On Wed, 2013-03-27 at 19:57 +0100, poljar (Damir Jeli?) wrote:
> Hi Tanu. Thanks for the review.
>
> On Wed, Mar 27, 2013 at 03:22:32PM +0200, Tanu Kaskinen wrote:
> > On Tue, 2013-03-26 at 21:54 +0100, poljar (Damir Jeli?) wrote:
> > > +void pa_device_port_new_data_set_direction(pa_device_port_new_data *data, pa_direction_t direction) {
> > > + pa_assert(data);
> > > +
> > > + if (direction == PA_DIRECTION_OUTPUT)
> > > + data->is_output = true;
> > > + else if (direction == PA_DIRECTION_INPUT)
> > > + data->is_output = true;
> > > +}
> >
> > I think the direction should be set either always or never with
> > pa_device_port_new_data_set_direction(). I'd prefer the "always" option,
> > but since alsa modules can have dual-direction ports, they can't really
> > use pa_device_port_new_data_set_direction(), so as a second best
> > alternative, I suggest that you remove this function.
> >
> > Or there's another alternative: I could push a patch set[1] that removes
> > the possibility of dual-direction ports. That would possibly require
> > some rebasing work from you. What do you prefer? Remove
> > pa_device_port_new_data_set_direction() for now, or rebase on top of the
> > mentioned patch set after I've applied that?
> >
> > [1] http://thread.gmane.org/gmane.comp.audio.pulseaudio.general/13010/focus=13626
> >
>
> The second solution sounds better to me. If we're cleaning up this stuff we
> should do it properly. Isn't the feature freeze a problem here? (also I
> have an exam next week but I should find time to do this anyway)
Problem in what way? It's a problem in the sense that this patch won't
go in 4.0 unless you demand it, but the freeze doesn't stop us from
working on 5.0. The patch set that removes dual-direction ports is now
in the next branch in
git://gitorious.org/~tanuk/pulseaudio/tanuk-clone.git
--
Tanu
