hello, can somebody send me some link (website) where I can learn how to use video codec (compression) on linux ?
I mean.. is there a common API to record video independently of the codec ?
sailus: I answered your comments about patch 4/5... I appreciate if you could take a look on it as soon as possible, as my intention is to send the patches fixing au0828/snd-usb-audio early next week, during the merge window
mchehab: I'll check that within the next few hours.
mchehab: media_device_init() doesn't really register anything at the moment.
For that reason I wouldn't change the naming so. Naming-wise it's also a bit confusing to re-purpose functions for the users.
What I'd propose is to simply rename the resulting media_device_unregister() as media_device_unregister_cleanup().
Then its name corresponds what it really does.
What do you think?
That should be more user friendly.
works for me
mchehab: Excellent.
sailus: I'll send a separate patch with the rename
Ok for me.
good
patch sent
if you're ok with patches 4/5, 5/5 and this new one, please ack/review
I'll check them next.
thanks!
Done.
sailus: ping
janaszewski: Pong.
sailus: hi, how are you?
Fine, thanks!
How about you?
fine too, thanks
did you see my reply: http://www.spinics.net/lists/linux-media/msg97620.html ?
Uh... I did but I forgot to reply. :-(
I'll do that over the week-end.
(Just ping me if you don't receive a reply within a reasonable amount of time.)
it's most vital to me on how to proceed with opeining subdevs  before executing ioctl in libv4l2subdev.c
Yeah.
I've continued working on libmediatext since the initial patch.
I asked about a possibility of removing v4l2_subdev_open from that functions and rely on the media device open
I'm using it as a testbed for request API test program.
That's extending the scope a bit. The code is experimental so far.
ok, I am not sure when I will be able to post the updated set anyway
I've just got back to v4l-utils for a moment and rebased my patch set to master branch
and there is a problem with media-bus-format-codes.h - it isn't being created
despite that I see relevant entries in Makefile.am in media--ctl
I am doing autoreconf -vfi before build
isn't it sufficient for updating v4l-utils build system?
There's a script called bootstrap.sh or such.
But I think autoreconf -vfi should be enough.
That's what I mostly use.
ok, bootstrap.sh created the file
thanks :)
Hmm.
configure.ac:85: error: required file 'build-aux/config.rpath' not found
Did you get that?
I'm not sure everything gets build correctly if you get that error.
Just creating the file helps though.
sailus, pinchartl: is an enpoint a required sub-node for each port, or it is optional?
neither Documentation/devicetree/bindings/graph.txt nor Documentation/devicetree/bindings/media/video-interfaces.txt are clear about it
sailus: no, I don't get rpath not found error
javier__: Good question. I have to let pinchartl to respond that one.
sailus: Ok, thanks
my understanding for reading the docs is that at least one endpoint should be required but it seems other think that is optional so drivers are not consistent on how that case is handled when parsing the nodes
s/for/from
I guess there could be some value on making optional so for example a video interface with ports could be in a .dtsi file and a .dts only adding an endpoint
but IMHO there is little point on doing that and is more confusing
anyways, I'll see what Laurent says