[mythtv] EIT table parser

Stuart Auchterlonie stuarta at squashedfrog.net
Tue Mar 21 11:35:04 UTC 2006


On Tue, Mar 21, 2006 at 11:15:56AM +0000, Chris Birkinshaw wrote:
> 
> 
> A couple of questions about the table parser:
> 
> 1. It seems to only start when the backend is recording - is this correct? If so
> why? It caused me problems recently when the machine had been off for a few
> weeks.

The eit scanner, if it has been enable, will sit there and walk your channels
downloading the eit data

So you might need to check if you compiled with --enable-dvb-eit and have
the datasource for your card configured to use the eit datasource.

> 
> 2. Which sub-tables does it look at from PID dec 18?

All of the ones you are interested in.
0x4e/0x4f: now next tables
0x50-0x60: future event tables

> 
> And some info:
> 
> In the near future the system which produces the DVB SI for the UK network will
> receive triggers from the program providers at program junctions. For example:
> 
> Now: 10:00 Sport
> Next: 10:30 Weather Update
> 
> If the Sport overruns and a trigger is not received until 10:34 then the EIT for
> this service will not change until 10:34, at which point it will become:
> 
> Now: 10:30 Weather Update
> Next: 10:45 Something else
> 
> And the table version number will be incremented.

There is currently work in progress to better handle the eit data
that is broadcast, to prioritize according to what tables it comes
from.

> 
> If mythtv were to hold off recording until the scheduled recording event became
> the EIT "now" event then it would always start recording exactly at the correct
> time.

Reschedules are done at the end of collection eit data or 1 minute since the
last update.

There is currently no explicit support for what you are after,
but it should work okay...

> 
> Its worth mentioning that this is currently the case for the EIT"actual" (i.e.
> for originating transport) for multiplexes 1 and B (would allow testing) but
> not the case for the other multiplexes EIT"actual" or the EIT"other"
> cross-carry SI information.
> 


Stuart



More information about the mythtv-dev mailing list