[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Re: bit.lshift and performance - luabitop v.s. lua-5.2.0-work4
- From: steve donovan <steve.j.donovan@...>
- Date: Wed, 13 Oct 2010 15:45:45 +0200
On Wed, Oct 13, 2010 at 2:40 PM, Enrico Colombini <email@example.com> wrote:
> I remember having to write horrible Lua bit ops workarounds just to be able
> to make small changes to data files.
Yes, it's often a question of notational convenience and knowing when
to get closer to the machine.
Part of the problem is that we need a higher-level way to express
these operations. I'm thinking of a recent thread about describing
packet structures in Lua and either 'interpreting' them directly or
'compiling' them to optimized C. The first makes testing easy, the
second makes coding with a micro with 4K RAM possible.
Naturally, this is not a new idea, for instance vstruct:
Basically, trying to think of a way to transform a problem from the
Von Neumann space to a set of declarations ;)