[mythtv-users] trunk: mythfrontend does not show all recordings, SOLVED

Michael T. Dean mtdean at thirdcontact.com
Wed Jan 13 23:37:29 UTC 2010


On 01/13/2010 06:30 PM, Juergen Sachs wrote:
> Am Mittwoch, 13. Januar 2010 23:59:33 schrieb Michael T. Dean:
>   
>> I could have sworn I had already asked you to do this, but looking
>> through the ML archive, I don't see my asking you.  Perhaps I had asked
>> someone else, or perhaps the message is still in the "Unsent" folder on
>> my laptop (being busy with travel last week, I may well have forgotten).
>>
>> Anyway, do you see an error in your frontend log files about, "length
>> mismatch between programinfo"?  Can you please run:
>>
>> cat << "EOF" | mysql -umythtv -p mythconverg
>> SELECT COUNT(*) FROM recorded WHERE bookmark NOT IN (0, 1);
>> EOF
>>
>> If the value is anything other than 0, that's the problem, and you can
>> fix it with:
>>
>> cat << "EOF" | mysql -umythtv -p mythconverg
>> UPDATE recorded SET bookmark = 1 WHERE bookmark != 0;
>> EOF
>>
>> as mentioned at http://svn.mythtv.org/trac/ticket/7467#comment:2
>>
>> Note, also, that eventually Qt will be fixed.  There are bug reports at
>> http://bugreports.qt.nokia.com/browse/QTBUG-5758 and
>> http://bugreports.qt.nokia.com/browse/QTBUG-5765 .  QTBUG-5758 has been
>> fixed for Qt 4.6.1 (so with 4.6.1, Myth would actually see the failure
>> and tell you things are wrong).  QTBUG-5765 is still awaiting review
>> (though you can apply that patch if you happen to compile Qt yourself),
>> and it will fix the problem completely (i.e. you won't have the
>> truncation, so everything will work--even with invalid data).  However,
>> it's much easier to do the update above, if you do have invalid data,
>> and you'll never see this problem again (even with a buggy Qt).
>>     
> The select reports "1".
> So I did an "UPDATE recorded SET bookmark = 1 WHERE bookmark not in (0,1);" an 
> it SOLVED. (Changed only the one line that seems to be buggy, hope thats fine 
> too)
>   

Good.  Thanks for reporting back.  I'm sorry it took me so long to get 
you to the right answer--I /really/ thought I had already had you check 
that.

Too bad that by the time Qt is fixed, no one will have this problem anymore.

> By the way I do not saw the error message about length mismatch before running 
> the SQL query. I had that in the past where I saw nothing, but now it is gone. 
> Also when I saw it in the past I found out, that mythbackend did not a schema 
> update, I think it was because of mythfrontend running already.

Yeah, that's another case where you'd get the same error message--but 
unrelated to the root cause here.

Mike


More information about the mythtv-users mailing list