[mythtv] Re: Teletext

[Martin Moeller]

I just updated the driver. I had to download a v4l2 tarball and install
that first, but after that there were no problems - BUT - It still seems
to need the patch to get any data...

fre, 2003-07-25 kl. 10:53 skrev Erik Arendse:
> At 25-7-2003 10:39, Martin Moeller wrote:
[is teletext stored as real text?]
> Yes, charbuffer with color & positioning info. Out of my head all 
> filelayout is defined in frame.h.

[various AleVT experiments]
> I presume you did get teletext with the patched AleVT?
> I don't know about this AleVT patch, but you could try to apply the same 
> changes to the libvbitext in myth (I borrowed that code from AleVT, it is 
> the part that is doing the vbi grabbing nitty-gritty).
Okay. The only file the patch touches on in version 1.6.1 is vbi.c
Is there any myth-specific code in that file or might it work if I
simply copy the prepatched version in there? 

BTW: I activated vbi_debug on this last invocation. MythTV is presently
recording from VH-1 and it seems the vbi is doing *something*:

saa7134[0]/vbi: buffer_activate [c53b46e0]
saa7134[0]/vbi: buffer_activate [c53b4760]
saa7134[0]/vbi: buffer_activate [c33bf6c0]
saa7134[0]/vbi: buffer_activate [c33bfb40]
saa7134[0]/vbi: buffer_activate [c53b46e0]
saa7134[0]/vbi: buffer_activate [c53b4760]
saa7134[0]/vbi: buffer_activate [c33bf6c0]

There are so many of these messages that dmesg -c is virtually useless..
That doesn't guarentee that it is getting what it should, though..?
There shouldn't be any subtitles on that program anyway, but it seems it
is at least looking for them.

hamm.c seems basically the same, except for variable types:

199c199
< hamm8(unsigned char *p, int *err)
---
> hamm8(u8 *p, int *err)
207c207
< hamm16(unsigned char *p, int *err)
---
> hamm16(u8 *p, int *err)
217c217
< hamm24(unsigned char *p, int *err)

and some cosmetic regarding arrays of numbers.

The same with hamm.h and vbi.h, except for this:

> #include "cache.h"
7d7
< //#include "cache.h"
9c9

Hmm.. except for certain lines being commented out and different
indentation, the unpatched vbi.c is exactly like the one in libvbitext.

That seems to indicate that if I can comment out the same lines in the
patched vbi.c there's a fairly good chance of it 'just working' (of
course, I could be wrong :)).

I assume that the commented parts are not used by mythtv. Are they a
hazard or just a waste of CPU cycles?

Thanks for your help so far Eric. I actually have hope now :)

/Martin.

> 
> Erik
> 
> 
> _______________________________________________
> mythtv-dev mailing list
> mythtv-dev at snowman.net
> http://lists.snowman.net/cgi-bin/mailman/listinfo/mythtv-dev