[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: Mike Pall's standalone FFI: a dream?
- From: steve donovan <steve.j.donovan@...>
- Date: Wed, 11 May 2011 14:46:08 +0200
On Wed, May 11, 2011 at 2:19 PM, Lorenzo Donati
<lorenzodonatibz@interfree.it> wrote:
> Well, I don't think sleeping a process for a given amount of time, or
> detecting the title of the Windows of a running process or accessing
> Window's registry can be defined "demanding". There are things that Lua
> simply can't do without a C extension.
Very true. By 'direct' I meant LuaJIT-style FFI, as opposed to a
little extension for wrapping some common calls. I've used Alien for
that sometimes; there have been a couple of attempts to do WinAPI
bindings but no-one seems to agree on what 'everyone needs'. There's a
lot of functions in that API ! POSIX is a model of simplicity in
comparison. Although I'd say that a binding to the top one hundred
useful functions would go most of the way for console tools: window
and process management, enumerating resources, etc.
Small GUI applications? Well, one of the things that can be done when
LuaJIT gets callbacks is the ability to run native Windows GUI
applications. It will be small and insanely fast compared to the big
packages, at the cost of portability. But I suppose we no longer care
about lean and mean that much anymore - unless in the embedded game,
of course.
The other way to get small self-contained GUI applications is to use
the browser as the UI, i.e. use a dinky little local webserver.
steve d.