[GEM-dev] Re: pix_video ieee1394 on linux

IOhannes m zmoelnig zmoelnig at iem.at
Thu Jun 30 16:46:57 CEST 2005


B. Bogart wrote:
> Hi,
> 
> Thanks for the clarification Chris, I had a feeling you would know the
> answer.
> 
> I got some ffmpeg sample code from the net and I'll take a look to see
> if that loads the file fine and go from there.
> 
> I thought I would mess around with my iBot camera today, I had to
> manually create the /dev/video1394/0 device, but now coriander sees the
> camera and will display the output. So the question is how do I tell
> pix_video to look at it? pix_video reports both 1394 and v4l support. I
> messed with "driver 0|1" and "device 0|1" but could not get anything out
> of it. I find these messages really confusing, often devices and drivers
> are considered the same thing.. blech. "device /dev/video1394/0" would
> actually be easier...

actually i am wondering now, why i didn't do it like that... i have no 
idea what happens if i open an ieee1394-device with the v4l-api. would 
it crash ? so that is why i chose to differentiate between the 2 
"driver"-APIs with a message, and to separate between hardware devices 
(camera1, camera2) with another message.

apart from that:
[pix_video] (in ieee1394 mode) tries to open 
/dev/ieee1394/dv/host#/{PAL,NTSC}/in, where "#" can be defined via the 
"device"-option. this device is created (with devfs) by the module "dv1394".
i do know that devfs has been deprecated for some time, and i don't know 
how udev behaves in this way.
the fallback (if there is no /dev/ieee1394/dv/host#/{PAL,NTSC}/in) is 
/dev/dv1394 (which might be created by udev (or by hand :-())

mfg.asd.r
IOhannes




More information about the GEM-dev mailing list