[linux-dvb] Re: Can video_ioctl2() be unlocked?
hermann-pitton at arcor.de
Fri Oct 20 01:47:20 CEST 2006
Am Donnerstag, den 19.10.2006, 08:28 -0300 schrieb Mauro Carvalho
> Hi Hermann,
> Em Qui, 2006-10-19 às 02:53 +0200, hermann pitton escreveu:
> > > Regarding Mauro's new framework for the ioctls: the proof of the pudding
> > > is to actually convert an existing driver of at least medium complexity
> > > to the new code and see how it goes. I'm not convinced of the concept
> > > either, but then I haven't had the time to investigate it closely.
> > Those discussions should also be send over to the DVB ML,
> > since there will hardly be hardware in the future _not_ using digital
> > video and that is already the main feature on most products.
> Yes. For sure one needed API is an unique API for both analog and digital parts.
> As I've already pointed on V4L ML, one of the advantages of the newer
> approach is that API parsing will be done inside videodev.c. At the
> drivers, we will have callback routines, for each required function.
> This is basically the same generic approach as DVB API took.
> If tomorrow we decide to replace V4L API v2 to another API , this
> would be easy, since it would be just a matter of calling the same
> callbacks at the newer API.
> This will open a path to create an hybrid V4L/DVB API, capable of
> working with all sorts of multimedia capture boards.
>  In fact, I mean to keep the old V4L2 callbacks, to keep compat to
> older binary userspace apps, while creating a newer handler for the
> newer API.
> > Compared to it cams and encoders are minor issues ...
> Both issues are important. Hybrid devices is clearly a tendency for TV
> reception boards.
> We shouldn't forget, however, that analog to digital TV conversion won't
> affect Webcams, currently supported by V4L API. They should remain in
> the market, introducing newer features and improving resolutions.
> Also, video stream encoding from S-Video/RCA should also be there for a
> long time. Probably, this is the basic point why there are so many
> hybrid devices.
> Of course, analog broadcast TV tend to disappear in some years,
> especially on North America, Europe, Japan and Australia, but, o poorest
> countries, there will still be some demand for analog boards for some
> time. This probably means that analog TV development will stop, or
> reduce, but the manufacturers should still sell their pure analog board
> for a long time, maybe replacing tuners by some cheapo ones.
thanks for your explanations and I know you have lots of work, even
without any changes on the API.
Hopefully it becomes visible for everybody, that you don't do this for
your personal favours, but that there are requirements imposed by
already existing and upcoming hardware.
Hopefully everyone else of the experienced programmers will speak freely
up like Hans and Laurent did, that we come to a discussed and well
accepted conclusion, how to further safely sail this fine ship through
the rough waters.
More information about the linux-dvb