[mythtv-users] 0.25, zero-size recordings caused by DST?

[Michael T. Dean]

On 03/12/2013 09:17 AM, Kevin Kuphal wrote:
> On Sun, Mar 10, 2013 at 5:58 PM, Robert M. Riches Jr. wrote:
>
>> With the version 0.25-20120418.3.mga2 packages on Mageia 2, all
>> three items that were supposed to record Friday evening (March 8)
>> were basically zero length--effectively failed to record.  Two
>> were based on the off-the-air program guide, and one was a manual
>> entry.  Also, the red light on the Hauppauge USB tuner was off,
>> and live TV was not working.  After a restart of mythbackend,
>> live TV worked, and yesterday evening's two recordings appear to
>> have been recorded.  I was not able to find a needle in the
>> haystack of logs that appeared to indicate the cause.
>>
>> With other/past experiences with MythTV not correctly handling
>> the transition to DST (recordings off by an hour), I'm wondering
>> whether the switch to DST last night (just over 24 hours after
>> the missed/empty recordings) might have been the trigger.  It
>> seems a little unlikely that a temporary USB tuner outage would
>> happen so close to a DST transition.
>>
>> Has anyone else noticed a correlation between DST transitions
>> and missed/empty recordings?
>>
> My understanding is that 0.26 is the only version of Myth that current is
> DST "proper" in how it handles the change.  I could be wrong but I believe
> that is the case.

That said, even on an old version of MythTV, there shouldn't be any real 
issues with DST--except that you'll need to use start early/end late on 
recordings that happen at the actual boundary.  And I can't imagine any 
possible way that DST could affect recordings that happen before the DST 
change (i.e. that occur 24 hours before the change).

I'm guessing it was just another case where the capture card failed to 
provide any data for MythTV to write.  The real question, then, becomes 
why didn't the capture card provide data (need a channel rescan, need a 
capture card reset/reboot, there was a signal or broadcast issue that 
caused problems while tuning, ...).

Mike