[mythtv-users] [mythtv] CC benchmarks and procedure list ? Was: how would one speed up the channel changing in mythtv ?

[Doug Young]

On Mon, Oct 20, 2008 at 11:00 AM, Rob Smith <kormoc at gmail.com> wrote:
> Like I said, it takes my ir blaster script a long time to run, because
> I slowed it down so my cable box doesn't miss presses. That's not
> myth's fault, but it adds into the overall time that channel changing
> takes.

I'm also moving to the users list...

I had a similar problem...my blaster script (I use a MyBlaster, which
unfortunately was discontinued) was taking several seconds, and mythtv
was waiting for it to finish before it would proceed.  I also had
problems when children, who were frustrated with the slow channel
change, pressed multiple keys on the remote, and multiple processes
got launched contending for the serial port, sending mixed signals,
and generally making things worse.

My solution was straightforward.  I took the actual perl to change the
channel and made it into a daemon that listens on a fifo object.  Now,
the script that myth calls simply writes the channel change (or power,
or volume, or whatever) out to the fifo and exits, assuming the daemon
will take it from there.  Now myth changes channels fast enough that I
can see the tail end of the blaster change in the new buffer.  It also
has the advantage of lack of contention...since the daemon has full
control of the serial port, there's no contention, no bad
signals...everything is automagically serialized and waits it's turn.
I've never got out a stopwatch, but my kids have stopped complaining
about the wait, and I haven't missed a channel change since.
-- 
Doug