[mythtv-users] What might cause a recording to mysteriously stop?

[f-myth-users at media.mit.edu]

    > Date: Tue, 11 Apr 2006 10:40:49 -0400
    > From: "Michael T. Dean" <mtdean at thirdcontact.com>

    > On 04/11/2006 03:36 AM, f-myth-users at media.mit.edu wrote:
    > >     > From: "Michael T. Dean" <mtdean at thirdcontact.com>
    > >
    > >     >				 			 The only fix is to 
    > >     > close the device and reopen it.  0.19-fixes SVN (which is being used by 
    > >     > many packagers) and SVN head will do this for you.  ( 
    > >     > http://www.gossamer-threads.com/lists/mythtv/commits/182149#182149 )
    > >
    > > [I need this to tell my
    > > automation that the recording is defective and should be rerecorded if
    > > reaired.]

    > Why?  0.19 will finish the recording.  0.18 won't.  You won't be able to 
    > get any changes in the 0.18 branch, so there's no need to "rerecord" 
    > because of this.

I think I was unclear.

(a) Even if 19 will finish, the recording is glitched.  If it's
possible to automatically snag a fresh copy from some later airing
that isn't glitched, so much the better.  I want to enable that, by
making 19's log output slightly more informative.  (This is part of a
much larger task of checking for a whole variety of failures---some
caused by my end, some caused by the studio end---and rerecording
whatever we can.  The idea is to do as much machine checking as
possible so the request to rerecord happens as soon as possible,
possibly snagging a reairing that happens a couple of hours later
[common on many of my feeds] which might -not- still be airing the
next day by the time a human looks at it.)

[Note that if the ivtv errors aren't totally randomly distributed
across tuners, logging the tuner involved might also raise a red flag
if it always happens to a particular one---that would be a good way to
tell the user "throw this tuner away" or something.  I have no idea if
the distribution is random or not, since this is the first time I've
caught mine doing it and others' reports are sporadic.]

(b) Even though I'm running 18, I'm looking forward to the point at
which I'll be running 19.  If I can fix this error message now for 19,
so much the better.  Especially if my fix to the error message in 19
makes it into fixes and thus into packagers' outputs, which would be
even more convenient.