00:00:00 --- log: started forth/06.06.18 00:12:26 --- quit: JasonWoof ("off to bed") 00:21:50 --- join: Raystm2_ (n=Raystm2@adsl-69-149-45-51.dsl.rcsntx.swbell.net) joined #forth 00:37:49 --- quit: Raystm2 (Read error: 110 (Connection timed out)) 01:10:46 --- join: Cheery (i=Henri@a81-197-32-96.elisa-laajakaista.fi) joined #forth 01:52:14 --- join: Robert (i=bob@unaffiliated/robert) joined #forth 01:56:18 hi# 01:57:17 Hi 02:12:56 Hi Robert 02:12:59 Long time no Talk 02:16:32 True.. 03:46:33 --- join: PoppaVic (n=pete@0-1pool47-95.nas30.chicago4.il.us.da.qwest.net) joined #forth 03:55:25 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #forth 04:01:57 --- join: michaelw (n=michaelw@lambda.foldr.org) joined #forth 04:04:18 if I have a forth word BAR, and then define : FOO BAR ; then redefine BAR, which version of BAR is executed when executing FOO? 04:04:40 the older one 04:04:52 ah, good, thought so 04:07:33 Well, that may depend on the implementation. 04:08:07 If you have an indirect-threaded forth where a new definition of a word keeps the old header, but overwrites the code pointer, FOO would use the new version. 04:08:13 But that's rather uncommon I think. 04:08:50 Robert would an ID forth allow for shuffling whole words around memory? 04:09:30 how you mean? 04:09:58 indirect is lowlevel, but essentially akin to funcptrs 04:10:00 well, spose i want to reposition words to a different order 04:10:06 You don't have to shuffle around anything. The old code could remain on the heap. 04:10:09 PoppaVic !!! 04:10:19 reposition? 04:10:25 but i'm thinking the whole word structure 04:10:55 ohh... You mean like small, modular PIC code? 04:11:03 i dont know PIC 04:11:12 position-independent-code 04:11:13 The word structure can remain, too. 04:11:15 oh yea. 04:11:29 well, i'm seeing ***words 04:11:36 or ****words :) 04:11:48 well, you'd have to be over 100% sure that every word was relocatable - and relative - and this seems unlikely. 04:12:02 in asm coding, that's not a problem 04:12:27 lfa can be a reladdr to parent 04:12:30 or prev 04:12:31 it will be, when you start using secondaries (colon-defs), and want to move them around as well 04:13:51 I think tathi's Fovum-stuff might allow this, but I've been beating a vm for weeks and can't see how I can do the same as-is 04:14:17 Why would you want to shuffle dictionary _entries_? 04:14:22 PoppaVic you also wnat to be able to move words around? 04:14:35 it's a design flexibility thing i've been thinking of 04:14:50 I'm writing an entire vm thingie that isn't really forth, just forthlike. 04:14:51 i just wnt it to be possible, not that i have a particular use for it :) 04:14:56 nods 04:15:24 I would not suggest doing this, but instead think of a way to use "modules" 04:15:44 loading and unloading, yea, exactly 04:15:51 like linux kernel modules 04:16:01 that's one reason to design forth to be able to do that 04:16:15 itinerent loadable code 04:16:26 However, I've been racking my brain on this for quite a while, and (frankly) I think it's mostly a LPGL sort of idea, rather than "convenient". 04:17:11 well do as kernel does, allocate some ram, load and link the vocabulary to the forth-trunk 04:17:31 and then run it 04:17:38 or wait for user to run it 04:17:39 I'm already allocating, and thinking of dlopen & friends - and it's still not thrilling me. 04:17:48 ah yea 04:17:59 heh, we're actually in the same neighborhood!!! :) 04:18:38 I'm thinking it's actually not _too_ tough to load modules of "klasses", but I don't much care for the idea of relocatable "instances" 04:18:57 well, they'd have to be PIC/relative 04:19:06 instead of absolute addys 04:19:17 yeah, adding even more overhead 04:19:28 and both abs and rel fit in a cell 04:19:31 I'm just not sure it's sensible 04:19:42 nah, the overhead is mucho min 04:20:01 if you're target is x86, then speed obviates the concern 04:20:06 your 04:20:19 No... Not for moving - the moving would then be simple. The extra overhead would be jammed into inner-interpreters 04:20:33 ho wmany do you have? 04:20:53 every klass will have their own interpreter/executor. 04:21:00 wow 04:21:02 why is that? 04:21:04 subclassing will add layering 04:21:29 inner-interpret[0] would not have overhead that klassing would 04:21:31 I'm trying to get into a klass-based, OO-sorta' arrangement. 04:21:37 yea 04:21:54 I've seen too much code replicated over and over 04:23:21 who's code? 04:23:21 Plus, I am treating gcc as the "assembler", (which lets us abstract a bit immediately), and then trying to standardize "registers" and structs/types/klasses 04:23:39 k 04:23:49 the last I really delved was into gforth - it's one of the few that will actually build for macosx 04:23:59 nods 04:24:22 well i'm of the opine that gnu has stretched portability to the limits 04:24:35 i'm working my way back to asm 04:24:47 No - actually, they can back down a notch and thrash it very, very hard and improve it 04:25:05 they've abused it :> 04:25:29 the prob with all the gcc-related stuff is that every target mucks things in different ways, and may or may not report consistently, etc. 04:25:38 yep 04:25:58 Their front/back-end stuff was _supposed_ to be an improvement.. And they did the same goddamned thing AGAIN. 04:26:23 lol 04:26:33 now tell me how you REALLY feel! :) 04:26:43 i t oo am sick of auto* 04:26:47 i too * 04:27:03 rebelling actually 04:27:08 I never use it - I'm writing this vm to act as the core to a "metabuilder" shell. 04:27:15 cool 04:27:47 I want to sort of fold forthish, make & sh together. 04:27:56 vcool 04:28:15 i'm working on a better emacs 04:28:29 well, it would be if I had some clones. Most of the "interested" have sorta' vaporized. 04:28:41 they'll return 04:28:53 maybe 04:30:29 http://rafb.net/paste/results/tSNzd775.html 04:30:43 above is the basic/working abi header 04:32:10 all that union crap gives me a headache :) 04:32:15 and typedefs too 04:32:41 asm makes me ill - and I love structural 04:32:49 lol 04:32:54 touche 04:33:30 actually, the unions bred last week - I got really fuckin' PISSED at assumptions repeated over and over 04:33:40 lol bred 04:33:58 multiplied 04:34:07 yeah.. You have to remember: cpp doesn't allow sizeof-tests at compile-time. 04:34:22 k 04:35:28 And, (of course), you'd get VERY unhappy if a 'Cell' was in one register-type here, but in a _different_ register-type on another system. 04:35:38 well yea 04:38:03 --- quit: neceve (Read error: 113 (No route to host)) 04:38:46 http://rafb.net/paste/results/tGeezX49.html 04:39:14 above is the vm_object.h - I'd rafb the objects "input", but I've done a lot of breakout. 04:40:26 It's looking like I can pull almost all the defs/undefs out of the header and into the .c file. 04:40:59 interesting use of defines and includes 04:41:27 yeah.. I saw FICL doing something similar and it jogged some memories. 04:41:34 cool 04:43:21 http://rafb.net/paste/results/4EntCN98.html 04:44:13 --- quit: madgarden (Read error: 110 (Connection timed out)) 04:44:22 http://rafb.net/paste/results/hXSchc35.html 04:44:29 heh, the 3rd line of the intro comment reminds me of something Loeliger wrote in his book 04:44:53 Good book, if fairly dated 04:44:58 absolutely dated 04:45:00 :) 04:45:05 I look thru it regular 04:45:18 mines in storage but i too think of it 04:45:25 he advocated "diy" 04:45:25 ..just like the F-PC tech manuals and F83 stuff. 04:45:31 yep! 04:45:33 "diy"? 04:45:43 do it yourself 04:45:55 I guessed that just as you typed it ;-) 04:46:05 heh 04:46:20 Yes, but you can NOT do dick until you start to see how shit conceptually works _somewhere_ 04:46:52 Loreliger does that for one to learn from 04:47:00 and all the other implementors too 04:47:24 except for ANS 04:47:35 heh 04:47:51 hmm.. ANS bugs me like ANSI-C 04:47:58 committees dont build race-horses 04:48:06 they only build rhinos 04:48:11 worse - they ignore posix and such 04:48:17 who? 04:48:31 both i spose 04:49:04 ..and, in either forth or C, the ANS/ansi stuff ignores basics, but then allows you to do "options" or posix - and THEN shit starts getting consistently-inconsistent. 04:49:16 agreed 04:49:29 sysctl is an example. 04:49:52 k 04:49:56 First, it seems like it's going to be really USEFUL - then you try it several places and begin to scream 04:50:02 heh 04:51:17 Also, bear in mind: I've pretty well concluded that the heads should be detached from bodies, both abstracted. 04:51:24 why? 04:51:40 i've always resisted that idear 04:51:42 One reason would be to facilitate a turnkey strip 04:52:01 also, to use database-capabilities. 04:52:10 well, that's an idear about hiding, not sharing 04:52:20 not entirely - see above 04:52:27 where above? :>~ 04:52:32 database 04:52:48 what benefit does db get from headless? 04:53:27 Instead of just slapping byte after byte into the bloody heap, you store things more abstractly and allow the platform, os and users to override where and what matters. 04:53:54 ok, that's similar to my idear of ***words 04:54:13 similar/related 04:54:24 user=lookup(context, "Bob"); 04:54:45 what would that accomplish? 04:54:56 or, context(lookup, "Bob"); 04:55:00 abstracting out. 04:55:17 leaving more room for "overloading" and extension. 04:55:39 leaving "implementation details" out, but providing for reporting. 04:55:44 k 04:56:06 not the baseline klass "Entries" 04:56:14 not/note/ 04:56:26 k 04:56:53 By expecting every klass to support a "baseline", they can override, passdown, or extend 04:57:01 ok 04:57:52 This is a basic of "abstraction".. Which, (I pray), I can get organized a bit better so it works for secondaries AND primitives. 04:58:49 ok 04:59:08 I suspect I need to beat the klass "stacks" with a 4x4 club.. It's already as simplified as possible, but I am having a BITCH of a time trying to reconcile "colon-think" versus "C-interfaces" 04:59:10 Vic, i gotta go afk; bbl to continue if you're still around. 04:59:16 good to chat with you again. 04:59:23 sure.. I'll be here to about noon ;-) 04:59:29 okidoke 04:59:31 ttyl 04:59:35 yeppers 05:50:31 --- join: neceve (n=claudiu@86.104.205.201) joined #forth 05:51:31 --- quit: PoppaVic ("Pulls the pin...") 05:53:03 --- join: PoppaVic (n=pete@0-1pool75-144.nas24.chicago4.il.us.da.qwest.net) joined #forth 06:59:27 --- join: I440r (n=mark4@24-177-235-246.dhcp.gnvl.sc.charter.com) joined #forth 07:01:41 crc albert van der horst has a message for you on clf :) 07:01:52 Hey 07:03:48 hi robert 07:04:01 i did a pre-release release of isforth 07:04:06 with a debugger 07:05:30 Oh, yay. 07:05:54 No more ." Test!" then. ;) 07:07:22 lol dont count your chickens :P 07:07:28 brb gotta go run an errand 07:07:49 Yeah, I need a beerrun too ;-) 07:08:04 i dont drink beer :P 07:08:10 make mine a jd 07:08:28 lord, I couldn't function w/o beer between the family, net and myself. 07:09:06 Coffee hones the blade, the family and net bind and dull it. 07:09:43 It's like weather: COLD is easy - add a few layers; HOT is a whore - cops frown on running around w/o layers! 07:11:11 I440r: it's almost impossible to explain to folks "Whay do you bother with IRC if it bothers you?" *sigh* Should I stop watching and helping and read a goddamned romance-novel? 07:11:35 I dunno' - it's almost depressing, though 07:45:59 --- quit: Quiznos ("BitchX: the un-Cola") 07:48:06 what ? 07:49:08 np, I440r just rambling anymore - it's been that sort of day. 07:49:41 ok lol 07:50:06 Robert, u check out my debugger yet > 07:50:48 I440r: you ever come with a portable object/lib/loader-linker mess? 07:51:50 no not yet 07:51:56 why would it need to be portable ? 07:52:28 I440r: the "no" was sufficient. ;-) 07:53:57 I've had issues with "compilers" plus "linkers/loaders" for decades. Research will reveal why. 07:54:15 assemblers just add more issues ;-) 07:59:17 Nah, not yet 08:04:06 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth 08:05:51 hi 08:08:44 --- quit: PoppaVic ("Pulls the pin...") 08:08:46 Hi, snowrichard 09:14:07 --- quit: snowrichard ("Leaving") 09:17:33 --- join: JasonWoof (n=jason@pdpc/supporter/student/Herkamire) joined #forth 09:17:33 --- mode: ChanServ set +o JasonWoof 09:22:16 --- quit: neceve (Read error: 110 (Connection timed out)) 09:38:44 --- join: Snoopy42 (i=snoopy_1@dslb-084-058-104-118.pools.arcor-ip.net) joined #forth 09:47:54 good afternoon 09:48:00 Hi 10:11:22 --- join: snowrichard (n=richard@adsl-69-155-177-154.dsl.lgvwtx.swbell.net) joined #forth 10:11:52 hi 10:13:22 H 10:13:23 i 10:13:38 hey Robert 10:21:57 --- quit: snowrichard ("Leaving") 10:35:41 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #forth 11:20:25 --- mode: ChanServ set +o crc 11:29:41 --- nick: Raystm2_ -> Raystm2 11:56:19 here's your homework for the day: Formulate a series of experiments that will allow you to objectively test the hypothesis that the universe is conspiring to help you. 11:59:13 I've thought about that... but there's obviously no use in trying. The universe is slightly more powerful than you are. 12:00:06 And in reality (as opposed to Star Trek), I don't think an all-powerful conspiracy would make such silly mistakes which allow you to discover and defeat it. 12:01:10 i find your lack of faith in star trek disturbing 12:02:27 Tsss! 12:02:32 How's the smalltalk research? ;) 12:03:57 Robert: no use in trying what? 12:04:15 Trying to uncover the conspiracy. 12:04:16 nice :) I found a irc->telnet proxy 12:04:32 I was just playing discworld mud 12:05:04 and wishing it was running in an irssi window 12:05:18 Robert: not too well. my migraine has returned and i have RL things i'm tending to somewhat 12:05:20 probably can't do color though :( 12:05:23 JasonWoof: Heh, interesting. 12:05:39 saon: Tss... that RL seems like an annoying thing. 12:06:47 Robert: indeed, but a girl i'm somewhat courting exists in it 12:07:40 hmm, anybody knows a description how to bootstrap a forth systematically? 12:08:09 from scratch? 12:08:22 i did to some degree, but it was a lot of trial and error 12:08:26 crc: yes# 12:09:27 try to find the "Moving Forth" papers, they were a help to me when I implemented my forths 12:09:34 i know the zetetics stuff, but it is somewhat confusing and somewhat incomplete 12:09:49 ok, that's what I was thinking of 12:10:16 I am not really aware of any real resources describing the bootstrap process 12:10:37 yeah, i searched a bit but didn't find much either 12:13:07 michaelw: you want to write a forth that can compile its own binary? 12:14:23 JasonWoof: I am considering that, but i haven't found much on how to build a sane assembler in forth yet. 12:14:40 http://www.nondot.org/sabre/os/articles/TheBootProcess/ 12:15:18 Ray: that's a different sort of bootstrap ;) 12:15:24 yes :) 12:15:39 What kind of problems do you run into? I think it's rather straightforward. 12:15:49 JasonWoof: ive experianced the universe conspireing to my aid. Or atleast i've attributed several co-incEdinks to that end. 12:16:13 Just make a list of the state variables you use (stack pointers, perhaps TOS cache, etc.), initialise them, and then jump to the interpreter code. 12:16:20 * Raystm2 reads back to see where ray-link-find failed. 12:16:31 Robert: i am interested in a minimal set of words from which i can start defining the rest of the system already "from inside" 12:16:40 michaelw: Ah! 12:16:52 michaelw: AH! ( as well) 12:16:55 michaelw: I made an attempt at constructing such a minimal system. 12:17:31 P'' p-prime-prime 12:17:44 perhaps take a look at eforth, or foerthchen for ideas? 12:17:44 http://robos.org/code/edforth2.tgz if you're interested. x86 assembly language. 12:17:53 even more minimal then brainF*** 12:17:56 foertchen... heh. 12:18:00 neat, i will have a look 12:18:48 http://en.wikipedia.org/wiki/P_prime_prime 12:19:08 retroforth also has a pretty small assembly (x86) core, with most code written in forth 12:19:53 of course my P'' link is also not what you are looking for... 12:19:53 some of the remaining assembly could be moved to forth, but hasn't been for various reasons 12:20:18 BUT i've built some very interesting things using P'' and colorforth. 12:20:25 Interesting. 12:21:34 you need a way to span memory, and a way to adjust memory at location. after that, you can build practically anything. 12:48:40 well, I couldn't get that irctelnet thing to work 12:49:03 looks like it hasn't been touched in 4 years 12:49:53 seems like it would be cool to have an irc->telnet proxy with color support 12:50:03 but I don't feel like writing it 13:18:47 --- quit: saon ("leaving") 13:44:00 --- join: madgarden (n=madgarde@Quebec-HSE-ppp3617608.sympatico.ca) joined #forth 13:56:44 --- join: dauterive (i=51b53c90@xii.mine.nu) joined #forth 13:57:43 --- quit: madgarden ("?OUT OF DATA ERROR") 13:58:20 --- join: madgarden (n=madgarde@Quebec-HSE-ppp3617608.sympatico.ca) joined #forth 14:02:21 --- quit: Cheery ("Leaving") 15:17:00 --- join: T0paz (n=top@spc1-horn1-0-0-cust255.cosh.broadband.ntl.com) joined #forth 15:57:03 --- join: segher_ (n=segher@dslb-084-056-144-091.pools.arcor-ip.net) joined #forth 16:04:34 --- quit: segher (Read error: 110 (Connection timed out)) 16:16:27 --- quit: T0paz (Remote closed the connection) 16:33:29 --- quit: Zymurgy (Remote closed the connection) 16:49:23 --- join: Zymurgy (i=zymurgy@cat.delfax.net) joined #forth 16:52:40 --- join: ttuttle (n=tom@unaffiliated/ttuttle) joined #forth 16:52:42 Quartus: Hey. 17:37:06 Hey T. 17:37:35 * ttuttle likes having an i386 emulator because he can add debug messages to what would normally be a hardware device. 17:37:48 I see. 17:40:08 And what did PoppaVic have to contribute about that? :) 17:40:26 He's messed up. 17:40:27 Seriously. 17:40:39 He seems even more erratic lately. 17:41:00 It's a shame, because he occasionally has good ideas, but he's too insane to implement one well. 17:42:58 Really? Like what? 17:43:17 pv is not rational 17:43:19 ...awkward silence 17:43:30 Seriously, there was a good idea? 17:43:34 His forth build system could have been useful. 17:43:37 But he's insane. 17:43:41 Forth build system? 17:43:54 forth build system = forth 17:44:06 You mean his mythical autoconf improvement? 17:44:10 Quartus: Yes. 17:44:21 Quartus: But he's far too philosophical and annoying to ever get it done. 17:44:30 So, the one where he keeps calling autoconf 'autoshit' but can't code anything better? 17:44:36 Quartus: Yeah. That one. 17:44:50 Calling something names doesn't qualify as an idea. 17:44:54 Let alone a good idea. 17:46:16 * ttuttle takes back his earlier statement. PoppaVic is just loony. 17:51:30 What are you working on, ttuttle? 17:59:32 Quartus: My bootable Forth. I'm debugging a fault in the interrupts. 18:03:44 Figured out your table problem? 18:04:05 I'm not sure exactly what the problem is. 18:04:37 Bootstrapping is a tricky process. 18:04:47 Seems obvious once you've got it running, but getting there can be troublesome. 18:47:01 --- quit: Raystm2 (Read error: 104 (Connection reset by peer)) 18:47:44 --- join: Raystm2 (n=Raystm2@ppp-70-248-34-163.dsl.rcsntx.swbell.net) joined #forth 18:53:19 --- quit: uiuiuiu (Remote closed the connection) 18:53:24 --- join: uiuiuiu (i=ian@dslb-084-056-236-131.pools.arcor-ip.net) joined #forth 18:55:56 --- quit: neceve ("Leaving") 19:08:30 --- quit: I440r ("Leaving") 19:15:50 --- quit: ttuttle ("leaving") 22:21:44 --- quit: virl ("Verlassend") 23:59:59 --- log: ended forth/06.06.18