Re: [xine-devel] Viz refresh rate

[Michael R. <mr...@us...>]

Hi Miguel,

> > Miguel what do you think ?
>
> I'm a little lost with all these flying patches... ;)
>
> but from what have i seen:
>
> - having a new "something" to carry the common code for smoothing the
>   frames vpts: good idea.
> - one metronom for all x viz metronom. i prefer "one metronom to rule
>   them all" ;-) but we should weight the pros and cons.
> - new discontinuity type DISC_VPTS: sounds hacky.

The "one rin^H^H^Hmetronom to rule them all" solution would require a 
way to tell the metronom "listen up fellow, I will be feeding you with 
PTS values from now and I want the first PTS to result in the VPTS I 
give you". That's because we do not have the original discontinuities, 
so our reference are the VPTS values from the audio stream.
Does that explanation make it sound less hacky?

> - put_buffer() returning vpts: good idea (even if we have no use for
>   it now).
> - viz plugins tracking discontinuities: ugly ugly ugly! ;)

We are lucky here: That's going to be removed by either of the two 
solutions.

> > > Btw, since we are only discussing the syncing inside the post
> > > plugins now, can I commit the engine patch? (which makes metronom
> > > xine_stream_t-less, since you do not have that available, when
> > > initializing the post plugins)
> >
> > i think you can commit this.
>
> me too.

I will have to resolve some conflicts, but expect this in cvs this 
evening.

Michael

-- 
panic("Unable to find empty mailbox for aha1542.\n");
	2.2.16 /usr/src/linux/drivers/scsi/aha1542.c