16:11:28 doublec, map is going away and then inject will be renamed to map 16:11:39 ok 16:11:46 doublec, don't use expand either 16:11:59 ok 16:12:11 i think map/expand were a mistake 16:12:36 doublec, check out library/math/simpson.factor <-- (slow) numerical integration 16:12:49 at the moment you manually specify accuracy 16:12:57 i'm going to hack it to keep doubling the accuracy until the result converges 16:14:14 cool! 16:14:41 A few of the cont-responder examples work fine under native factor. I need to tweak some of them for the changes (like map, etc). 16:15:09 nice 16:15:37 factor has now absorbed many of the features of my earlier mtac project 16:16:00 I like the test responder. nifty. 16:16:05 weirdo. why not just make one lang that answers all? 16:16:10 or are you searching for that. 16:16:14 jDoctor, i wrote mtac like 4 years ago 16:16:17 jDoctor, its not really a language 16:16:22 jDoctor, just an interactive algebra system 16:16:47 ah. thats what I got so far. well, mine doesnt even do doubles correctly 16:16:50 i had some nice code for computing various special functions like Erf, Ei, Li, Si, Ci etc in there. I might port that some day 16:17:48 when I type a period to signify a decimal, the parser thinks the command ended. 16:17:57 hehe. thats like that last thing I need to fix 16:18:09 why not use , for decimal 16:18:14 or ; for end of command :) 16:18:17 otherwise its ambiguous 16:18:25 THAT is ambiguos, slava 16:18:30 for seperation of params 16:18:37 add(4,4). 16:18:39 shilly 16:19:01 my listener does add one number, also. it adds it to 0 16:19:09 so add(4). returns 4 16:19:19 jDoctor, you're slowly reinventing lisp ;) 16:19:35 what about add()? that should be zero :) 16:19:38 well, mine is still better 16:19:42 hehe 16:19:44 and multiply() should be 1 16:19:48 * (+) 16:19:51 0 16:19:53 * (*) 16:19:54 yah! you looked at the code? 16:19:57 1 16:20:03 no, its what lisp does 16:20:13 /(9,3,3). returns 1 also 16:20:22 also in lisp (/ 3) == (/ 1 3) 16:20:42 whaa 16:20:49 returns 1/3rd ? 16:20:58 that's not what i meant 16:21:04 (/ x) is the same as evaluating (/ 1 x) 16:21:07 but yes it returns a third 16:21:18