Re: gstmm add element to pipeline test

[Murray Cumming <murrayc murrayc com>]

On Wed, 2007-12-05 at 17:57 -0500, José Alburquerque wrote:
> José Alburquerque wrote: 
> > Murray Cumming wrote: 
> > > I think we should do this in C++ much as it is done in C:
> > > - We should have an add_watch(priority) that just causes the signal to
> > > be emitted.
> > > and
> > > - We should have an add_watch(slot) that provides a sigc::slot that is
> > > called. We have lots of these set_func(slot)-like functions in gtkmm.
> > 
> > Hi.  I'm trying to implement these GstBus "watch" functions and I
> > have a quick question.  AFAICT, GstBus has the following "watch"
> > functions:
> > 
> > gst_bus_add_watch(GstBus*, GstBusFunc func, gpointer)
> > gs_bus_add_watch_full(GstBus*, gint priority, GstBusFunc func,
> > gpointer data, GDestroyNotify notify)
> > gst_bus_add_signal_watch(GstBus*)
> > gst_bus_add_signal_watch_full(GstBus*, gint priority)
> > gst_bus_remove_signal_watch(GstBus*)
> > 
> > As I understand, the first two will be wrapped using slots while the
> > last three deal with enabling/disabling the Gst::Bus "message"
> > signal.  My question is:  When these functions are wrapped, should
> > similar names be used for the methods (eg.
> > Gst::Bus::add_signal_watch() for gst_bus_add_signal_watch)?
> > 
> > -Jose
> > 
> > 
> I was sort of thinking of having a Gst::Bus::enable_message_signal()
> and a Gst::Bus::disable_message_signal() (for the message signal)

I do wonder if these can be called more than once, and what
remove_signal_watch() does if I've called it twice. If they can only be
called once then enable/disable would be clearer, but I think we need to
discover more.

>  and then having methods like Gst::Bus::add_watch() and
> Gst::Bus::add_watch(priority) for the slots. 

This seems correct.
-- 
murrayc murrayc com
www.murrayc.com
www.openismus.com