00:00:00 --- log: started retro/06.07.31
00:38:36 --- join: Cheery (n=Cheery@a81-197-20-242.elisa-laajakaista.fi) joined #retro
01:35:43 <Cheery> Hmm.
01:36:18 <Cheery> to lessen complications, I think I'd need something which reads ahead and evals it.
01:53:48 --- join: virl (n=virl@chello062178085149.1.12.vie.surfer.at) joined #retro
02:21:43 <Cheery> btw.
02:21:53 <Cheery> now you guys see the first results from my work.
02:22:20 <Cheery> http://retroforth.net/paste/?id=97
02:22:41 <Cheery> This proofs that at least some of my assemble, -features works.
02:23:43 <Cheery> because ebx and eax are compile-time stuff, I put them after.
02:23:59 <Cheery> therefore it got easier to give them to reg-move
02:24:31 <Cheery> I still have no the opcode solver, it is quite complicated stuff.
02:26:24 <Cheery> hmm.
02:26:31 <Cheery> How things like case: works?
02:27:39 <Cheery> [ 0 ] [ebx] [edx*4] store eax
02:28:01 <Cheery> :/
02:29:56 <Cheery> of course I could do something like:
02:30:17 <Cheery> [ 0 [ebx] [edx*4] ]store eax
02:30:20 <Cheery> or
02:30:33 <Cheery> [ 0 [ebx] [edx*4] eax ]store
02:30:45 <Cheery> what do you think?
02:30:51 <Cheery> which looks better?
02:31:01 <Cheery> reg-move ebx eax
02:31:03 <Cheery> or
02:31:15 <Cheery> [ ebx eax ]reg-move
02:31:28 <Cheery> or: [ ebx eax ]mov
02:32:54 <Cheery> hey! I'm asking opinions here. :/
02:34:16 <Cheery> virl? crc?
02:34:22 <Cheery> docl?
02:34:57 <Cheery> then there could be also: [ 0 [ebx] [edx*3] store ]
02:35:07 <Cheery> + eax
02:44:26 <virl> hi Cheery how are you?
02:45:50 <virl> well I think reg-move ebx eax looks quite good.
03:09:47 <Cheery> Hmm.
03:09:50 <Cheery> great.
03:09:59 <Cheery> Thought I wonder if I shouldn't make them macros.
03:10:26 <virl> the versions with [] are: a) too confusing b) too long to type ;-)
03:10:44 <Cheery> but you could chain them.
03:10:45 <Cheery> like:
03:11:24 <Cheery> [ eax ecx mov 0 [ebx] [edx*3] eax store ]
03:12:11 <Cheery> And they wouldn't require to be state smart neither macros.
03:12:16 <virl> ok, I try to decode...
03:12:38 <virl> eax gets moved into ecx and then something with 0 happens...
03:13:17 <Cheery> if you'd move here -pointer to tib for a moment, you could also do temporate compiles and execute them immediately.
03:14:39 <virl> what did the [ ] in this context actually mean? not the [] around the register names.
03:14:51 <Cheery> asm: eax ecx mov ;asm could be anywhere to express it'll be executed even in interpreter.
03:15:00 <Cheery> virl: don't you remember?
03:15:38 <Cheery> they prevents compiling inside word.
03:15:46 <virl> switching between compiler and interpreter?
03:16:05 <Cheery> : word [ ."executed now" ] ."executed in call" ;
03:16:24 <virl> iirc, [ exits from compiler and  ] enters again.
03:16:27 <Cheery> I just don't smart to oversmart myself.
03:16:40 <Cheery> *want.
03:16:43 <virl> ok, well then I know them ;-)
03:17:01 <Cheery> ok, I must go, see ya.
03:17:03 <virl> because it seemed that you use it outside from definitions
03:17:08 <virl> ok..
03:18:05 <Cheery> I make them forth words that you can use them that way, it seems most straightforfard and simple.
03:18:34 <Cheery> and one can build the macro things from them later.
04:51:40 <Cheery> crc: http://retroforth.net/paste/?id=98
04:51:55 <Cheery> virl:  I think I'll use that version.
04:52:26 <Cheery> It's the most flexible of those versions.
04:57:22 <virl> ok and it looks good. you know I can understand it without exactly knowing what the code does and that's good.
04:58:27 <virl> the only question is, is it moving from ebx to eax or from eax to ebx?
05:03:17 <Cheery> from ebx to eax
05:03:29 <Cheery> virl: do you think that is sufficient?
05:03:36 <Cheery> from is before to in registers.
05:04:00 <Cheery> in memory addresses, it varies from command because the memory address is always before register.
05:04:02 <Cheery> like:
05:04:17 <Cheery> $10000 eax store
05:04:24 <Cheery> $10000 eax recall
05:04:27 <Cheery> ;)
05:04:45 <virl> afk...
05:46:11 --- quit: nighty- ("Terminated with extreme prejudice!")
06:06:17 --- join: Ray_work (n=Raystm2@199.227.227.26) joined #retro
06:39:06 <docl> hiya Cheery, still up?
06:39:21 * docl just got up
07:09:34 <Cheery> ye
07:20:04 --- join: nighty_ (n=nighty@66-163-28-100.ip.tor.radiant.net) joined #retro
07:30:23 <docl> howdy
07:30:45 <docl> impressive work
07:34:07 <docl> I agree with virl, the simpler one without brackets looks best.
07:54:59 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #retro
07:59:04 <Cheery> But I wonder if it is the right choice.
07:59:09 <Cheery> Think the macros.
08:08:25 <docl> eax and ebx are macros?
08:09:38 <Cheery> nope, and I don't want to use macros.
08:09:43 <Cheery> at this point.
08:09:51 <docl> oh, ok
08:10:05 <Cheery> reg are constants
08:10:17 <docl> I take it they need evaluated at compile time though
08:10:20 <Cheery> [reg] and [reg*n]s are functions.
08:11:12 <Cheery> yes, they goes to the word which contains the compiling mechanics for each word.
08:11:31 <Cheery> and uses the backbone of my assembler to do the final work.
08:12:10 <docl> perhaps we need a compile-time evaluation prefix
08:12:30 <Cheery> we already have one.
08:12:33 <Cheery> [
08:12:35 <Cheery> and ]
08:12:50 <docl> '_ prefix: [ name eval ] ;
08:13:07 <Cheery> the prob is that there are lots of those things to be needed evaluated instead of compiling.
08:13:27 <Cheery> oh well.
08:13:36 <Cheery> with simple commands the prefix could be useful.
08:13:38 <Cheery> thought.
08:13:55 <docl> hmm, wait isn't x: about the same thing?
08:15:12 <docl> maybe not
08:16:50 <docl> here we go: '_ prefix: x: [ eval x: ] ;
08:17:34 <docl> (underscore is not necessarily the best prefix)
08:26:50 --- join: timlarson_ (n=timlarso@65.116.199.19) joined #retro
12:31:58 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-110-053.pools.arcor-ip.net) joined #retro
12:32:24 --- quit: Snoopy42 (Nick collision from services.)
12:32:35 --- nick: snoopy_1711 -> Snoopy42
13:29:30 --- quit: timlarson_ ("Leaving")
13:29:45 --- quit: Cheery ("Download Gaim: http://gaim.sourceforge.net/")
14:26:49 --- nick: Raystm2 -> nanstm
14:29:11 <crc> good evening
14:39:53 * crc is going to continue with his source-wide cleanups
14:41:29 <virl> hiho...
14:46:21 <crc> what's happening?
14:47:42 <virl> well, I play with the idea to produce an assembler written in retro for some nice microcontrollers, or at least for one of them ;-)
14:53:45 <crc> cool
14:53:56 * crc coded an assembler for his virtual machine
14:54:26 <virl> your virtual machine?
14:54:53 <crc> I coded a stack based virtual machine with memory mapped hardware earlier this year
14:55:04 <crc> portable, written in C :)
14:55:37 <virl> bah.. 
14:55:47 <virl> what is it able todo?
14:55:58 * tathi is working on a PPC assembler/disassembler in forth
14:56:27 <crc> virl: run simple programs
14:57:22 <crc> virl: long term goal is a portable target to develop a varient of RetroForth
14:58:24 <tathi> which you plan to run on phones or palmtops IIRC
14:58:37 <crc> yes
15:00:08 <crc> my last experiment was some code to walk through a dictionary that is almost identical to retro's in structure
15:00:45 <crc> next on the list (when I get around to it) is to implement "lookup"
15:01:51 <tathi> in your copious spare time :)
15:03:22 <crc> yes
15:04:20 <Ray_work> Ray_work: departing. 
15:04:31 <Ray_work> see ya from la casa. 
15:04:34 <virl> bah, now I'm asking myself if there is a sense with xell. puh..
15:04:43 --- quit: Ray_work ("Client exited")
15:05:35 <crc> virl: my vm is a very simple approach. inefficient, but enough to give me a new world to play with
15:06:25 <crc> not to mention, it provides nothing but a barebones text console, and eventually a block i/o system for simulated hardware
15:07:01 <virl> ah, ok :-)
15:08:01 <crc> if you want to see it, I can paste a link to the repo :)
15:28:54 <virl> yeah, I'm always interested into something new which kicks the old forth world in the ass ;-)
15:29:34 <crc> http://retroforth.net/maunga
15:29:54 <crc> http://retroforth.net/maunga/doc/Toka.txt describes the instruction set 
15:31:57 <virl> hmm..
15:34:19 <virl> well, we share a lot of the instruction set ;-)
15:34:30 <virl> well, it's forth
15:34:46 <crc> yeah
15:35:02 <crc> I am trying out a machine forth/misc approach to the virtual hardware
15:35:43 <virl> well me too ;-)
15:36:09 <crc> http://retroforth.net/maunga/images/poharu is the assembler
15:36:14 <crc> (written in retro)
15:36:15 <virl> only that my one is more orientated  on streams
15:37:00 * crc needs to redo the assembler using the newer features in retro 9.2.x
15:40:48 <virl> hmm.. you use a switch for the opcodes. 
15:41:22 <crc> yes
15:43:19 <virl> well, and probably your source is much shorter than mine..
15:43:53 <crc> 617 lines of C, including comments and whitespace, and the debug execution tracer
15:45:59 <virl> well, and mine a little bit about 1000
15:46:38 <crc> I expect mine to grow to about 1000-1200 lines once I finish the hardware devices and finalize the instruction set
15:47:38 <virl> well, I'm about 1193 lines, but if/while/for is missing. well it will grow more.
15:48:16 --- quit: nighty_ ("Disappears in a puff of smoke")
15:49:11 <virl> I think that's because of my stacks and my api interface.
15:52:15 <crc> mine is pretty limited
15:52:47 * crc does expect to have a mostly functional cross-compiler implemented by mid-September
15:53:38 <nanstm> Hi.
15:53:42 <nanstm> oops 
15:53:47 --- nick: nanstm -> Raystm2
15:55:07 <crc> hi Ray
15:56:22 <Raystm2> Hi crc: trying to settle in here. 
16:02:28 <Raystm2> How was your day, so far?
16:03:30 <crc> pretty good
16:14:08 <Raystm2> oh yeah. I remember toka
16:14:12 <Raystm2> this is cool.
17:25:24 <crc> 9 new patches in the repo; another day or two until 9.2.5 will be ready (almost entirely cleanups; one new feature; "remove-prefix" from the lookup table)
17:32:43 * crc wanders off to bed; goodnight
17:32:55 * Raystm2 needs to look into the prefix code to see what all is available now.
17:33:00 <Raystm2> Goodnight.
18:27:53 --- quit: tathi ("leaving")
18:31:13 --- join: rabbitwhite (n=trip_n_s@pool-71-248-84-154.bltmmd.east.verizon.net) joined #retro
18:32:14 --- quit: rabbitwhite (Client Quit)
20:00:27 --- join: nighty_ (n=nighty@CPE00119576a9c5-CM0012c90d36fc.cpe.net.cable.rogers.com) joined #retro
20:52:15 --- nick: Raystm2 -> Raymole
20:53:06 --- nick: Raymole -> Raymole_St
20:53:30 --- nick: Raymole_St -> Raystm2
23:59:59 --- log: ended retro/06.07.31