Re: Ripping out GLib bindings, depending on GLib

[Andy Wingo]

On Thu, 04 Dec 2003, Andreas Rottmann wrote:

> 1) Is it reasonable for g-wrap to depend on GLib?

I don't really think so, unless there's a dire need. In this case, I
don't think it's necessary, as outlined in a mail to guile-gtk-general.

> 2) It's not possible combine usage of GLib 1.2 and 2.X. Since g-wrap
>    comes with a GLib 1.2 binding (which is superceded by the GLib 2.X
>    in guile-gobject). This means I either can't use GLib in the core
>    g-wrap, if I want guile-gobject stay operational, or rip GLib
>    binding stuff (along with the GTK+ 1.2 binding) out of g-wrap. I
>    definitely favor the latter.

Kill the (g-wrap internal) GLib binding. Don't even worry about it. GLib
1.2 has been outdated for years.

> While writing up 2), I wondered if this is a reasonable point for a
> g-wrap version fork: Release g-wrap 1.3.4+fixes as g-wrap 1.4.0 and
> start a 1.5 branch with my modifications/extensions. I guess it is not
> feasible to drop the GLib 1.2 binding in the 1.3 series, since gnucash
> (still) depends on it.

I agree. But this depends on Rob. If we don't get a response, we're
going to need to fork, I think. Rob: If you're swamped, consider (again)
adding a co-maintainer. g-wrap is worth more as a live project than as a
dead baby.

Regards,

wingo.