[mythtv] QGVector errors

[Stuart Auchterlonie]

Torbjörn Jansson wrote:
> mythtv-dev-bounces at mythtv.org <> wrote:
> 
>>Torbjörn Jansson wrote:
>>
>>
>>>
>>>I've done a test, i deleted program and credits table so i know all
>>>data i get is new data. Now most EIT programs don't have any
>>>descriptions and the descriptions on hallmark looks cut off, just
>>>like in the begining of the EIT code when it didn't know how to
>>>assemble those long descriptions that was cut of into multiple
>>>sections/packets/whateveritscalled. 
>>>
>>
>>Ahh, there was a ticket in the last two weeks about EIT data not
>>being inserted. It turned out that the problem in that case was
>>that the complete set of sections was never collected and thus
>>never emitted to the database.
>>
>>*research*
>>http://svn.mythtv.org/trac/ticket/344
>>
>>There was a patch that went in based upon 344, which
>>basically short circuited that and faked the response that all
>>the sections had been collected. This may be the source of your
>>problem. 
>>
>>I'd suggest backing out the changes in the changset
>>http://svn.mythtv.org/trac/changeset/7573 and see what
>>happens. (ie. Just get the patch and apply it with patch -R)
>>
>>John, any comments?
>>
>>
>>Stuart
> 
> 
> I'll give that a try in a while.
> 
> About the empty descriptions, i've dumped the eit data with dvbsnoop and the
> descriptions is there.
> I've checked my fixup function to verify that i have a description at the
> end, and i do.
> 
> How do i revert one changeset? Is there a way to get a diff of a changeset
> so i can use patch to apply it in revers?
> 

Follow the link to the changeset.
Down the bottom of the changeset there is an option to download
the diff in unified diff format.

Then at the myth toplevel directory do
patch -p1 -R < thepatch

and off you go :-)