[mythtv-users] Got my Ceton infiniTV 4...any else have one yet?

[Ronald Frazier]

On Tue, Jun 28, 2011 at 12:06 PM, Tom Bongiorno <two.bits.11 at gmail.com> wrote:
> I am glad to see progress, and I am very interested in trying this.  Which
> version or branch of MythTV did you end up using?  I was hoping .24-fixes,
> but I assume since you said "compile myth from source" that you are using
> .25 master, dkristjansson/mythtv-rec, or mythtv-rec2.  Did you use the
> import recorder as you tried early on or something else?

I've played around with all 3 branches (.24 and .25 trying to get it
working with the import/demo recorder, and daniel's branch before I
learned his code wasn't yet usable). When I finally started toying
around with the code directly, I think I was working off of .25 (It's
certainly not daniel's branch, as the OCUR recorder files are not
there). However, since I plan on putting this into production on my
main system, I was thinking about going with .24 since it's probably a
bit more stable for now. I don't think it should take too much effort
to get it working, since (as far as I know) I don't think I'm relying
on anything .25 specific. I probably should have started with .24 to
begin with, but I've done so much tinkering with this...I kept
switching versions, and then before I knew it I had started writing
code on whatever branch I had last tried to get working :-)

My code is based off a very stripped down version of the hdhomerun
recorder. I didn't go with the import recorder as my code base because
it seemed like it was designed to function in a much different way,
and I lacked the knowledge of myth to know how to make it do what I
wanted. The HD Homerun recorder, on the other hand, seemed like a
perfect fit for me to start with. I'm already intimately familiar with
that device's functions and protocols, so when I look at the myth code
it's much more apparent to me exactly what everything is trying to do,
and it makes it easier for me to figure out what the corresponding
code should be.

That's made it a lot easier, though there's still a ton I don't
understand how to make work properly. For instance, I've got a very
basic knowledge of the MPEG stream format and the different types of
packets it contains. However, I was not able to figure out how to get
the MPEGStreamData class to like the packets I was sending to it. It
was able to find the PAT/PMT information, but it complained that the
tables didn't contain info for the program I wanted to record, and as
a result, it was never forwarding the packets on to the recorder to
store in the ring buffer. After being unable to figure out what I did
wrong (and I wasn't even sure if I was on the right path yet, since I
had a bunch of hacked code that had yet to produce anything usable), I
hacked it a bit more, bypassed the MPEGStreamData helper completely,
and threw the data straight into the  ringbuffer after reading it from
the /dev device. With this, I had video from the ceton in mythtv for
the first time. I'll figure out tonight if I broke anything serious by
doing that. If so, then at least I know what I have works and I can
start adding any necessary elements back in one by one.


-- 
Ron Frazier