[mythtv-users] Issues since upgrading to 0.25

[Enigma]

Wow, lots of replies, I will try to reply inline to all points.

Michael Dean wrote:

> > 2. Commercial Flagging. Since the upgrade most of my recorded
> > programs don't get the commercials flagged.
> > ...
>
> There are rumors that jobs aren't being started for the 2nd recording
> with back-to-back recordings.
>

OK, looking at the pointer to the bug that David Engel posted, it looks
like this is what is happening to me.  I do a lot of recordings so it
stands to reason that I would be seeing some back-to-back recordings and be
hit by this bug.  I will monitor the bug and upgrade once a fix is released.

> 4. MythMusic. Everybody else hates it, why shouldn't I? j/k! The
> > new music interface does seem to have a problem remembering what
> > playlist I had selected before, each new time I go into music it tells
> > me I haven't selected any tracks to play. Is this the normal behavior
> > or do I have something configured wrong?
>
> Perhaps you're not exiting mythfrontend nicely, but instead letting it
> get killed? Start up mythfrontend, queue up some playlist, then exit
> mythmusic and exit mythfrontend--using the EXIT key, and selecting to
> exit (don't use the shut down or reboot options you may have). Then go
> back in and see if it remembers things better.
>

If that is the case I'm sure that's what is happening.  The only time the
frontend restarts either a) the frontend froze and I had to kill it or b)
the mceusb module crashed and I need to reboot the frontend.  I don't think
I have ever exited the frontend from the interface.  Is there any chance of
this behavior changing (like saving the playlist out when exiting MythMusic
rather than the whole frontend or writing it when getting a SIGINT) or is
this pretty much "the way it's going to be"?

xtragb at gmail wrote:

> > 1.  HDPVR.  I am currently using a HDPVR hooked up to a DirecTV HD box.
> I
> > am using http for channel changes, the HDPVR is not using the blaster.
> > Video is through component cables, audio is analog through the RCA
> jacks.
> > Since the upgrade the HDPVR has been extremely unstable.
>
> I have experienced this on my fully updated Arch Linux box. In my
> case the lockups happen on a specific channel, apparently caused by
> audio dropouts on the STB which is connected to the HD-PVR using the
> optical input. Disconnecting the SPDIF cable while the HD-PVR is
> recording will reproduce the same lockup every time. Interestingly,
> the HD-PVR can recover from a brief audio dropout but a longer one
> will take it out. I don't see any recent changes in the HD-PVR driver
> that would have affected this issue one way or another. It could be a
> Myth change though; I've been running .25 exclusively since I got my
> HD-PVR so I'd have to take your word for it that it's worse with this
> version than it was with the previous one.
>

I don't think my freezes are related to the audio as I am using analog
inputs, from what I understand the issues with audio on the HDPVR was for
the SPDIF input.  Under 0.24-fixes it was reliable but since the upgrade it
has been very flaky.  I assume it is a driver regression but I posted the
issue on the MythTV list on the off chance that 0.25 is using the device in
a different manner than before which might account for the different
behavior.


acstadt at stadt wrote:

> > Some TVs only output stereo over their spdif out connector. So it's
> > hard to say where the problem lies.
> >
> > Does your receiver have HDMI inputs? If so you may be better with
> > mythbox->receiver->TV
> >
> Never saw the op's post, so:
>
> The trouble is with his setup is that the eeeBox is only seeing the edid
> info from the TV (which probably only says it supports 2 channel
> stereo). Really he has two options, the easiest, as you suggested, (if
> possible) is to go to the receiver then the TV. The other option is to
> hack the edid extended data block and then load the hacked block with
> nVidia's customEdid option.
>
> I'm not entirely sure this would work with the latest nvidia drivers
> (they seemed to have changed some of the edid handling/validation), but
> you theoretically could plug the computer into the receiver, capture the
> receiver's edid, plug into the TV, get a mode debug of the TV's edid.
> Build a set of custom modelines for Xorg based on the TV's edid. In
> xorg.conf, load the receiver's edid instead of the display's edid (via
> the customEdid option), but then force X to use your custom modelines
> for the display. The catch hear is that you can't ignore the edid after
> loading a customEdid (at least in my experience) so you have to use the
> modevalidation, noedidmodes, usedisplaydevice, connectedmonitor, dpi,
> and exactmodetimingsdvi options to accomplish your overrides. FWIW, I
> had done something similiar to this on one machine, but in the end went
> and out and got a better receiver - still had to use the custom
> modelines to get the most out of the display (and which also
> incidentally fixed a few other minor issues with Xorg) but at least the
> audio just worked.
>

Yep, I think that is what is going on here.  I assumed that if I was using
the passthrough output on the TV it would report the capabilities of the
device on the passthrough rather than its own capabilities but obviously
that assumption is wrong and the TV is just reporting its 2 channel
capability.  I don't have any HDMI inputs on my receiver but the frontend
does have what appears to be an optical output (1/8" jack labelled as
S-PDIF so I assume it is one of those dual-purpose analog/optical jacks) so
I should be able to just run  the digital audio directly to the receiver.




On a side note, were there big changes to mythfilldatabase in 0.25?  I used
to get through a run fairly quickly but since the upgrade it is taking a
loooong time to run.  Last night's run started about 9 hours ago and it is
still running.  has anybody else seen a degradation of performance with
mythfill?


Overall, I can't say that I have been pleased with the upgrade.  I was
hoping some of the issues I was seeing on 0.24 would be resolved - while
some things are working better (like starting channel in live TV) some
issues from 0.24 remain (like all the frontends choking when a new
recording is started).  0.25 seems to have introduced some new bugs,
hopefully it will all work out in -fixes.  Thanks for all the replies, I
will see what I can do to work around the issues until the fixes are
committed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mythtv.org/pipermail/mythtv-users/attachments/20120604/f9ea22f2/attachment.html>