00:00:00 --- log: started retro/11.01.27
06:57:36 --- join: aisa (~aisa@173-10-243-253-Albuquerque.hfc.comcastbusiness.net) joined #retro
07:53:29 --- join: qbject (417dcfe2@gateway/web/freenode/ip.65.125.207.226) joined #retro
11:18:21 <crc> I implemented a new combinator: preserve
11:44:50 <cfa> interesting
11:45:45 <cfa> the documentation in Words.txt could be clarified slightly -- it reads like there's a typo until you realise that 'a' is referencing the stack value under the quote
11:46:40 <cfa> also: how necessary do you think preserve is?
12:21:10 <crc> it's not strictly necessary, but is something I've thought about adding for a while
12:22:28 <crc> would this be better?
12:22:28 <crc> preserve        (   aq-     )  Given a variable (a) and a quote (q), preserve
12:22:28 <crc>                                the contents of (a) while executing the quote,
12:22:28 <crc>                                and restore the original contents of (a) after
12:22:28 <crc>                                execution completes.
12:23:04 <cfa> much clearer, yes
12:23:17 <crc>                                (a) is removed from the stack before (q) is 
12:23:17 <crc>                                executed.
12:23:23 <crc> ok
12:31:50 <crc> preserve can be useful for macros, as it makes it cleaner to use the heap for temporary things. e.g., heap [ ...code allocating space at 'here' for temporary things... ] preserve
12:34:51 <cfa> ah, clever
14:25:09 <qbject> Anyone around?
14:25:29 <crc> me
14:25:54 <qbject> Sweet. I was curious about the versions of ngaro that can run on their own.
14:26:15 <qbject> As in, without a kernel under them. Am I understanding that correctly?
14:26:31 <crc> currently there are no standalone ngaro implementations
14:27:59 <qbject> I recall that, but didn't you say at one point that there were from Retro9?
14:28:30 <crc> retro9 didn't use ngaro; it had a kernel in x86 assembly
14:29:01 <qbject> Ahhh. That puts a tidy end to my question. Thanks. :)
14:29:31 * crc would love to have a native implementation of ngaro, but hasn't had time to work on it :(
14:38:48 <qbject> All things in time, I suppose.
14:40:01 <qbject> crc: is there a comprehensive description of an ideal ngaro somewhere?
14:40:54 --- quit: aisa (Quit: aisa)
14:45:49 <crc> the only semi-detailed notes are in the main repository, in doc/The_Retro_Language.rst
14:56:03 <qbject> Thanks. I'll check that out.
14:59:49 <crc> http://rx-core.org/rx.fossil/artifact?name=8c0ec8f6be5da57199861090d430390bddca1800 should be the latest
15:00:36 <crc> there might be additional stuff on http://retroforth.org/ngaro/description.html
15:04:11 <qbject> Great.
15:09:05 <qbject> Very rudimentary question, I know: what's the point of having both an instruction AND an opcode number?
15:23:57 <crc> each instruction corresponds to a particular opcode number
15:34:45 <qbject> crc: but in normal operation, retro is just passing opcodes to the vm, right? It's not like the instruction names are ever processed unless they're being displayed to a human?
15:37:36 <crc> correct
15:42:58 <qbject> Thanks. And now I shall vanish, perhaps to return tomorrow.
15:43:19 <crc> ok. have a good night.
15:43:26 <qbject> You too.
15:43:35 --- quit: qbject ()
17:51:10 --- join: aisa (~aisa@c-68-35-167-179.hsd1.nm.comcast.net) joined #retro
20:27:23 --- quit: aisa (Ping timeout: 240 seconds)
20:30:09 --- join: aisa (~aisa@c-68-35-167-179.hsd1.nm.comcast.net) joined #retro
21:23:36 --- quit: aisa (Quit: aisa)
23:48:58 --- quit: Sgeo (Read error: Connection reset by peer)
23:49:48 --- join: Sgeo (~Sgeo@ool-18bf618a.dyn.optonline.net) joined #retro
23:59:59 --- log: ended retro/11.01.27