[MPlayer-dev-eng] libvo changes

[Reimar Döffinger]

On Wed, Apr 02, 2008 at 08:22:42PM +0300, Uoti Urpala wrote:
> These patches remove variable arguments from VO control() functions. I
> haven't tested most of the changes as I don't compile those VOs, but
> hopefully there aren't too many bugs.
> 
> I created these in preparation of creating a data struct for VOs, but
> that isn't complete yet and I thought I'd post these first.

I can't see how this would be related to creating a data struct for
VOs...


> diff --git a/libmpcodecs/vf_vo.c b/libmpcodecs/vf_vo.c
> index f0daa59..63306f7 100644
> --- a/libmpcodecs/vf_vo.c
> +++ b/libmpcodecs/vf_vo.c
> @@ -103,13 +103,13 @@ static int control(struct vf_instance_s* vf, int request, void* data)
>      {
>  	vf_equalizer_t *eq=data;
>  	if(!vo_config_count) return CONTROL_FALSE; // vo not configured?
> -	return((video_out->control(VOCTRL_SET_EQUALIZER, eq->item, eq->value) == VO_TRUE) ? CONTROL_TRUE : CONTROL_FALSE);
> +	return((video_out->control(VOCTRL_SET_EQUALIZER, VOCTRL_SET_EQUALIZER_ARGS(eq->item, eq->value)) == VO_TRUE) ? CONTROL_TRUE : CONTROL_FALSE);
>      }

Personally, I find this line rather too crowded as it is, a
struct voctrl_set_equalizer_args param = {eq->item, &eq->value};
in a extra line would look much better IMO.

Not related to this patch, but also some VOs use strcmp and some use
strcasecmp. Is there any reason why we _have_ to use strcasecmp? strcmp
seems a much saner choice to me otherwise...

Greetings,
Reimar Döffinger