> On Wed, Oct 14, 2009 at 11:56 AM, Wayne Meissner <wmeissner@gmail.com> wrote:

> >
> > Yep, should be changed, although in that specific case, it doesn't
> > matter, since we no longer use that part of libffi; ruby-ffi
> > implements its own closure allocator.

> 
> Yet still the condition in extconf is being used to define options for
> the build process, right?
> 
> As long that is true, the evaluation of the platform is still relevant

So far, extconf.rb was the only issue I saw.  The main Rakefile appears to be correct for platform evaluation.  Probably just a quick grep of host_os is need to verify.

Jon