[Larceny-users] Larceny-x86-iasn on Win32 & FFI

[Felix Klock]

David-

On Jan 16, 2007, at 3:42 PM, david rush wrote:

> I'm *really* excited about the native x86 codegen in Larceny along
> with the fact that it appears to work under Windows. So my *next*
> question has to do with interfacing to native Windows libraries - is
> anyone doing it and is the FFI basically the same as it was under
> Unix?

I am trying on and off to improve the std-ffi.sch code and to  
construct a layer on top of the current FFI layer that is nicer to  
interact with.  So you hopefully will be able to use the old FFI  
layer if you want, but you will also have the option of using  
something a bit prettier.

With my progress so far since Christmas, I have made some demo  
programs utilizing gtk and opengl, so the FFI seems to be potentially  
useful, if not necessarily well documented . . . or even well  
defined; (e.g. Jesse and I are not sure what to do about first class  
continuations, so we're punting on them right now; here be dragons).

I do not test often on Windows machines, but I think the log message  
associated with changeset:3923 should make you more "chuffed":

http://larceny.ccs.neu.edu/larceny-trac/changeset/3923

Under the assumption that you would like to get your hands on these  
new developments and are willing to put up with alpha-quality code to  
get it, I will try harder to get our nightly builds posted to a web- 
accessible download area (or maybe delegate that task to someone  
else).  In the meantime, if you want to get your hands dirty, you  
could try grabbing your own checkout of the repository and making  
your own build of IA32 (Sassy) Native Larceny:

http://larceny.ccs.neu.edu/larceny-trac/wiki/SvnHttpCheckout

The Larceny developers would love to know what roadblocks people hit  
along the way doing that; all the better to improve our documentation  
with. . .

-Felix