00:00:00 --- log: started retro/10.09.16 04:06:01 --- quit: erider (Quit: ChatZilla 0.9.86 [Firefox 3.6.9/20100824153629]) 04:09:48 --- join: Espy (~Espy@166.137.141.151) joined #retro 04:13:13 --- part: Espy left #retro 06:49:27 --- join: docl (~luke@97-120-241-117.ptld.qwest.net) joined #retro 07:21:41 --- quit: crc (Ping timeout: 252 seconds) 07:23:30 --- join: crc (~charlesch@184.77.185.20) joined #retro 11:06:40 --- join: roarde (~sixforty@pdpc/supporter/active/sixforty) joined #retro 11:19:28 --- quit: roarde (Quit: Leaving.) 11:38:56 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 11:39:10 hello 12:43:57 hi 12:47:51 --- quit: foucist (Ping timeout: 252 seconds) 13:09:01 --- join: foucist (~foucist@ps14150.dreamhost.com) joined #retro 13:09:01 --- mode: ChanServ set +v foucist 13:15:11 --- quit: foucist (Ping timeout: 276 seconds) 13:16:32 --- join: Mat2_ (4d177cf6@gateway/web/freenode/ip.77.23.124.246) joined #retro 13:16:48 --- nick: Mat2_ -> Mat2 13:16:59 hello 13:20:35 crc: I had take a look at retro 10.7 13:22:32 and want to ask if the compilation of DTC code is now prefered against STC ? 13:26:44 Mat2: at this point, yes 13:31:07 would it be, in this case, not better to create a standart forth vm which executing DTC code and implement the Ngaro instruction set as optional byte code representation at Forth level, like Helforth did ? 13:31:18 the threading has been (mostly) dtc since 10.5 13:31:42 just wondering 13:32:28 I have no problems with my current implementation and no real desire to rewrite things again at this point 13:33:12 yes, that would lead to a complete rewrite 13:34:55 anyway I like the new approach 13:35:23 any suggestions on a good opcode to replace VM_CALL? 13:35:37 (assuming that I drop it) 13:36:41 because of DTC, the vm doesn't need such an opcode, I think 13:36:54 agreed 13:37:06 but you can replace it with an combined cmp and branch instruction 13:37:34 I already have combined compare and branch instructions 13:38:08 yes, I mean an opcode for counted loops like the rep prefix on IA32 13:40:11 probably an pick instruction can be useful too 13:41:20 I will *not* add a pick instruction 13:41:52 I like the rep idea... it'd make some things much cleaner 13:42:22 I think so 13:48:56 hmm, because DTC code is the default the interpreter can be reduced to two lines of C code (for the replicated-switch threading version) I think 13:49:44 (not including the replicated switch constructs) 13:51:55 ok, thanks, I will experiment with my idea now 13:51:56 ciao 13:52:03 --- quit: Mat2 (Quit: Page closed) 15:53:08 evening 17:28:13 --- join: foucist (~foucist@ps14150.dreamhost.com) joined #retro 17:28:13 --- mode: ChanServ set +v foucist 18:09:00 --- quit: erider (Remote host closed the connection) 18:13:55 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 18:14:25 evening 18:41:48 good evening 18:43:05 --- join: roarde (~sixforty@pdpc/supporter/active/sixforty) joined #retro 19:45:05 I'm don't see a graphical vm in 10.7. 21:35:28 --- quit: crc (Ping timeout: 252 seconds) 21:36:07 --- join: crc (~charlesch@184.77.185.20) joined #retro 22:16:20 --- quit: roarde (Quit: Leaving.) 23:59:59 --- log: ended retro/10.09.16