Bug #5757
New autorecordings should not record actual running streams partly
0%
Description
To me it happens very often that I see something worth in EPG for recording that's actualy on air.
If it's a series or a set of movies I create an autotimer out of the EPG.
The problem ist, that after finishing the autotimer TVH starts to record the actual running stream.
Combined with the parameter set "Record only once based on title" (which I normaly use), TVH records this stream partly an never gets a complete one.
TVH should wait until the next EPG entry to record right from the beginning.
If I want to record the stream instantly, I have use the normal record function.
History
Updated by Joe User about 5 years ago
Some people ( like me ) want to record and keep that partial recording in case a rerun does not occur. So, it would have to be added as an option and not the default...
Updated by Em Smith about 5 years ago
There is a "complex scheduling" tickbox in the recording profile on 4.3. That tries to find a better timeslot at the expense of more CPU and logging. One of the criteria for "better" is "not half-way through the programme", though other criteria may override it such as "I want HD and this is the only HD recording ever available."
It works for manually created autorecs and should probably work for EPG-created timers, though (over-the-air) EPG-timers sometimes include other IDs (that are unfortunately hidden on the UI) which have to match and can limit re-scheduling. For example "series link" on OTA EPG is often "this timeslot only" so can't be rescheduled to a better timeslot. Manually created autorecs have no such issues since they are definitely "title only".
There's a bit more information here:
https://github.com/tvheadend/tvheadend/commit/027377c53a08c6e713653d66f932a1217ffabde8
If it is working then the logfile will have entries stating "Replacing existing dvr recording".
The way I create my autorecs is to find the programme and then copy+paste the title in to a manually created autorec since I always have to alter the duplicate policy to "unique only", and it has the benefit of ensuring there are no hidden IDs.