00:00:00 --- log: started forth/04.07.19 00:29:13 --- join: kc5tja_ (~kc5tja@66-74-218-202.san.rr.com) joined #forth 00:35:29 --- quit: crc ("http://www.retroforth.org/dev/beta_releases/retroforth-7.beta4.tar.gz") 00:42:32 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 00:46:42 --- quit: kc5tja (Read error: 110 (Connection timed out)) 01:06:27 --- join: arke (arke@melrose-251-251.flexabit.net) joined #forth 01:06:59 kc5tja!! 01:07:14 :) 01:07:19 talk to me! 01:10:11 wtf 01:15:01 :/ 01:15:17 kc5tja_: are you asleep? 01:17:58 :/ 01:18:04 and I already have to leave 01:18:15 oh life, so cruel... 01:20:08 oh well 01:20:16 I'll talk to you guys later4 01:20:31 I might just have a surprise for the forth world when I get back to the US :) 01:21:05 you're pregnant?! 01:21:09 --- quit: arke ("Ich hab' alles was Ich brauch', und 'nen bauch hab' Ich auch!") 02:22:17 --- quit: jDoctor (Client Quit) 02:34:25 --- join: O3BEPH (~z@212.34.52.140) joined #forth 02:34:57 --- nick: O3BEPH -> Serg_Penguin 03:54:22 i'm not sure why the forth world would care about him being pregnant.... unless he delivers forth cpu's to this world, but then it wouldnt be just a surprise to the forth world... ;) 04:23:48 http://weeble.fluent.ltd.uk/toons/37/ 05:18:21 --- quit: Serg_Penguin () 06:03:24 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 06:06:28 urre. my swap consistes of... about 76 calls, and 31 executions (of CODE words). 06:06:43 and it doesnt even work :p 06:12:11 --- quit: aum () 06:13:04 qFox: how'd you manage that? 06:13:34 well. for now, because my >r word is not functioning correctly 06:13:35 :) 06:13:43 i'm messing stack movement up 06:16:36 :) 06:20:39 * Robert returns. 06:20:59 Mmmm... 24 hours without sleep, almost as long without food. 06:21:14 * Robert enjoys his tea. 06:27:16 is there a reason for this intensional amnesia? :) 06:29:45 Are you accusing me of having a bad of having a bad accusing me of having a bad memory? 06:32:39 uhm, yes? 06:33:59 :( 06:34:11 And what makes you think it's intentional? 07:01:44 re arke 07:01:48 No, I'm getting ready to go to work. 07:03:10 doh, didn't see him log off. 07:03:27 Oh well, I'm off to work now. Laters. 07:03:32 Bye 07:03:44 --- quit: kc5tja_ ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 07:28:06 --- join: warpzero (~warpzero@mi078.dn183.umontana.edu) joined #forth 08:09:16 --- join: Serg[GPRS] (~z@193.201.231.126) joined #forth 08:09:36 Privet. :) 08:10:15 needs more I440r 08:10:49 hi 08:11:00 i got new tforth this morning 08:11:11 Cool :) 08:11:21 Did you see my inline brainfuck (bf.f) code? ;) 08:11:57 not unzip yet 08:12:25 OK. 08:13:56 i just cleaned monitor w/ water, and it left TERRIFIC traces 08:14:12 use baby wipes 08:14:21 isopropyl alcohol 08:14:31 * Robert uses window cleaning spay 08:14:45 Right, also some isopropyl alcohol (and tons of other stuff) in that. 08:14:48 CRT was all in spit spots - speaking at CRT sucks, but i have to speak sometimez 08:14:48 that'd work =) 08:15:22 ok, tnx 08:15:28 won't it dissolve anti-flare layer ? 08:18:03 hehe, this is the most exact named lang on earth !! 08:22:05 what turn-based game u'd advice me to code after Sokoban ? 08:22:21 (chess are too hard) 08:22:22 Hmmm... good question. 08:22:30 Gomoku? :) 08:22:43 5-in-line ? 08:22:48 Right 08:23:18 nooo ! i'm not ready to code decision tree and a-b pruning in Forth 08:24:13 plus, this takes real trees and lotsa RAM to save calculations 08:25:26 some easy-rest travel/rpg - hard to win, hard to lose, explore and have fun 08:28:18 textmode Pirates ? 08:28:40 but no fencing, just trade, sailing and sea fights 08:30:29 I don't know.. Never played it :)= 08:31:04 i played some modern games ;) w/ original Pirates, i did not cope w/ keys 08:31:38 hm. honestly, when i think of modern in russia, i see relatively ancient hardware 08:32:03 i think there's still this mental picture that russia never really left the 80's or something :p 08:32:12 * qFox hids 08:32:20 qFox: down to Spectrum in poor places 08:33:03 poors live post-nuke life: looting for dead things and making 1 live out of 2 dead 08:33:30 even i did so, if too greedy for really modern 08:34:27 Canon EOS 300 w/ "soapy" 28-80 lens is 300$ - over half of my salary 08:35:11 so i shoot w/ soviet fake-Leica, 10$, 20-30-40 years old 08:35:18 :\ 08:36:02 most likely 40 ;)) no even "outer" meter not to say TTL 08:37:41 but the lens, 50/2 Flectogon clone, beats 28-80 zoom hands-down 08:40:01 DDR "Practica" is 50$ - TTL, 70-80's 08:40:20 but i like RF more than SLR 08:41:54 shot some portraits - razor-sharp at ~1m, easy to catch w/ RF and almost impossible - w/ SLR, for my eyesight 08:44:06 even our "Consumer Photo" magazine issue came out w/ that "FED", fake-Leica M39 rangefinder 08:44:22 on cover 08:44:40 --- part: segher left #forth 08:49:03 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth 08:50:02 Robert: are where words to drop entire data and R stack in TFORTH ? 08:57:42 Hmm... 08:58:09 sp0 sp! <--- reset data stack 08:58:17 rp0 rp! <--- reset return stack 08:58:21 Those should work 08:59:32 aha 08:59:59 i used to use so in main loop and on GOTO from Forth to game and back 09:08:08 "quit" usually checks for stack underflow 09:08:20 Er, I mean, it does unless you change it 09:08:29 heh 09:08:52 i do this to lessen the consequence of errors 09:09:17 i mostly tend to overflow than underflow 09:09:52 I underflow much more often. ;) 09:10:23 The new tforth has another prompt - it prints the data stack before it asks for a line. 09:10:27 Pretty handy 09:10:45 i noticed ;) 09:10:57 i insert depth . in loop while debugging 09:16:59 --- join: I440r_ (~mark4@64.47.44.254) joined #forth 09:25:35 --- join: arke (~arke@melrose-251-251.flexabit.net) joined #forth 09:26:00 Hi. 09:28:45 howdy arke, I440r_ 09:28:46 Hi 09:32:21 hi 09:36:56 hi 09:38:21 bye 09:38:24 btw 09:38:35 i have a surprise for you guys when I'm back :) 09:38:39 br 09:38:41 brb 09:38:44 --- quit: I440r_ ("Leaving") 09:39:10 --- quit: arke ("www.asianpr0n.com admin passowrd is admin! Go go go go go !") 09:42:12 can you make photo shutter out of _one_ metal piece ? 09:42:34 yes 09:42:45 a sheet of metal 09:43:11 just cover the hole 09:43:14 slide over 09:43:36 no, i mean short exposures :) 09:43:55 answer - disbalanced disc w/ sector cut 09:45:41 * Serg[GPRS] is trying to invent home-makeable shutter capable for 1/500 or even 1/2000 10:02:46 --- quit: Serg[GPRS] () 10:24:31 --- quit: tathi (niven.freenode.net irc.freenode.net) 10:24:31 --- quit: kuvos (niven.freenode.net irc.freenode.net) 10:25:46 --- quit: Herkamire ("bbl") 10:25:56 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 10:25:56 --- join: kuvos (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 10:33:28 --- nick: TomasuDlrrp -> Tomasu 11:21:29 --- join: wossname (wossname@HSE-MTL-ppp66652.qc.sympatico.ca) joined #forth 11:25:24 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth 11:33:42 --- join: crc (crc@0-1pool88-42.nas48.philadelphia1.pa.us.da.qwest.net) joined #forth 11:47:30 --- quit: slava ("Leaving") 11:58:59 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 12:02:39 --- part: slava left #forth 12:03:16 --- quit: crc ("http://www.retroforth.org/dev/beta_releases/retroforth-7.beta4.tar.gz") 12:23:52 ahaha 12:23:59 i am listening to this song about lisp 12:24:09 ahahahahha 12:25:52 warpzero: is it sung with a lithp? 12:26:53 no 12:26:57 its awesomely funny 12:27:04 i am disturbing my office i am laughing so hard 12:44:55 --- quit: wossname (":~(") 12:49:52 --- join: crc (crc@0-1pool176-34.nas6.philadelphia1.pa.us.da.qwest.net) joined #forth 12:50:00 --- join: Bullrush (~Bull@196.44.2.197) joined #forth 12:51:59 warpzero: what is it? 12:56:05 --- quit: tathi ("leaving") 12:58:15 --- join: Topaz (~top@spc1-horn1-6-0-cust117.cosh.broadband.ntl.com) joined #forth 13:45:59 chandler: i don't know... the ID3 and filename are mangled 13:46:35 what is the chorus line? 13:46:43 --- quit: crc (Client Quit) 13:49:01 chandler: that god wrote the world in lisp code 13:49:24 oh, "the eternal flame" or something like that 13:49:30 Bob Kanefsky - The Eternal Flame 13:49:31 yeah 13:49:35 this is funny shit 13:49:57 http://www.songworm.com/db/songworm-parody/EternalFlame.html 13:50:31 hahaha 13:50:41 they got the original author to sing it 13:57:06 "[...] in spite of my admiration of the general ethos of the world of Linux, I must condemn the backward, slavish, unimaginative, harmful, 13:57:10 and generally unwashed and ill-informed human-computer interfaces that 13:57:12 -- Jeff Raskin 13:57:14 Linux has sported." 15:20:19 --- join: crc (crc@0-1pool88-17.nas48.philadelphia1.pa.us.da.qwest.net) joined #forth 15:34:58 --- quit: Topaz ("Leaving") 15:36:31 --- quit: Bullrush ("Download Gaim: http://gaim.sourceforge.net/") 15:49:09 --- join: wmg (~weldon@c208-37-6-201.ip.newc.com) joined #forth 15:55:31 --- quit: Klaw (Success) 15:56:39 Herkamire: that didn't copy quite right 16:00:52 "[...] in spite of my admiration of the general ethos of the world of 16:00:56 Linux, I must condemn the backward, slavish, unimaginative, harmful, 16:01:00 and generally unwashed and ill-informed human-computer interfaces that 16:01:03 Linux has sported." 16:01:07 -- Jeff Raskin 16:09:02 --- join: blockhead (default@dialin-590-tnt.nyc.bestweb.net) joined #forth 16:19:45 --- quit: qFox ("meh ircii's netsplit detection appearantly sux ^^") 16:26:39 --- part: crc left #forth 16:27:05 --- part: wmg left #forth 16:43:00 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 17:03:08 --- quit: blockhead ("FOOMP!") 17:38:19 --- quit: tathi ("leaving") 17:42:21 --- quit: warpzero ("Tried to warn you about Chino and Daddy Gee, but I can't seem to get to you through the U.S. Mail.") 18:22:25 --- join: LOOP-HOG (~jdamisch@sub22-119.member.dsl-only.net) joined #forth 18:28:14 --- quit: LOOP-HOG () 19:06:55 --- join: doublec (~doublec@coretech.co.nz) joined #forth 19:39:53 --- quit: kuvos (niven.freenode.net irc.freenode.net) 19:41:16 --- join: kuvos (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 19:42:36 --- join: Sonarman (~matt@adsl-66-124-254-118.dsl.snfc21.pacbell.net) joined #forth 19:48:55 --- join: lalalim_ (~lalalim@p508AAC0D.dip.t-dialin.net) joined #forth 19:49:09 --- quit: lalalim (Read error: 110 (Connection timed out)) 20:06:20 --- quit: doublec ("Leaving") 20:07:48 --- join: kc5tja (~kc5tja@66-74-218-202.san.rr.com) joined #forth 20:07:53 --- mode: ChanServ set +o kc5tja 20:18:06 --- quit: kuvos (niven.freenode.net irc.freenode.net) 20:25:39 One of my friends is recommending we move the #forth channel off of this network and onto a new IRC network that he's creating from scratch. 20:25:41 --- join: ayrnieu (~julian@205.241.56.30) joined #forth 20:26:20 re ayrnieu 20:26:40 --- join: kuvos (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 20:26:43 (to repeat for the benefit of ayrnieu) One of my friends is recommending we move the #forth channel off of this network and onto a new IRC network that he's creating from scratch. 20:26:49 What is the opinion of the folks here on this? 20:27:03 hello, kc5tja. 20:27:17 it wouldn't be much of an inconvenience or anything, but what's wrong with this network, aside from the occasional netsplits? 20:28:01 we'd be (temporarily) losing the benefits of clog and cmeme 20:28:22 That's what I asked. His response was something along the lines of, "No need to deal with lilo, less politics, and an attempt at tighter quality control, thus yielding what we hope to be a better user experience for everyone." 20:28:52 I'm talking to him about getting something like cmeme on the new server, offered as an intrinsic service of OCCN. 20:29:10 ooh, nice 20:29:20 OCCN is the network's name? 20:29:29 Yeah. 20:29:42 Open Collaborative Communications Network 20:30:46 I have a slight preference for staying here 20:31:04 i have no preference either way, really 20:31:22 as long as the logger is ready before we move :) 20:32:51 Personally, I don't care what we do. 20:33:22 I already told him I see no need to move, and that I think a move would disrupt the currently excellent participation we have had these past couple of weeks. 20:35:31 --- join: MrReach (~chatzilla@209.181.43.190) joined #forth 20:35:31 --- mode: ChanServ set +o MrReach 20:35:39 well, as long as it's decided at least semi-democratically, i.e., not on the whim of one person, then i'm happy with whatever happens 20:36:16 That's why I sought the opinions of those here. 20:36:19 * MrReach waves 20:36:25 re MrReach 20:39:41 --- quit: juhammed (Read error: 104 (Connection reset by peer)) 21:35:25 Morning MrReach, kc5tja, Sonarman 21:36:10 * Robert _woke up_ at 06:30, for a change. 21:36:50 re 21:39:46 --- quit: MrReach ("ChatZilla 0.9.61 [Mozilla rv:1.7b/20040421]") 21:40:52 * ayrnieu woke up at 5:45, but he didn't want to =) 21:41:21 kc5tja - curiously, do you still deal with J at all? Or do you still find it interesting? 21:42:12 I find it very interesting, but as of late, I haven't worked with it. 21:42:49 ah, OK. 21:43:30 I've probably forgotten much. 21:43:38 ayrnieu: I'm so proud - this is my usual bedtime. ;) 21:43:58 What is J, anyway? 21:45:17 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 21:45:20 hi all 21:45:30 Hi slava 21:45:48 i think for writing network code, Forth has an advantage over C, just because a skilled Forth programmer will read the source very quickly, and this means it will be easier to audit for potential security problems 21:45:59 Robert: It's the latest heir to the APL programming language legacy. 21:46:12 you could write a couple of array words that do bounds checks and use these for all untrusted input 21:46:14 kc5tja: Oh, OK. 21:46:20 in C this adds a lot of syntax overhead 21:46:34 kc5tja, will we see K for kestrel? :) 21:46:37 kc5tja, or APL? :) 21:46:51 Robert: It is an array-based programming language, that can express some mightily compact programs because it largely does away with explicit iteration, AND program constructs are replaced with symbols. 21:47:04 slava: I don't know enough about the languages to offer a port. 21:47:06 kc5tja, how would the current cpu design fare with such code (if it was compiled with a good compiler) 21:47:31 kc5tja, i've always wanted to develop a vector language, with math stuff only, with apl-like syntax. 21:47:51 kc5tja, but very simple and forth-like (but infix syntax for everything) 21:48:01 slava: I don't know. I'm building the processor for Forth, so I expect it to run moderately OK as long as you can fit the stuff into a Forth perspective. 21:48:07 Maybe I should look it up... After breakfast and a shower, anyway. 21:48:27 However, vector processing is the reason why I provided three address registers -- the intention was to have address registers for two sources and one destination pointer. 21:48:31 kc5tja, since it would have to emulate FP anyway, performance is not a top concern :) 21:48:48 or have fixed point as well maybe? 21:48:54 slava: Sorry, but you are not correct. Graphics represents an enormous amount of vector work. 21:49:11 kc5tja, true 21:49:45 A later generation of the hardware will offer additional hardware support for vector operations. 21:50:00 i'm trying to figure out the unix system calls for networking... 21:50:21 so does one use socket(2) or something from section 3 of the manual? 21:50:53 Yes, but then you have to configure the socket. 21:50:57 It's hard to describe. 21:51:16 Your best bet is to learn how to do sockets programming in C first, then transcribe that experience to Factor. 21:51:16 i should look at the source to a tcp/ip app. 21:51:16 networking is a PITA on linux 21:51:31 kc5tja, in freebsd, everything is in /usr/src its nice 21:51:45 Herkamire: no, Sockets programming is a PITA, no matter what platform you're on. It's intrinsic in the nature of sockets itself. 21:51:53 slava: Wha? 21:52:09 ok 21:52:13 I've never done it elsewhere 21:52:16 kc5tja, i'm only doing a very simple api (connect to a host name and port number, or listen on a local port number) 21:52:40 so its just a matter of stringing together some libc functions. 21:53:34 slava - on linux, you'll have to use a 'socketcall', a syscall that then calls a socket-related kinda-second-class-syscall. 21:53:55 slava - oh, easier if you can use libc, yes. 21:55:46 here's the C code to open a tcp socket: http://jason.herkamire.com:5000/svn/jjj/socket.c 21:55:56 Herkamire, thanks 21:56:09 and functions to check if you can read or write to it without it blocking 21:56:12 slava - the GNU C Library Manual also has TCP Server and TCP Client examples. 21:56:19 i don't need nonblocking i/o for now. 21:56:25 ayrnieu, gnu libc ? 21:56:34 slava - no, the manual. 'info libc'. 21:56:50 ayrnieu, ain't no gnu libc here :) i'm using bsd, all it has is manpages for each function :) 21:57:09 slava - that doesn't matter; 'info libc' has TCP server and TCP client examples :-/ 21:57:10 bsd has its own c library? 21:57:12 info is only emacs and gcc docs pretty much 21:57:27 yes, it has its own libc, its own utilities in /bin, and its own tcp/ip stuff in /usr/bin 21:57:36 the latter shows up in linux distros (ftp, telnet, finger, ...) 21:57:48 and /etc is very nicely organized 21:58:05 i thought it used glibc 21:58:21 no 21:58:25 i have glibc installed though 21:58:32 only binaries, not docs, in /compat/linux/ 21:58:43 slava - as for Forth socket APIs, please just remember to implement a read(2)-like word of some kind. Some languages (GHC's extensions to Haskell) get confused and pretend that sockets work exactly like files. 21:58:43 slava: That's a pretty narrow-minded view of what info is. 21:58:48 Info is a lot more than those two things on my box. 21:58:49 if you run a linux program, it uses shared libraries from that directory (its basically linux libc, ncurses, X11, etc) 21:58:55 kc5tja, i know what info is 21:59:00 And I'm going to use GNU Info format to document the Kestrel and write its developer/user manuals. 21:59:16 kc5tja, but most stuff in info on a linux box is fileutils, findutils, textutils,... gnu stuff, which has other bsd alternatives 21:59:25 kc5tja - yay, I've always liked Texinfo. 21:59:33 it seems archaic to me 21:59:38 gforth uses info 21:59:39 i like docbook for pdf and html output 21:59:49 info has links 21:59:55 * ayrnieu fairly despises docbook, for no sane reason. 21:59:57 my info has loads of documents in it 22:00:12 ayrnieu, the tool chain is a nightmare to set up. 22:00:31 slava: I wanted to use docbook, but the damn thing is so freaking complicated that it's actually far less work for me to use TexInfo. Like, we're talking getting Texinfo sources up and going in minutes, not days. 22:00:34 ayrnieu, i use a mix of xslttools, openjade, pdftex... 22:00:48 And Texinfo puts out HTML too. 22:00:48 kc5tja, setting up openjade is the tough part. 22:00:49 398 lines 22:00:54 even has an entry for "factor" 22:00:58 latex and texinfo definitely appeal to me -- but since I dislike docbook for no sane reason, I won't argue =) 22:01:01 kc5tja, its a freaking scheme interpreter with its own ideas of how it should be used. 22:02:33 kc5tja, you should use a wiki for kestrel docs :) 22:02:54 and make it a simple format so they can be viewed online in the kestrel. 22:03:15 wiki in blocks? 22:03:41 My order of preference is TeX, because it produces beautiful printed output, then DocBook (not quite as beautiful, but workable), but then I realized that I also wanted a plain-text copy of all my books too. 22:03:46 or have it shown as one big file, with hyperlinks that jump around -- like the humane interface. 22:04:13 slava: :) 22:05:05 slava: Hey! Thanks for the idea! I might put that in VIBE 3.0 for the Kestrel. 22:05:14 (and for gforth for that matter too.) 22:05:24 Not sure how to represent Wiki links though. 22:05:48 kc5tja: what about texinfo? 22:06:03 Sonarman: Huh? 22:06:10 texinfo for your docs 22:06:15 Sonarman: Didn't I say that's what I was using? 22:06:30 * Sonarman shuts up :) 22:06:40 :) 22:14:24 * ayrnieu grumbles at MOPS for distinguishing between Enter and Return (and ignoring the convenient one) 22:14:32 --- join: Serg (~z@212.34.52.140) joined #forth 22:14:50 hi ! 22:15:37 howdy, Serg 22:15:50 sleepy ;( 22:16:07 today gonna go to "pocketers" meeting 22:16:21 Me, too, but I won't notice it until this MSG-ridden Pu-Erh tea wears off =) 22:16:42 what does a "pocketers" meeting involve? 22:17:19 what means "pocket thief" for 99% of folks and "palm+PPC" for the sanctified 22:17:47 ah =) 22:18:07 Palm and PPC? That seems like an odd combination. 22:18:10 That still doesn't make any sense to me. That doesn't even parse to me. 22:19:52 everything handheld what can be named a computer 22:21:59 ayrnieu, what is MOPS like? 22:22:03 ayrnieu, is it very OOP? 22:23:13 slava - I haven't really gotten deep into it, yet, due to the annoying interface. I remember that it claimed much OO-ness. 22:25:21 ayrnieu, what is the forth itself like? is it bloated? 22:27:06 I *really* haven't gotten deep into it =) 22:27:30 slava: Well, it's designed to run under MacOS, and to write MacOS applications, so I'm sure there is sufficient bloat in it to support at least that much. 22:27:38 * ayrnieu likes to play around with '1 1 + .' before he graduates to reading documentation, and he'd rather look around for nicer forths before continuing with mops right now. 22:29:33 ayrnieu, are you on a 32-bit powerpc? 22:30:14 slava - on a G3 iMac with MacOSX, yes. 22:31:24 ayrnieu, what byte order is powerpc? 22:31:32 I really should finish porting herkforth to ppc 22:31:39 I mean to mac os X 22:31:49 slava: big-endian 22:31:58 slava: actually, the chip can do either endian 22:32:15 but mac os and linux ppc use big-endian 22:32:23 Actually, even though it can do both, it *really* prefers big-endian. 22:32:38 kc5tja: that explains why linux ppc uses big-endian :) 22:32:51 Yep. 22:32:56 kc5tja: how? is it faster? or just more instructions support it? 22:33:17 More instructions support it, and those that do support little-endian, have caveats to their use. 22:35:12 ayrnieu, i think then if I byte-swap a factor image, it should work on macos no? 22:35:22 ayrnieu, the C part only uses standard library stuff 22:35:31 and the image is relocatable 22:35:52 but its read into memory and becomes the heap in one fread() 22:38:59 I wouldn't know, sorry. I've used MacOSX for 4 days, now, and don't have Xcode or gcc or such niceties, yet. I've mostly downloaded binary systems -- OpenMCL, J, etc. 22:39:48 slava: byteswapping files generally only works if the entire file is all 32-bit entities, (or all 16-bit) 22:39:48 ayrnieu, ok. 22:39:59 if there's any text or anything then it won't 22:40:08 Herkamire, the text is 16-bit characters 22:40:36 is everything in the file 16-bit? 22:40:46 funny, Forth Inc used to have a 'Power MacForth'. 22:49:30 Hmm...having some difficulty designing conditional control flow into the new TTA architecture. 22:49:44 Everything else I have down to one, maybe the occasional two, clock delays. 22:50:06 why did they make the CPU support little endian? 22:50:25 Probably political pressures. It's hard to say. 22:51:34 --- join: wmg (~weldon@bgp02689673bgs.flrdav01.dc.comcast.net) joined #forth 22:54:17 anybody got tips for making an x86 get over its aversion to writing executable sections? 22:54:38 wmg: Yes. Map a data segment on top of a code segment. 22:55:03 clever 22:55:45 It's the only way to do it in protected mode. 22:56:16 is there any way to just load the entire program as a data or bss segment? 22:56:23 Note I don't recommend mixing data and code anyway, if you care about performance of the code. 22:57:01 kc5tja, for a JIT would it be wise to store generated code in a completely different heap? 22:57:28 BSS segments are uninitialized variables space (Blank Segment Space); it's not actually recorded in the executable file -- only its size is. 22:57:34 well I only will need to do this when my forth is being run interactive rather than in batch mode; so a performance hit would be expected 22:57:39 slava: Ideally, yes. 22:58:06 wmg: Frankly, the performance hit isn't really important anyway unless you're doing some serious number crunching. 22:58:45 wmg: What OS are you developing for? 22:58:52 eh, it's going in a video game ultimately. But the dictionary won't be changed so I won't need to write to the text segment 22:59:20 Linux right now; I've got about half of the BSD port 22:59:30 I'm kind of adding interactivity as an afterthought 23:00:01 OK, I'm not sure about how BSD handles the differences between code and data sections, but I do know that Linux will allow writes to the text section if you map it as writable. 23:00:17 If not, there is some kind of weird syscall you can make to change that. 23:00:34 that's what I gathered from poring over gnu as's alleged documentation 23:01:02 BTW, both Linux and BSD map the code and data segments to the entire 4GB address space. So if you find your text section to be unwritable, it's because of how the page tables are set up, not the segment tables. 23:01:41 Thanks 23:01:47 Thus, you'd need to use (I *think*) mmap() to change the permissions. 23:01:54 I'd have to research it to be positively sure though. 23:01:56 I440r would know. 23:02:04 bleh. Maybe interactivity is a bridge too far 23:02:53 Well, like I said, I found that setting the .text section's permissions to writable in the ELF executable does let you modify code in the .text section at run-time (on my box at least) 23:03:11 in herkforth, (don't know if this is relivant, it being PPC, but it is linux) 23:03:20 I just specified the segment in the ELF file to be rwx 23:03:23 I think I440r said that the X permission for the .text section is ignored on BSD. 23:03:42 err, the W permission, rather. 23:03:50 Obviously, .text should have X set. :D 23:03:55 one would hope so 23:04:09 :) 23:05:18 the other thing I like about batch mode is it means I don't have to implement a control stack 23:06:30 ? 23:06:38 why would you have to implement a control stack? 23:06:58 I just use the data stack 23:06:59 OK, I have to get to bed now. night. 23:07:07 goodnight kc5tja 23:07:08 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 23:07:16 later kc5 thanks for the advice 23:07:28 Herkamire: "if" needs some portion of memory to affect 23:11:00 wmg: data stack 23:11:09 : if here 0 , ; immediate 23:11:27 eh? 23:12:52 : then here over - relitive-branch swap ! ; immediate 23:12:54 it has to mark a branch somehow; how does yours do that? 23:13:10 "here" 23:13:11 slava - do you have Factor online? 23:13:39 : relative-branch ( offset -- branch-instruction ) 23:13:40 that still leaves here on the data stack 23:13:48 oh but the then clears it 23:13:49 I see 23:13:51 ayrnieu, factor.sf.net 23:13:58 slava - ah, thanks =) 23:14:02 wmg: yep. you can use the data stack as the control stack 23:14:55 --- part: wmg left #forth 23:17:09 ayrnieu, i have some big endian support 23:17:14 ayrnieu, i'm not sure if it works 23:17:27 ayrnieu, there is a switch in the image generator 23:25:49 do you have an idea: what turn-based game to code after Sokoban ? (chess etc. - not interesting) 23:26:35 Serg - if you've written Sokoban, and if you've done it in the textual way that, e.g., the sokoban that comes with gforth does, then I'd suggest making it a bit more complex and writing Lolo =) 23:28:38 what's Lolo ? 23:31:33 --- join: ball (~ball@dialup-4.158.165.198.Dial1.Chicago1.Level3.net) joined #forth 23:32:21 Serg: minesweeper? 23:33:03 Serg - nintendo game. http://tbd.yi.org/nes/ seems to only have Lolo 2 and Lolo 3 =( 23:34:37 --- part: ball left #forth 23:34:51 sweeper suxx 23:35:13 i wanna break loose out of 20x20 boundary, write the exploration game 23:35:51 addictive, hard either to get hight score or to die completely 23:36:16 but turn based? 23:36:19 like nethack? 23:36:24 yeah 23:36:44 sure, nethack :) but rather open-land than dungeon 23:36:53 Morning Serg 23:36:56 hi 23:37:38 maybe, RPG where experience = health ? 23:37:46 slava: factor starts up on my PPC after the image has been run through a byte swapper, but this is the sign-on message i get: aFtcro.0066.oCypirhg tC( )0230 ,0240S alavP seotnEet r``ehpl''f roh le.pnEet r` :P 23:38:07 Serg - why not first write a sokoban-map editor, translators for common online descriptions of sokoban levels to your game's database, a generic rule-based configuration system that can encapsulate sokoban maps with specialized sokoban rules, such as RULE: "@X" " X" msg" You've been destroyed!" game-over ; ? 23:38:37 monsters beat you but not to death - they lose interest after beating you below some EXP 23:39:08 ayrnieu: i feel claustrofobic in a puzzle grid 23:40:09 Sonarman, LOL 23:40:13 Sonarman, there is a better way to do it 23:40:25 Sonarman, if you have java, you can bootstrap a big endian image 23:40:45 huh, Factor starts with no fuss, here -- but maybe I miss something. 23:40:52 ayrnieu, java or C? 23:41:00 slava - Java. 23:41:03 ayrnieu, if you run Factor.jar it will work on ppc :) 23:41:08 ayrnieu, try this 23:41:18 in the java factor: "big-endian" on make-image 23:41:26 this will create a file native/factor.image 23:41:33 though the graphical listener starts without the initial greeting... which goes to the tty. 23:41:49 me very sleepy 23:41:51 ayrnieu, can i have a screenshot of the listener on oS X? 23:41:51 --- quit: Sonarman ("leaving") 23:42:12 i imagine its very sluggish. 23:42:14 slava - sure. Just, um, let me figure out how to do that. 23:42:19 slava - no, not sluggish at all. 23:42:23 well, so far. 23:42:24 native factor is much snappier -- it starts instantly 23:42:32