[vdr] Re: ANNOUNCE: Softdevice plugin. Use /dev/fb for output

[alexw@undercover.mine.nu]

On Thursday 04 December 2003 23:50, Roland Praml wrote:
> > I must admit that I'm not familiar with Vidix, so I can't say much how
> > the other libraries relate to Vidix, but in general it depends on what
> > you want to do and how portable you want to be.
>
> the Vidix library is very hardware depend and there also exists only
> a few hardware-drivers. But it was easy to implement (more or less just
> copy and paste from an example codelet).
>
> > The most portable way in the PC/Mac world would be based on SDL or
> > OpenGL - if you design your code properly then your code will run on all
> > cards that are supported by these libraries: this is virtually all
> > available hardware in conjuction with all commonly used operating
> > systems.
> >
> > When you want to let your code run once on 'real SetTopBoxes' then
> > DirectFB is probably your first choice.
>
> I think I change to DirectFB...
> It shouldn't be a problem to get linux running on a cheap client like a
> Xbox or an Mini-ATX MB based system ;-)
>
> > I personally would choose OpenGL for a bigger project - SDL and DirectFB
> > for smaller ones. If you are interested I can send you off-list some
> > codelets for DirectFB and SDL showing how you can implement this.
>
> Yes, I'm very interested. (Especially how to implement the YUV and ARGB
> layer.)
>
> TIA
>
> Roland


Hi Roland,


That's really a nice job. I hope you still improve your plugin ;-)

I did some tests with a Matrox G250 card. The result is not fabulous. As you 
mention in the REAME, the fbdev output is not OK. I am able to see the 
startup channel message OSD and a VDR crash. Sometimes with a I-Frame in 
background.

As fbdev was not working I decided to switch to the directfb output 
possibility. I took today's CVS (directfb and dfb++). After complete 
installation I am able to run the fb_cpuload test program. That part is OK.
The plugins compiles without any warning. But when I run VDR I receive an 
error dealing with dfb_config. 

.../libdirectfb_fbdev.o: undefined symbol: dvb_config

I did some investigation and the error comes from the Creation of the 
dfb = DirectFB::Create(); line. 

Q1: Did you ever saw such an error in the past?

Q2: Is there any special reason to use a IDirectFB object instead of an 
IDirectFB_C object?

Cheers,

Alex

PS: I am running all my tests with a 2.6.0 kernel and VDR v1.2.6








-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe vdr" as subject.