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

[Ronald Frazier]

Well, I've had some good luck, after a bit of bad luck.

After getting things just barely working on Monday, I tried testing it
on Tuesday and found I had some major problems. In stripping out a lot
of code to get things going, I found that I had broken the seek table
generation, and the video was not seekable until you ran mythcommflag
to rebuild the seek table. Not being to thrilled about having to dig
into that, I turned my attention to tuning the card. To put things
bluntly, I absolutely failed at figuring out how to query it via UPNP,
and I spent hours going nowhere trying to find the right methodology
to make the SOAP calls to control it via the DRI interface. It was a
very disappointing day, but determined to salvage something of the
days work, before I went off for bed, I decided to skip the
UPNP/SOAP/DRI stuff and just make HTTP calls to the web server.
Hackish? Maybe a bit, but I was able to get channel changing working,
so at least something had gone good. I've now got a way to control and
query the state info for the card.

So, today's task was figuring out what to do about my other major
problem. Well, I started to look into things, and I discovered the
signal monitor did a LOT more than I thought it was responsible for.
So I started to look through the code and realized it was starting to
make a bit of sense. So I added it back in, along with a bunch of
other stuff I had stripped out (and thus I also disabled my hack of
writing directly to the ringbuffer). Of course, everything was now
broken, so I had to work through things one step at a time. It took me
the better part of the day to deal with, but just a short while ago, I
had some great success. Playback is once again working, seek table
generation is now working, and my 15-second channel changing is now
only about 5 seconds. And since it's now utilizing all of myth's
existing MPEG packet stream processing, it probably fixed a whole slew
of bugs I had yet to discover.

So, my task for tonight is going to be to clean up some of the code a
bit, and then tomorrow, I'm going to try and start from scratch again
and see if I can get this all going in 0.24-fixes (since it seems like
there's a lot going on in 0.25 right now, and I'm not sure I want to
run it on my production system). Once I get that taken care of, my
next task will then be to order my cable card and get that all
working. I think I already understand how everything there works, and
I suspect I've already got the code necessary to deal with it, but you
know how things go...I'm sure something will come up.

So at this point, I'd say I have a pretty high degree of confidence
that I'll be able to get this all working quite well in about another
week or so . . . though I still don't think I'm going to tackle
channel scanning properly. I'll still probably write a perl script to
scan via the web interface and generate the SQL to insert it right
into the database (hey, I've gotta keep Michael on edge somehow).

PS. My first scheduled recording just finished commflagging, and it
all worked like a charm. :-)

-- 
Ron Frazier