Mailing List Archive: MythTV: Dev

length mismatch in programinfo on QT4

Index | Next | Previous | View Threaded

kkuphal at gmail

Apr 9, 2008, 6:59 PM

Post #1 of 5 (610 views)
Permalink

length mismatch in programinfo on QT4

Hi. I just took the plunge to QT4 on my two FC6 systems. The backend was a
clean compile of new sources (16983) and started up fine. My frontend is a
brand new built FC6 frontend from the same revision but when I try to view
Watch Recordings I get an error of a mismatch in the length of programinfo
even though that struct hasn't changed in *forever*. Any ideas why this
would be? A verbose log from the system only shows the command to get the
PLAY list right before the error. I checked by includes both in and out of
my source tree and both have 46 as the length.

Kevin

kkuphal at gmail

Apr 9, 2008, 7:15 PM

Post #2 of 5 (578 views)
Permalink

Re: length mismatch in programinfo on QT4 [In reply to]

Something seems wrong

In remoteutil.cpp at RemoteGetRecording list, I have:

cerr << "lines: " << NUMPROGRAMLINES << "\n";
cerr << "recordings: " << numrecordings << "\n";
cerr << "size: " << (int)strList.size() << "\n";
Where I get:

lines: 46
recordings: 316
size: 1

And then the test:

if (numrecordings * NUMPROGRAMLINES + 1 > (int)strList.size())

fails

Why is that occuring?

Kevin

kkuphal at gmail

Apr 9, 2008, 8:05 PM

Post #3 of 5 (581 views)
Permalink

Re: length mismatch in programinfo on QT4 [In reply to]

On Wed, Apr 9, 2008 at 9:15 PM, Kevin Kuphal <kkuphal[at]gmail.com> wrote:

> Something seems wrong
>
> In remoteutil.cpp at RemoteGetRecording list, I have:
>
> cerr << "lines: " << NUMPROGRAMLINES << "\n";
> cerr << "recordings: " << numrecordings << "\n";
> cerr << "size: " << (int)strList.size() << "\n";
> Where I get:
>
> lines: 46
> recordings: 316
> size: 1
>
> And then the test:
>
> if (numrecordings * NUMPROGRAMLINES + 1 > (int)strList.size())
>
> fails
>
> Why is that occuring?
>

I put some debugging in the backend and it is failing on query.next() in
HandleQueryRecordings despite the size of the query returning 316 as
expected.

If I run the same query in a mysql window I get the same 316 rows
returned. I'm very confused here

Kevin

gtgj at pacbell

Apr 10, 2008, 12:44 AM

Post #4 of 5 (576 views)
Permalink

Re: length mismatch in programinfo on QT4 [In reply to]

>>>>> Kevin Kuphal writes:

k> Hi. I just took the plunge to QT4 on my two FC6 systems. The backend was a
k> clean compile of new sources (16983) and started up fine. My frontend is a
k> brand new built FC6 frontend from the same revision but when I try to view
k> Watch Recordings I get an error of a mismatch in the length of programinfo
k> even though that struct hasn't changed in *forever*. Any ideas why this
k> would be? A verbose log from the system only shows the command to get the
k> PLAY list right before the error. I checked by includes both in and out of
k> my source tree and both have 46 as the length.

I ran into this problem, too. In my case, it seemed to be
caused by some recordings with bad bookmarks. After I did this in
mysql:

update recorded set bookmark=0 where bookmark=127;

... the problem went away. I haven't tried to debug it further than
that.
--
Gregorio Gervasio, Jr.
_______________________________________________
mythtv-dev mailing list
mythtv-dev[at]mythtv.org
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev

kkuphal at gmail

Apr 10, 2008, 6:39 AM

Post #5 of 5 (564 views)
Permalink

Re: length mismatch in programinfo on QT4 [In reply to]

On Thu, Apr 10, 2008 at 2:44 AM, Gregorio Gervasio, Jr. <gtgj[at]pacbell.net>
wrote:

> >>>>> Kevin Kuphal writes:
>
> k> Hi. I just took the plunge to QT4 on my two FC6 systems. The backend
> was a
> k> clean compile of new sources (16983) and started up fine. My frontend
> is a
> k> brand new built FC6 frontend from the same revision but when I try to
> view
> k> Watch Recordings I get an error of a mismatch in the length of
> programinfo
> k> even though that struct hasn't changed in *forever*. Any ideas why
> this
> k> would be? A verbose log from the system only shows the command to get
> the
> k> PLAY list right before the error. I checked by includes both in and
> out of
> k> my source tree and both have 46 as the length.
>
>
> I ran into this problem, too. In my case, it seemed to be
> caused by some recordings with bad bookmarks. After I did this in
> mysql:
>
> update recorded set bookmark=0 where bookmark=127;
>
> ... the problem went away. I haven't tried to debug it further than
> that.

Well, I reverted to my pre-QT4 upgrade DB backup, recompiled after putting
back the QT3 rpms and the machine works fine again. Maybe it's something
with the QT4 MySQL driver on FC6. I may try this again but with some
virtual machines. Did any other developers move to QT4 on FC6?

Kevin

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact lists@gossamer-threads.com