Hi!
I'm now, since Rob Browning has apparently little time for it, the
only person hacking more actively on G-Wrap. I've done a few hacks on
it, and currently am doing a rewrite, called G-Wrap TNG (The Next
Generation). You can find some more details on [0].
ATM, there are 2 major users of G-Wrap: GnuCash, and Guile-GObject[1]
(GNOME2 bindings for Guile). I today tried a GNOME2 build of GnuCash,
and found it uses (g-wrap gw-glib-spec), which will lead to trouble
since the latest released G-Wrap (1.3.4) has (minimal) bindings of
GLib 1.2, so these bindings are essentially useless for the G2 port. I
have agreed with Rob that eventually, G-Wrap will not ship GLib
bindings anymore. The development of the GLib2 Guile binding goes on
inside Guile-GObject.
My question now is: what plans do the GnuCash developers have
regarding the Guile bindings? I think it the there are two
alternatives:
1) Use Guile-GObject's GLib wrapping instead of the one in G-Wrap
1.3.4, leave everything else mostly as-is.
2) In addition to switching to the Guile-GObject GLib binding, use the
wrapping engine of Guile-GObject. This would offer the following
advantages:
* (Mostly-) Automatic header scanning instead of writing/maintaining
long -spec.scm files.
* For GObject-based objects, Guile-GObjects offers a very nice
object-oriented API, saving one from having to deal with all
those annoying prefixes. See [3] for an example.
* It will be easier for interested people (no, I'm not one of them)
to crank up Python bindings, because Guile-GObject and PyGTK
share the header scanner.
* When G-Wrap TNG is ready, people should be able to add bindings
to any language, since a major goal of TNG is to be
language-agnostic[4]
I volunteer for doing either one, but I think 2) is a worthwhile
mid-term goal (though not really necessary for the Gnome2 port).
You might also want to have a(nother) look at my other mail[5].
[0] http://yi.org/rotty/GWrap
[1] http://www.gnu.org/software/guile-gtk/
[2] http://article.gmane.org/gmane.comp.gnome.apps.gnucash.devel/10255
[3] Link missing, cause www.gnu.org down, but it is linked from [1]
[4] Yeah, I must admit I had the evil thought of recruiting
G-Wrap developers via this route ;-)
[5] http://article.gmane.org/gmane.comp.gnome.apps.gnucash.devel/10255
Cheers, Andy
--
Andreas Rottmann | [EMAIL PROTECTED] | [EMAIL PROTECTED] | [EMAIL
PROTECTED]
http://yi.org/rotty | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint | DFB4 4EB4 78A4 5EEE 6219 F228 F92F CFC5 01FD 5B62
A. Because it breaks the logical sequence of discussion
Q. Why is top posting bad?
_______________________________________________
gnucash-devel mailing list
[EMAIL PROTECTED]
http://www.gnucash.org/cgi-bin/mailman/listinfo/gnucash-devel
