[mythtv] [mythtv-commits] Ticket #1888: EIT failures with Finnish channels (DVB-T)

Michael T. Dean mtdean at thirdcontact.com
Mon Jul 10 19:56:54 UTC 2006

On 07/10/2006 02:31 PM, Otto Kolsi wrote:
>>> If this PES packet situation is not related to the bogus EIT data,
>>> then I think we have two separate problems to be fixed.
>> Probably, I'm debugging the bogus data now.
> I haven't really paid attention to 'airdate' column. Now that I look my 
> program data in DB, I can see a bit strange values. Query "select 
> distinct(airdate) from program" gives me two values:
> 0001-01-01
> 2000-01-01

Jumping into the middle of a thread I haven't read, so ignore me if this 
info isn't helpful...

What DBSchemaVer are you using?  The airdate column of program is 
supposed to be a YEAR(4) with a default of 0000, and the comments in 
dbcheck.cpp say:

'airdate' is a string representing the year of release for movies and 
may have no meaning for other types of shows.

In DBSchemaVer 1051, airdate was changed to a DATETIME NOT NULL but in 
the oldprogram table, not the program table.

Note that there's also an originalairdate column of type DATE with a 
default value of NULL in the program table:

'originalairdate' if provided is the date when a show was, or will be, 
first televised.

That's the one that should look like a date.

I checked 0.18, 0.19, 0.19-fixes, mythtv-eit, and trunk, and none seem 
to have ever had airdate as anything other than a YEAR(4).  You 
mentioned "SVN 10047" (which had DBSchemaVer 1141), so I'm assuming 
either mythtv-eit or trunk r10047 when I say it definitely looks wrong.

> Obviously this isn't right.

Corrupt database?  Invalid schema?

>  Maybe we should check this out and then that 
> might help or even correct this bogus EIT data problem..?



More information about the mythtv-dev mailing list