Re: Bonobo Unique Application v3.1

[Michael Meeks <michael ximian com>]

Hi Gustavo,

On Fri, 2003-04-18 at 21:16, Gustavo J A M Carneiro wrote:
>   Changes:
> 	- Message discovery capabilities
> 	- some more gtk-doc documentation comments.

	This is beggining to look really nice; I still think we should do a GEP
on it, but I'd like to get most of the great stuff you've done in in
bits first.

	Can you commit the bonobo-activation part - that looks great; can you
split out the get_custom_poa stuff into bonobo-main.[ch], and add a:

enum {
	BONOBO_POA_ALL_AT_IDLE,
	BONOBO_POA_ONEWAY_AT_IDLE,
	BONOBO_POA_THREAD_PER_REQUEST
} BonoboThreadHint;

POA... *bonobo_poa_get_threaded (BonoboThreadHint *hint);

	Since we want other people to be able to start using this, pwrt.
gnome-vfs' MT daemon.

	The bonobo-arg.c fixes look great; can you commit those too ?

	I'd also like to get most of the nice signal stuff into CVS in a more
generic form as well, which looks like it could be generally useful. I
was envisaging that the 'getMessages' thing would return a
sequence<SignalDescription>:

struct SignalDescription {
	string name;
	Any    retval;
	Any    params;
	long   flags;
};

	And that we could build this out of: g_signal_list_ids, and a
g_signal_query on the remove GObject; this would then allow us to
introspet / invoke arbitrary signals on remote objects - which would be
a very nice feature - which we could then re-use inside libbonobo. Would
you be interested in splitting that functionality out into a
'BonoboSignalServer/BonoboSignalClient' object pair ? [ as a future
extension, we could (most likely) hook things up so we could
g_signal_emit on the client without a special method [ your argv /
collection thing looks great for now ], and it would 'just work' ;-].

	How does that sound ? either way, we should get as much as we can of
what you've done in now I think.

	Regards,

		Michael.

-- 
 mmeeks gnu org  <><, Pseudo Engineer, itinerant idiot