00:00:00 --- log: started retro/13.06.03
04:03:06 <tangentstorm> crc: some versions of forth had a 'recurse' keyword
04:05:01 <tangentstorm> i don't think it really makes sense to inline recursive functions though.
04:06:59 <tangentstorm> here's a thought: if you know you're going to recurse then you could have an opcode that pushes a copy of the instruction pointer to the return stack
04:07:13 <tangentstorm> call it 'mark'
04:08:35 <tangentstorm> then in addition to RETURN you could have RECURSE and ... i dunno... FINISH
04:08:57 <tangentstorm> RECURSE would jump to the address on the return stack without popping it.
04:09:29 <tangentstorm> FINISH would drop the TOS from the return stack and then do a normal RETURN
05:46:11 <crc> the recompiler would be less troublesome I think. (it could scan the quotes and detect recursions, and not inline the recursive functions)
07:06:30 <tangentstorm> very likely. you can be a whole lot less tricky when you don't have to emit code immediately on each word :)
09:32:16 --- join: ncv (~quassel@79.114.118.238) joined #retro
09:32:16 --- quit: ncv (Changing host)
09:32:16 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
11:11:29 --- join: Mat2 (5b419085@gateway/web/freenode/ip.91.65.144.133) joined #retro
11:11:35 <Mat2> hello
11:15:42 <tangentstorm> heya :)
11:25:54 <Mat2> hi tangentstorm
11:26:40 <Mat2> how goes your vacation ?
11:28:03 <Mat2> crc: your idea of inlined compilation at demand lead naturally to native-code generation
11:28:38 <tangentstorm> oh i was already back home when i told you about it :)
11:29:03 <Mat2> ah ok
11:29:33 <Mat2> I'm slowly beginning porting parable from phyton to C via nimrod
11:30:52 <Mat2> Saiwa features now some kind of dynamic vocabulary support based upon btrees
11:31:46 <crc> I moved the repo to bzr branch lp:~crc-x/+junk/parable-py and have a snapshot (updated nightly now) at http://forthworks.com/parable/parable-snapshot.tar.gz
11:34:50 <Mat2> thanks
11:36:25 <Mat2> I think to have a working version next week
11:40:11 <tangentstorm> Mat2: oh yeah... i tried nimrod.. wrote a little brainf*** interpreter/debugger in it. pretty fun stuff :)
11:47:10 <Mat2> well, the compiler has some bugs but beside that its like Oberon with intentation
11:48:01 <tangentstorm> yeah
11:48:43 <tangentstorm> i think i'll probably port ngaro over eventually.
11:48:43 <Mat2> (I'm only not impressed about '==' as equal operator and '=' instead of ':=' for value binding)
11:48:53 <tangentstorm> yeah, i noticed that too
11:49:19 <tangentstorm> it reminds me of a mix of python, oberon, and turbo pascal
11:49:43 <Mat2> I begin porting Saiwa to it, it makes sense
11:49:43 <tangentstorm> i like the macro support though.
11:50:06 <Mat2> that is very nice, specially for implementing interpreters
11:51:09 <Mat2> I'm a bit puzzled about there library concept
11:54:52 <Mat2> at moment
12:11:08 <Mat2> ciao
12:11:12 --- quit: Mat2 (Quit: Page closed)
12:40:26 --- quit: ncv (Remote host closed the connection)
23:59:59 --- log: ended retro/13.06.03