[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Re: [ANN] LuaJIT-2.0.0-beta1
- From: Mark Hamburg <mark@...>
- Date: Mon, 2 Nov 2009 22:30:35 -0800
On Nov 2, 2009, at 4:06 PM, Mike Pall wrote:
Mark Hamburg wrote:
To what extent did you change the instruction set v to what extent
you change the encoding of the instruction set? (I'm thinking about
whether a bytecode translator is even reasonable to consider.)
Summary: Nope, use source code. Bytecode isn't shorter, anyway.
Oh, and LJ2 doesn't even *have* a bytecode loader/dumper at the
moment. It simply wasn't a priority. Open source projects using
LJ2 have nothing to worry about, since they don't need to hide
their Lua sources.
The idea was to make a commercial add-on for bytecode handling.
But I'm not sure it would pay off (required work + support
overhead vs. money in return). There are cheaper ways to hide your
Lua sources: simply compress and encrypt them. And bytecode is
easy to reverse-engineer, too.
It would be nice to save the parse step when loading though perhaps
some of Michael Franz's work prior to tracing JITs is relevant to that