00:00:00 --- log: started forth/02.08.09 00:09:56 --- quit: proteusguy (Connection timed out) 00:35:01 --- join: goshawk` (goshawk@panix1.panix.com) joined #forth 00:53:57 --- quit: goshawk` (Killed (NickServ (Ghost: goshawk!goshawk@panix1.panix.com))) 00:53:58 --- join: goshawk` (goshawk@panix1.panix.com) joined #forth 01:00:40 --- quit: goshawk` ("No windows for this server") 01:10:41 --- join: FareAway (fare@samaris.tunes.org) joined #forth 01:23:33 --- join: goshawk` (goshawk@panix1.panix.com) joined #forth 01:29:23 --- part: goshawk` left #forth 01:39:30 --- join: goshawk` (goshawk@panix1.panix.com) joined #forth 01:43:58 --- join: Grunt (~nicad@12-245-48-122.client.attbi.com) joined #forth 01:44:34 What's new in Forth land? 01:51:19 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 01:53:02 howdy grunt 01:54:31 lots is new in forth land 01:55:01 new forth implementations, new ideas w/ forth, etc 01:55:25 I got a couple Forth interpreters. I'm curious 01:55:46 it is 3 am right now for me 01:55:49 and i'm going to bed 01:55:56 How difficult is it to integrate Forth proggies with stuff written in another language? 01:56:06 but stick around and you'll be able to discuss forth 01:56:06 5am here. 01:56:11 * Grunt nods 01:56:20 it's not too difficult 01:56:24 check out FICL 01:56:28 * Grunt nods 01:56:45 FICL was designed for integration with other langs i think 01:56:49 i'm not too sure 01:56:54 tired and all :) 01:57:03 futhin: Sleep well. 01:57:07 good night :) 01:57:10 * futhin is afk sleeping 02:11:07 --- quit: proteusguy (Read error: 110 (Connection timed out)) 02:11:55 --- quit: Grunt ("This isn't Happy Hour!") 02:26:22 hey, futhin u still here? 02:51:58 --- nick: FareAway -> Fare 03:35:08 --- join: Speuler (l@62.206.56.189) joined #forth 03:35:30 hi 03:51:33 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 04:30:38 --- part: Speuler left #forth 04:34:32 --- join: cleverdra (julianf@0-1pool37-206.nas2.florence1.sc.us.da.qwest.net) joined #forth 04:44:22 hi 04:48:51 hello onetom 04:55:23 --- quit: Soap` (Read error: 104 (Connection reset by peer)) 05:10:36 --- quit: proteusguy (Read error: 110 (Connection timed out)) 05:26:23 --- join: Serg_penguin (~Z@nat-ch1.nat.comex.ru) joined #forth 05:26:31 hi 05:34:22 eh, everybody seems busy.. :( 05:51:17 hi 05:51:25 busy, sorry... 06:24:53 --- quit: Serg_penguin () 06:50:56 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 07:11:49 --- quit: proteusguy (Connection timed out) 07:31:08 --- join: Serg_penguin (~Z@nat-ch1.nat.comex.ru) joined #forth 07:31:20 hi 07:31:32 --- quit: Serg_penguin (Client Quit) 07:40:26 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 07:59:28 --- join: kc5tja (~kc5tja@ip68-8-206-226.sd.sd.cox.net) joined #forth 07:59:54 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth 08:11:17 --- quit: proteusguy (Connection timed out) 09:11:34 --- quit: proteus (Read error: 110 (Connection timed out)) 09:31:01 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 09:45:00 --- join: ictx (ictx@sdn-ar-002flflauP177.dialsprint.net) joined #forth 09:56:31 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth 10:09:20 --- quit: proteusguy (Read error: 110 (Connection timed out)) 10:09:21 --- quit: ictx ("") 10:29:18 --- join: I440r (~mark4@1Cust199.tnt3.bloomington.in.da.uu.net) joined #forth 10:35:21 --- quit: cleverdra (Read error: 104 (Connection reset by peer)) 10:35:26 --- join: ayrnieu (julianf@0-1pool37-206.nas2.florence1.sc.us.da.qwest.net) joined #forth 10:36:50 --- nick: ayrnieu -> cleverdra 10:40:51 hi ppl 10:41:13 Howdy 10:42:39 im sorting out the instruction tables into all XYZ r1, r1 and xyz mem, reg or reg,mem 10:42:50 so all instructions with the same addressing mode are next to each other 10:52:30 kc5 u recon i need to have segment overrides ? 10:52:44 they dont realy make sense in linux do the? 10:52:44 they 10:52:50 can you modify a selector ? 10:52:57 Not really, and they're easy to implement if you need them. 10:53:17 and is the lock prefix needed ? 10:53:20 what is lock for ? 10:53:25 You can, but Linux doesn't define any method for creating new selectors 10:53:37 Lock is used to lock the bus while updating memory. 10:53:40 It's for SMP systems 10:54:16 so you would MANUALLY put a lock in your code if you were coding for smp ? 10:54:24 lock has been arround longer than smp tho 10:54:35 unless you call the old fpu smp heh 10:54:42 No, lock has nothing to do with the old FPU. 10:54:51 Intel has been builting multi-CPU boards for a lot longer than the PC existed. 10:54:57 aha 10:55:35 Which is why this whole SMP craze is nonsense. The older CPUs could handle the required bus access protocols just as well as the modern CPUs. 10:55:39 sop you think ishould have that one supported by default then, im not looking for a easy way out with these, they are to easy to implement, i just want to minimalize cruft heh 10:56:11 Lock might be handy, but I don't see any reason to use a segment override. 10:56:21 tswhat i was thunkin :) 10:56:28 And again, if you really need a segment override, : cs: $2E C, ; 10:56:49 lock and rep will be the only ones used then :) 10:57:08 --- join: empirion (~empirion@207.6.240.106) joined #forth 10:57:13 * kc5tja nods 10:57:27 greets 10:57:40 I supported the whole gamut of prefixes because the general mechanism was the same, and besides, I'll likely be using them when I implement FS/Forth on native hardware (ring 0-level stuff). 10:57:48 hello empirion :) 10:57:49 re empirion 10:57:54 do you forth empirion? 10:58:17 hi emp! 10:58:22 I'm starting forth on the long quest. 10:58:30 ;) 10:58:35 yea - you might do ring zero (actually, isforth might eventually too :) 10:58:42 long quest ? 10:59:25 Actually i'm just a c hacker, who's looking for a good beginners tutorial on forth. 10:59:34 Any suggestions? 10:59:53 Yes. 10:59:55 * kc5tja finds a URL 11:00:06 aha any c hacker who will look at forth and no go screaming for mama is alright by me :) 11:00:31 http://www.geocities.com/SiliconValley/Bay/2334/4thtutor.htm 11:00:42 no no no, you gotta start off with an intro ;P 11:00:48 lemme get some urls 11:00:54 That's a tutorial for a derivative of Forth called 4tH, but it's close enough to work with most ANSI Forths. 11:00:59 futhin: That IS an intro! 11:01:06 naw, it sucks :P 11:01:19 In what way does it suck? 11:02:06 it's not the best tutorial out there iirc 11:02:13 If we had an online copy of Starting Forth by Leo Brodie, I'd surely use that. But that (a) doesn't exist, and (b) we don't have permission to do it ourselves. 11:02:26 i emailed leo brodie, never got a response 11:02:31 PRECISELY! 11:02:45 He never responds to e-mails about such matters. 11:02:52 He doesn't care about Forth anymore. 11:02:56 somebody with a scanner should scan it up and post it online 11:03:07 If I had my copy of the book, I might consider doing that. 11:03:15 and wait to see if leo brodie responds 11:03:33 It's probably still in the publisher's hands. 11:03:45 Starting Forth isn't the greatest tutorial either, but it's pretty good i guess 11:03:55 well 11:03:57 --- quit: I440r (bear.openprojects.net irc.openprojects.net) 11:03:57 --- quit: proteus (bear.openprojects.net irc.openprojects.net) 11:03:57 --- quit: Fractal (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: onetom (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: empirion (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: goshawk` (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: Mongrel (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: cleverdra (bear.openprojects.net irc.openprojects.net) 11:03:58 --- quit: ChanServ (bear.openprojects.net irc.openprojects.net) 11:03:58 It's world class! 11:04:01 it's the best out there i mean 11:04:13 ugh 11:04:17 beautiful netsplit 11:04:19 What do you want from a language that you can learn in 15 minutes? 11:04:33 i'm gonna organize my forth links 11:04:39 i need to find my favorite tutorial etc 11:05:10 i forget which tutorial and which intro etc 11:09:30 --- join: Mongrel (~ant@ool-435249eb.dyn.optonline.net) joined #forth 11:09:30 --- join: goshawk` (goshawk@panix1.panix.com) joined #forth 11:09:30 --- join: empirion (~empirion@207.6.240.106) joined #forth 11:09:30 --- join: onetom (~root@novtan.bio.u-szeged.hu) joined #forth 11:09:30 --- join: I440r (~mark4@1Cust199.tnt3.bloomington.in.da.uu.net) joined #forth 11:09:30 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth 11:09:30 --- join: Fractal (ajta@h24-77-171-228.ok.shawcable.net) joined #forth 11:10:03 --- quit: proteus (Connection timed out) 11:10:23 echo "here" 11:10:26 --- join: cleverdra (julianf@0-1pool37-206.nas2.florence1.sc.us.da.qwest.net) joined #forth 11:10:46 ? 11:10:54 heh netsplat 11:11:08 just seeing if anyone wasn't affected by the netsplit. 11:11:14 and that should be .( here) 11:11:15 heh 11:11:27 :) 11:11:32 ,( is a comment to be echoed to the screen 11:11:40 it echos everything up to the closing ) 11:11:44 ( is similar 11:11:52 ( everything in here is ignored by the compiler ) 11:11:58 not echoed, ignored 11:12:13 \ this is a comment to the end of the line 11:12:14 in forth? 11:12:17 yes 11:12:25 ( and ) are usually used for a special sort of comment 11:12:35 but alot of people use them for normal commenting too 11:12:38 heh, for a moment I was asking myself what os your useing ;) 11:12:40 its for a stack comment 11:12:51 ( n1 n2 n3 --- ) 11:12:53 thats a stack comment 11:13:03 n1 n2 and n3 are on the stack when you enter the word 11:13:10 n3 is at the top 11:13:17 nothing is on the stack when you exit the word 11:13:23 ( n1 n2 --- n2 n1 ) 11:13:28 thats a stack comment for swap 11:13:37 coolio 11:13:50 left of the --- is stack picture on entry, right of --- is stack picture on exit 11:14:04 in forth you can have as many return values as you want 11:14:17 ( --- n1 n2 n3 n4 n5 etc etc etc ) 11:14:24 no parameters passed but alot returned :) 11:14:34 n or n1 usually means a number 11:14:39 c or c1 means a character 11:14:45 a or a1 etc mean an address 11:14:47 or you can do 11:14:57 ( address length --- ) 11:15:23 that doesnt actually put a name on the parameters, you cant refer to them as "address" or "length" in the code 11:15:26 that's messed heh 11:15:33 but it gives them a description 11:15:45 how messed up ? :) 11:16:06 oh no 11:16:08 in terms of c logic. 11:16:10 ;) 11:16:16 you are perverting his mind with stack comments! 11:16:23 lol 11:16:40 empiron FORTGET c logic, its an oxymoron heh 11:17:00 lol, i'll try. 11:17:01 Yes, let's insult a C programmer who's trying to learn Forth. Good idea. 11:17:08 kc5 lol 11:17:21 And people wonder why Forth has so few members in its active community. :) 11:17:28 emp your at the most difficult part of learning forth 11:17:41 empirion: the mistake lots of forth coders make is to think too much about the stack. forget the stack, ignore the stack, pretend that implicit data passing is the _main_ feature, not a side effect of the stack 11:17:42 kc5 if he gives in now he doesnt deserve to know forth :) 11:17:52 emp just kiddin :) 11:18:05 everyone should be forced to learn it :) 11:18:13 after spending 10 years of coding nothing but asm :) 11:18:35 brb 11:18:36 emp, c places alot of restrictions on what you can do and when 11:18:38 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth 11:18:41 forth places NON on you 11:18:44 hi proto 11:18:54 I440r - re your earlier comment about organization of x86 instructions: http://www.sandpile.org/ia32/index.htm 11:19:02 I440r - NON? 11:19:24 I think he meant to say none. :) 11:19:32 erm none 11:21:04 oh, I didn't follow that from the previous assertion. 11:24:53 back. 11:25:15 wb 11:25:35 thanks. 11:25:50 so basically what use is the stack in forth if you don't need it? 11:26:02 implicit data passing 11:26:08 that's the main feature you use it for 11:26:35 like you have a word that takes two parameters off the stack and returns one 11:26:39 1 2 + 11:26:43 --- quit: proteusguy (Read error: 104 (Connection reset by peer)) 11:26:53 so 3 is on the stack 11:26:56 then you have other words 11:26:57 that use 3 11:27:19 : square dup * ; 11:27:25 1 2 + square 11:27:33 ah, ok, so you have to specify implicitly what goes on the stack? 11:27:37 there's kind of a data passing between the word + and the word square 11:27:38 empirio - futhin is trying, I think, to preemptively attack the, he thinks, common perception that the Forth stack is a big confusing deal. 11:27:50 cleverdra: that's correct 11:28:15 lots of people get too involved with the stack 11:28:20 : ^2 dup * ; 11:28:21 they'll put wayy to much items on the stack 11:28:32 and then they'll get all confused about how to deal with the items 11:28:46 and they'll use a whole bunch of really slow words 11:28:49 like OVER 11:28:52 or PICK 11:28:57 heh 11:28:59 4 items on teh stack is a soft max for me, i might go as many as 5 or even 6 but after that you can forget it 11:29:13 PICK lets you get a stack item from a depth that you specific 11:29:14 er 11:29:15 specify 11:30:49 --- join: XeF4 (Tipula@12-245-116-85.client.attbi.com) joined #forth 11:30:49 I know this might be a stupid question but does forth allow any unix or socket functions or is it purely in and its own self. 11:30:49 Note that these "maximum depths" for the stack are local to a word. 11:30:57 empirion - it's not a stupid question, and the answer is yes and no, respectively. 11:31:02 In reality, you can have 18 different numbers on the stack, but you don't care about the bottom-most 16 or 15. Only the top-most elements should matter. 11:31:05 empirion: heed chuck moore's recommendations 11:31:06 1) a word should not be more than 1 or 2 lines of code 11:31:06 2) a word should not pass more than 2 parameters by stack 11:31:06 3) try to only manipulate 3 items on the stack 11:31:06 if there are more than 3 items on the stack that you need to manipulate, try solving the problem in a different way - the end results will be better 11:31:32 yes just because the stack comment says ( n1 n2 n3 --- ) 11:31:38 doesnt mean there are only 3 items on the stack 11:31:48 it means that this word only uses 3 items 11:31:58 * cleverdra becomes annoyed. 11:32:01 the stack might actually be ( a1 a2 a3 n1 n2 n3 --- ) 11:32:13 with what ? 11:32:13 cleverdra: about the stack comments? 11:32:23 i440r: yeah, shush on the stack comments :P 11:32:30 Speaking of making a mountain out of an anthill, I think we're beating a dead horse here. :) 11:32:35 futhin: alright, so just for as a good forth programming rule no one should really be concernet with more than 3 items on the stack? 11:32:52 empirion: correct.. the stack isn't an array :) 11:32:54 *concerned 11:32:55 empirion - the most useful answer depends on your particular implementation, but it's generally yes. I've written servers and clients in gforth, using libc, and in isforth, using linux syscalls. 11:33:04 kc5tja - precisely =) 11:33:04 empirion: if you want an array, you can CREATE one 11:33:26 cleverdra: :) 11:33:47 cleverdra: ah, ok. And how generally would you do that? 11:34:00 empirion: Depends on the Forth implementation. 11:34:33 futhin: ok, so basically the language doesn't rely on arrays and pointers like c? 11:34:37 empirion: Some Forths allow you to link against existing C libraries, while others prefer to call system calls directly. Forth environments that sit on bare hardware will, of course, directly drive the networking equipment. 11:34:57 empirion: there are pointers, arrays, linked lists, and other beautiful stuff 11:35:03 kc5tja: ok, got it. 11:35:21 empirion: but often you have to make your own data structures on the fly, when you need them for your program 11:35:41 I hope i'm not sounding like too much of an idiot asking all these questions, but when all you program and know is c your views are somewhat limited. 11:36:00 I started programming Forth back when I was a teenager; at least 8 years ago. 11:36:13 I'm only now starting to come to grips with the inherent philosophy behind Forth. 11:36:24 --- quit: cleverdra (bear.openprojects.net irc.openprojects.net) 11:36:26 heh 11:36:41 So don't be ashamed or embarrassed if you "just don't get it." 11:36:48 it will seem like a bit of a pain in the ass to make your own data structures, but for experienced forth coders, it is a good thing, because you create data structures that you need, there is less bloat and it is more customized to fit your tasks 11:36:49 --- join: ChanServ (ChanServ@services.) joined #forth 11:36:49 --- mode: bear.openprojects.net set +o ChanServ 11:36:51 alright, thanks :) 11:37:07 Even Chuck Moore himself had refined his programming methodologies over the course of 15 years before he finally arrived at a codified philosophy. 11:37:08 --- quit: ChanServ (bear.openprojects.net irc.openprojects.net) 11:37:11 --- join: ChanServ (ChanServ@services.) joined #forth 11:37:11 --- mode: bear.openprojects.net set +o ChanServ 11:37:12 --- join: cleverdra (julianf@0-1pool37-206.nas2.florence1.sc.us.da.qwest.net) joined #forth 11:37:13 --- mode: ChanServ set +o futhin 11:37:18 lol?! 11:37:22 so is forth a compiled language? 11:37:26 Yes and no. :) 11:37:29 It's both. 11:37:47 it's a compiled language with the features of an interpreter :) 11:38:03 Well, it's actually an interpretter that lets you extend the interpretter itself with a built-in compiler. 11:38:05 ah, alright. 11:38:28 most of the source code is compiled 11:38:35 so it's not a slow interpreter 11:38:39 and is it reasonably portable? 11:38:48 When you say something like : Hi! ." Hello World!" CR ;, the : word is executed when the interpretter finds it. : is itself is then responsible for compiling the remainder of the definition. 11:39:41 empirion: If you factor your code well, you'll find that you have an astonishingly little need for ANSI compatibility. 11:39:44 empirion - it's portable in the sense that it's easily ported, but not so much in the sense that it's automatically ported. 11:40:17 empirion: A Forth word serves two equivalent C functions: a C function (no pun intended), and a C #define macro. 11:40:28 forth is one of the most portable languages, because you can code a forth implementation for a new architecture in no time at all.. forth is very small and understandable, and coding a forth implementation in asm can be as small as 1k to 5k depending on what you do 11:40:44 where an 'automatic port' exists when your program, written for implementation A, is instantly compile-able on implementations B and C. 11:40:57 Thus, a Forth application can be made portable by factoring out platform-specific functionality, so that you can simply re-write the relavent words to map it to the new environment. 11:40:58 --- quit: cleverdra (bear.openprojects.net irc.openprojects.net) 11:41:19 --- join: cleverdra (julianf@0-1pool37-206.nas2.florence1.sc.us.da.qwest.net) joined #forth 11:41:31 got it :) 11:41:31 --- mode: futhin set -o futhin 11:41:32 It's exactly equivalent to conditional compilation. 11:41:48 wb cleverdra :) Getting sea-sick yet? :D 11:42:55 Hey, I didn't go anywhere. 11:43:01 =) 11:43:12 No, you only changed planes, that's all... :) 11:44:18 empirion - for maximal portability in Forth, you should design your programs in layers -- for instance, you might have a set of words for terminal I/O: CURSOR-UP CURSOR-DOWN etc. These would constitute a 'layer'. 11:45:12 empirion - words in higher layers only rely on these words -- on their stack effect and on their side-effect, so porting a program that used CURSOR-UP would only involve a possible redefinition of CURSOR-UP 11:45:28 empirion - this may sound complex, but it's how some of the most portable programs are written -- even in C. 11:46:00 forth could be automatically portable if forth implementations were written in a standard way 11:46:16 empirion - it's how a program like NetHack has X11, GTK, vt100 interfaces on Linux, and then ports for various other systems. 11:46:20 Yes, futhin brings up an important point. 11:46:42 ok, whats a good name for an operand or an address override prefix ? 11:46:49 The real reason C is "portable" today is because every single processor is designed to run C today. Every one. 11:47:08 and because a team of coders port a C compiler to new architectures 11:47:20 Through the process of evolution, various features were implemented and removed from processors that, basically, makes all CPUs behave in more or less the same way. 11:47:32 Thus, C compilers and their software are imminently portable "across platforms." 11:47:37 but i'll bet that a forth implementation could be ported by one guy faster than a team of C compiler coders :P 11:47:42 In the early days of C, however, this wasn't the case. 11:48:04 Writing code on a PC under DOS in C was a whole lot different than writing for AmigaOS, which in turn was worlds different from writing C software for Mac. 11:48:18 Today, it's trivial to write "cross-platform" software. 11:48:39 true enough. 11:48:53 Forth makes no basic assumptions about the underlying hardware of the virtual machine, and its performance therefore is greatly affected by that physical hardware. 11:49:11 As with the earliest of C compilers, modern Forth environments are typically tailored to the host environment they run in. 11:49:33 This has two beneficial effects: it breaks non-portable software quite visibly, and it makes that Forth environment quite fast. 11:49:46 isForth compiles at nearly a megabyte of source code per second, without the use of pre-tokenized Forth source, for example. 11:50:00 Contrast this to PygmyForth under DOS, which compiles a meager 32KB per second. 11:50:21 32k of source per second is nothing to sniff at 11:50:25 its still faster than c heh 11:50:39 pygmy is a damned nice forth, i just never liked block files 11:50:41 I440r: Yes, but you're three orders of magnitude faster. :) 11:50:46 hehehe 11:50:50 *snicker* 11:50:57 depending on what machine 11:50:59 But it's still relatively easy to port a Pygmy program to IsForth, because of the way Forth programs are designed. 11:51:18 True; mine is running on a 486DX-33 laptop under DOS without disk caching software running. 11:51:26 Yes, I was getting to that. 11:51:38 I440r - by the way, do you have any plans for ALLOCATE FREE RESIZE etc in IsForth? 11:51:44 I wrote my own block editor for Forth under GForth under Linux; GForth is an ANSI compliant Forth, and I used a lot of ANSI features. 11:52:02 When I ported it to Pygmy (which is NOT an ANSI compliant Forth), it wouldn't compile because those features were missing. 11:52:02 kc5 im gona implement blocks as an option in isforth 11:52:08 erm... eventually heh 11:52:18 My solution was simple: I wrote a screen or two of source code that implemented those features, and now it compiles just fine. :) 11:52:32 I440r - not too difficult. Just use your file interface, but have a special editor. 11:52:40 I440r: Blocks are pretty easy, actually. When I write my code, I'll send you my sources. 11:52:41 wats a good set of names for the $66 and $67 overrides ? 11:53:00 I440r: I use o16:, o32:, a16:, and a32:. :) 11:53:15 kc5 i was figuring make 1 meg a max size and memory map the block file and just calculate the offset to the given block 11:53:46 well, o16/32 are teh same thing as are a16/a32 heh 11:53:54 I440r: That's how it works. Although my Forth supports a full 64MB block space, each individual block file can only grow up to 1MB in size. 11:53:58 I440r - what's the point of the max size? 11:54:04 should i allow the USER to inject these himself or do it auto based on teh instruction ? 11:54:31 I440r: Since mine is only 16-bit, the user has to do it himself. When I implement the 32-bit extensions, I'll do both. 11:54:34 clever none realy - its arbitary hehe 11:54:51 aha rite 11:54:53 cleverdra: I limit file sizes so I can ensure enough space in the block map to map several files. 11:55:05 I don't want one file hogging all the blocks. 11:55:24 for manual injection 016 and 032 (identical overrides) makes sense because it shows context 11:55:25 nod. And I guess that 1MB is huge, anyway. 11:55:52 im not sure im gona support multiple block files open at the same time 11:55:54 maybe i will 11:56:05 still, it'd be easy enough to write a quick arbitrary-block-size in IsForth. 11:56:16 I440r - what about ALLOCATE and friends? 11:56:26 i got allocate and free written 11:56:31 need to do reallocate too 11:56:35 not a big deal heh 11:56:45 will be in 1.09b if i ever release it :) 11:56:47 'resize' is shorter -- but cool! 11:56:47 thought you finished dynamic memory management ? 11:56:59 futhin it is finished except for reallocate 11:57:23 I absolutely needed ALLOCATE RESIZE FREE for this last project I was working on. 11:57:24 ill call them o: and a: :) 11:57:32 I440r: You can write a very brute-force reallocate by allocating a new block, copying data over, and freeing the old block. Very simple, but also the slowest possible implementation. 11:57:40 I440r - release it now, please. Frequent revs is good. 11:57:44 I440r: Sounds good. 11:57:56 u cant realy resize an already allocated block, you can reallocate it to a bigger block 11:58:18 realloc will allocate the bigger block, move the data, deallocate teh smaller block and return the offset to teh new block 11:58:22 I440r: You can also shrink a block safely too, without copying it. realloc() does exactly this. 11:58:42 kc6 thers no realloc() in linux 11:58:48 nor is there any real alloc() 11:58:52 I440r: Yes, I know. It's in libc. :) 11:58:52 thers just memory-map 11:59:07 kc6 ya hehe libc uses mmap and munmap 11:59:08 and brk 11:59:14 Just as ALLOCATE, FREE, and RESIZE will be in your Forth. :) 11:59:41 i dont think you can sys_brk ANY allocated block 11:59:48 just the one linux gave to your process 12:00:01 and breaking out and then back in again is dangerous :) 12:00:55 I440r: No! You don't sys_brk on an allocated block. 12:01:10 You use sys_brk() to allocate the pool that ALLOCATE allocates from. 12:01:27 What you do inside those pool(s) is entirely up to your Forth environment. 12:01:59 You can safely sys_brk() backwards if you know for sure that nothing references the memory about to be reclaimed. 12:02:02 wow, this is the only channel i've been in on irc where people actually talk about stuff that's useful and interesting. 12:02:16 empirion: And it's a rare day here at that. :D 12:02:28 empirion - language-specific channels are like that, for some reason. 12:02:28 naw 12:02:39 we talk interesting stuff at least 2 days a week ;) 12:02:44 True... 12:03:03 sometimes you have to start the conversation 12:03:07 ask a lot of questions 12:03:10 * kc5tja nods 12:03:11 and eventually people start chatting 12:03:16 cleverdra: thanks, that's good to know :) 12:03:20 This whole coversation is empirion's fault, after all... ;D 12:03:34 yup, i admit it :( 12:03:38 heh 12:03:46 Unfortunately, I need to duck out of this conversation, as I need to get ready for work. 12:03:49 people start chatting about the conversation they were in, and then the conversation they are in, and then about the meta-conversation -- oh, it *never* ends! 12:04:01 --- nick: kc5tja -> kc-shower 12:04:11 kc5 - bye 12:16:16 BTW, this is fantasically good music: http://www.ffmusiconline.com/FF6/D3/14-Dancing_Mad.mid 12:16:25 What is it? 12:16:39 A track from Final Fantasy 6. 12:17:12 I've lately become somewhat enamoured with FF music, and this is exceptional. 12:17:40 heh 12:18:08 is it sorta technoy? 12:18:17 *technoie 12:18:48 More classicy, but it varies. Bachish, I think. 12:22:04 about 10 hours (kidding) into this, it breaks out in drums. 12:23:08 heh 12:46:36 --- quit: kc-shower ("[x]chat") 13:02:58 --- quit: I440r (Read error: 113 (No route to host)) 13:03:08 --- join: I440r (~mark4@67.241.61.199) joined #forth 13:10:58 --- join: jamc (~dne@as3-6-8.asp.s.bonet.se) joined #forth 13:16:31 --- quit: jamc ("[x]chat") 13:26:16 --- quit: cleverdra (Connection timed out) 13:32:05 --- quit: XeF4 ("->screen") 13:32:41 --- join: XeF4 (benzoglyox@12-245-116-85.client.attbi.com) joined #forth 13:38:43 --- join: kc5tja (~kc5tja@user-24-214-86-42.knology.net) joined #forth 13:38:52 * kc5tja is presently at work. 13:40:19 you are in a maze of little twisty passages, all alike. 13:40:25 you may go n,s,e. 13:40:29 you smell a boss. 13:41:17 I DISBELIEVE! 13:41:22 Do I make it? 13:41:58 Actually, it's funny you mention that. I just got done playing Atic Atac on the Jupiter ACE emulator I have. :) 13:53:14 hrm, i've seen juptier ace emulator links on forth pages but i haven't checked it out yet 13:53:49 Jupiters are so damn slow... Soooo slow... :) 13:54:10 Either this emulator isn't written write or it's way slower than a TRS-80. :) 13:54:54 But you can actually build a Jupiter using ordinary TTL components (including its video hardware!), except for the microprocessor. 13:55:19 I'd actually consider building one using a 65816 CPU instead of a Z-80; Z-80s are just so doggone slow... :/ 13:58:26 kc5 u think its important to make sure ppl dont try to do lock fs: rep stosb sorta crap ? or should i just let them shoot themselves in the foot 13:58:40 I think it's so slow largely because the Forth environment is certainly not a good fit for the Z-80. 13:58:43 checking for idiot code is gona be hard 13:58:56 Let them shoot themselves in the foot. 13:59:07 glad you said that hehe 13:59:08 I do basic error checking in my code, but I certainly don't make it a habit. 13:59:25 my code will abrot" on mov eax, bx 13:59:27 etc 13:59:31 abort even heh 13:59:40 * kc5tja nods 13:59:43 oh 13:59:53 mov byte [ebx], 10 13:59:53 oh? 13:59:57 is the w bit set ? 14:00:06 No; you're moving a byte. 14:00:31 oh! wbit isnt affected by base or index :) 14:00:31 erm 14:00:33 how do you do 14:00:42 mov byte [bx], 10 then ? 14:00:58 oh 14:01:08 shoot its got a 66 override heh 14:01:08 ok 14:01:54 67 actually. But I thought you weren't going to support 16-bit addressing modes? 14:02:05 no im not, i just wanted to figure it out 14:02:13 regarding teh w bit 14:02:23 * kc5tja nods/ 14:03:43 oh now i have to check for mov ax, # 5 byte 14:03:49 what is TTL again? 14:03:56 time to live :) 14:04:02 or transister transister logic 14:04:55