[Home] [Groups] - Message: [Prev in Group] [Next in Group]
12357: Re: [MUD-Dev] Storing tokens with flex & bison
[Full Header] [Plain Text]
From: Kevin Littlejohn <darius@connect.com.au>
Newsgroups: nu.kanga.list.mud-dev
Date: Mon, 03 Jan 2000 05:25:40 +1100
References: [1]
Organization: Kanga.Nu
>>> cg@ami-cg.GraySage.Edmonton.AB.CA wrote
> Byte-code execution really only makes sense, I believe, for a strongly
> typed language. If run-time checks and conversions are needed, they will
> likely cost far more time than the basic interpretation of the code, and
> so going to byte-code instead of staying with something more direct, could
> be a waste of time, in that byte-code won't speed execution up much.
...python seems to gain from using byte-code rather than straight
interpretation, and it's most definately _not_ strongly typed or static
in it's variable definitions.
I think there's a step or two there in just tokenising the code, and wrapping
it in the appropriate references to global namespaces, function/object
structures, and so forth, that you win on by pre-tokenising/bytecoding. It
also makes a nice compact way to shovel code around, if you're db'ing/undb'ing
it.
What about LPC? Does it do the same thing - I seem to recall the words
"semi-compiled", from my vague memories of working with it...?
KevinL
_______________________________________________
MUD-Dev maillist - MUD-Dev@kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev