[mythtv] [Q]: EIT giving overlapping time slots

[Emmanuel]

On 30/09/2011 09:40, Stuart Morgan wrote:
> On Friday 30 Sep 2011 09:36:33 Emmanuel wrote:
>> Yes I understand that but I would like to override this rejection if
>> possible, because this is all the data I can get (its for local channels
>> on the sat here in the FWI).
> I'm going to ask a silly question at the risk of getting a silly answer, but
> why are you collecting the EIT data externally to MythTV? Why not allow
> mythbackend to collect it instead?
The question is not silly so I hope my answer wont be either ;-)
In fact its not a standard EIT, it is a proprietary one that I have to 
decode myself. The good thing is that it is long term EIT (up to 8 days 
IIRC), moreover no grabber that I know of has these channels correctly. 
So instead of having no data, I have bad data, I guess it is still worth 
it though.
I also made a patch for mythtv to collect this by plugging my decoder 
into active scanning, it worked well but prevented my backend from 
sleeping. Also it is on one TP so you cant really do passive scanning as 
if you are unlucky (which has happened to me) you dont collect data for 
a long period of time (some other hacks can be done to fix that) so 
finally I decided to run it when mythtv goes to sleep and use mythfillDB.

My idea would be to ask mythfillDB to forget about the start or end time 
in case of a <=5 mins overlap. Otherwise I would have to scan the full 
xml file to look for overlap and correct these, whereas mythfillDB has 
probably already part of the machinery to do that.
But I may be wrong.
TIA
Bye
Manu