totally rewritten glib bindings

[Colin Walters]

On Wed, 2005-02-16 at 23:43 -0500, Havoc Pennington wrote:
>On Wed, 2005-02-16 at 16:43 -0500, Colin Walters wrote:
>> So I gave up on the mega-function for now; the reason is that to do that
>> I have to pass everything as one mega-valist, but this prevents me from
>> using dbus_message_get_args_valist, which means I have to duplicate a
>> lot of that function's logic.  Probably the right way to do this is to
>> create an array from the valist, and have a version of
>> dbus_message_get_args_valist which operates on that array.  Not sure.
>
>We have to duplicate get_args_valist anyhow, because we need to return
>g_malloc memory rather than dbus_malloc memory. I just haven't gotten
>around to it. There may be ways to partially share the logic, haven't
>thought it through yet. (Sharing the logic isn't that big a deal right
>now, but if we added support for more type recursion etc. it might be of
>interest)

Hmm.  Ok.  I guess we would need to redo it for the glib bindings
anyways because not only do we want to return g_malloc memory, we want
to follow the other glib binding conventions and e.g. return g_strdup()
strings.  I guess it'd also be where we would have the logic for
converting dicts into GHashTable etc.

I'll have a look at this.

But do you think the current patch (or one of the previous iterations
you reviewed more closely) could go in?  It's getting a bit unwieldy,
and I'd like to start working on annotations and such.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.freedesktop.org/archives/dbus/attachments/20050217/d419feb7/attachment.pgp