00:00:00 --- log: started retro/10.03.08
02:36:17 --- join: virl (~virl__@chello062178085149.1.12.vie.surfer.at) joined #retro
06:15:32 <retro-commit> [crcx/retroforth] 7226c2: add start of forth200x compatibility library
06:20:29 <retro-commit> [crcx/retroforth] d123eb: add PARSE-NAME to forth200x
06:25:28 <retro-commit> [crcx/retroforth] fe3b0a: add [DEFINED] and [UNDEFINED] to forth200x
09:01:45 --- join: cfa (~cfa@unaffiliated/cfa) joined #retro
09:01:46 <cfa> hi
09:48:50 --- quit: docl (Ping timeout: 265 seconds)
10:08:24 <crcx> hi cfa
10:08:50 <cfa> hey crcx -- how's it going?
10:09:03 <crcx> pretty good so far
10:09:22 <cfa> i see you've been working on more compat stuff?
10:09:27 <crcx> yes
10:09:43 <crcx> I started adding some stuff from the forth200x effort
10:10:03 <cfa> neat
10:10:12 <crcx> (I have no intention of making Retro into yet another ANS system, but the compat stuff is sometimes nice to have for quick ports)
10:10:28 <cfa> for sure -- a good first step in porting
10:18:29 <crcx> regarding the keys code: do you want it added to aa.retro?
10:19:49 <cfa> sure thing -- did you play with it further?
10:20:45 <crcx> yes
10:20:57 <crcx> I haven't found any problems with it
10:21:01 <cfa> good good
10:21:38 <cfa> i'd suggest adding the contents of associative.keys to the main associative vocab and associative.keys.test to associative-tests in test/library/hash_and_aa/aa-test.retro
10:31:00 <crcx> will do
10:31:38 <cfa> i guess my speil about the prime at the top of the keys file should be integrated too
10:32:01 <cfa> spiel*.  it's important for the user to note that changing hash-prime midway through hash usage will screw things up.
10:55:44 <crcx> ok
11:02:04 <yiyus> crcx: i like the new implicit calls behavior, but it is in conflict with my extensions... any suggestion?
11:02:30 <yiyus> hello, btw, :)
11:03:12 <cfa> yiyus: what's the conflict?
11:03:54 <yiyus> well, i was using opcodes with a prefix, now they would be interpreted as function calls
11:04:21 <cfa> oh i see
11:04:26 <yiyus> i could carefully chose the opcodes, but it does not seem right
11:04:28 <cfa> what is the prefix used for?
11:05:12 <yiyus> floating point operations, string manipulation, concurrency...
11:05:28 <yiyus> the last byte is the op, the rest tells what extension to use
11:05:53 <cfa> why not protect the extension code with a lit?
11:05:58 <crcx> yiyus: given that values are signed, could negative values be used for the extended opcodes?
11:06:56 <yiyus> yes, negative values could be the most obvious solution
11:06:56 <cfa> crcx: try ': foo [ -10 , ] ; foo' in the current implementation
11:07:52 <yiyus> I think I will try that, once I get this working
11:08:27 <crcx> cfa: I'll fix the behavior with negative addresses
11:08:41 <yiyus> (I already made most of the changes I wanted, but it does not work at all, I will have to port the changes one by one...)
11:08:42 <crcx> cfa: all valid addresses should be positive
11:09:15 <cfa> crcx: sounds good
11:24:10 <crcx> done for console vm; I'll hit the others tonight
11:24:28 <retro-commit> [crcx/retroforth] f844ca: dont allow for negative addresses in jumps/calls (...
12:10:43 <cfa> yiyus: sorry, i thought you were using two cells {prefix,op}
12:43:48 --- join: crc_ (~charlesch@71.23.210.149) joined #retro
12:47:01 --- quit: crc (Ping timeout: 264 seconds)
12:54:30 <yiyus> cfa: I think Mat is using two cells (not sure though)
13:04:44 <crcx> Mat's vm keeps changing
13:05:21 <crcx> retro-fast is based on the last code he sent me, sans the extended instruction set
13:05:22 <cfa> {prefix,op} => {op,prefix} should protect the prefix from being interpreted as an xt
13:05:31 <cfa> (if that's what Mat's using)
13:05:50 <crcx> cfa: mat has moved to a radically different architecture
13:05:54 <cfa> k
13:06:08 <crcx> VLIW inspired now
13:12:05 <cfa> will our threading change be a problem for any of the vms in the repo?
13:21:14 <crcx> only the emacs implementation
13:21:27 <crcx> all others (except for gonga) are updated
13:22:05 <crcx> I can try to update the emacs implementation, but am not experienced in emacs lisp
13:23:56 <cfa> okay
13:24:03 <cfa> incidentally: can we make the javascript implementations auto-scroll?
13:24:39 <crcx> I should be able to do that
13:26:40 <crcx> I'm looking into the terminal emulation stuff from shellinabox
13:26:56 <crcx> I might be able to have it handle the console library
15:01:09 <crc_> home now
15:01:12 --- nick: crc_ -> crc
15:01:17 --- mode: ChanServ set +o crc
15:02:00 <cfa> welcome back
15:36:39 --- quit: cfa (Quit: cfa)
15:47:47 --- quit: retro-commit (Remote host closed the connection)
15:47:51 --- join: retro-commit (~Karere@li125-93.members.linode.com) joined #retro
15:57:49 <retro-commit> [crcx/retroforth] 953a5b: Merge branch 'master' of ../assembla
16:09:51 <retro-commit> [crcx/retroforth] f35805: keys stuff from Marc to data/aa.retro
16:14:49 <retro-commit> [crcx/retroforth] cc8cb3: sdl vm disallows negative addresses
18:31:23 --- join: cfa (~cfa@unaffiliated/cfa) joined #retro
18:31:26 <cfa> hi
18:31:31 <crc>                                                                                                                                                           welcome back
18:33:03 <cfa> thanks
19:04:49 <retro-commit> [crcx/retroforth] a78555: notes on hash table not needed in aa code
19:06:49 <retro-commit> [crcx/retroforth] 404176: removed from wrong file...
19:28:49 <retro-commit> [crcx/retroforth] caad40: autoscroll the terminal in javascript implementati...
20:06:49 <retro-commit> [crcx/retroforth] fbec8f: fix warning in retro-fast build
20:13:40 --- quit: cfa (Quit: cfa)
23:59:59 --- log: ended retro/10.03.08