Re: [MIT-Scheme-devel] GTk+ in MIT-Scheme?

[Andreas Rottmann]

address@hidden (Matt Birkholz) writes:

> Please review the Caveats (below) and the FFI Patchfile on this web
> page.
>
>       http://birkholz.chandler.az.us/~matt/Scheme/
>
> I propose to commit that patch and thus create an FFI (a basis for an
> FFI?) for MIT Scheme.
>
I had a quick look, and it seems this FFI is "static" in nature;
i.e. you can not do bindings that rely on run-time data. Am I right?

> A Gtk-Screen/Canvas for Edwin in 2009 then.  Surely! ;-)
>
Since you mention GTK+ bindings: are you aware of
gobject-introspection[0]? It is designed to make the creating of both
static and dynamic bindings much easier, as it should be possible to
create fully-functional bindings using only the introspected data (you'd
probably add a few niceties here and there, but the effort for doing
bindings of a library there is gobject-introspection data for should be
several orders of magnitude less).

I do a (runtime) binding library based on gobject-introspection called
sbank[1] (currently for R6RS implementations that provide a FFI that
supports runtime bindings -- currently working are Ikarus and MzScheme,
but Ypsilon is in the works). If you decide to do runtime bindings
(although, if I'm right, it doesn't look like it), this could be an
opportunity for sharing code -- a large part of the sbank code could be
reused in a R5RS implementation with a bit of refactoring (which could
be done in a way that allows the same code base for use in a R6RS and
R5RS environment).

[0] http://live.gnome.org/GObjectIntrospection
[1] http://live.gnome.org/sbank

Cheers, Rotty
-- 
Andreas Rottmann         | address@hidden      | address@hidden | address@hidden
http://yi.org/rotty/     | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint              | C38A 39C5 16D7 B69F 33A3  6993 22C8 27F7 35A9 92E7
v2sw7MYChw5pr5OFma7u7Lw2m5g/l7Di6e6t5BSb7en6g3/5HZa2Xs6MSr1/2p7 hackerkey.com

The best way to accelerate a Windows machine is at 9.81 m/s^2