[mythtv-users] playback performance regression: show of hands

Guy Paddock gap7472 at rit.edu
Mon May 22 16:40:35 UTC 2006


> Additional evidence of Myth's poor playback performance is that using
> mplayer or xine on the exact same hardware _always_ plays back without
> a blip.
>
> There is something wrong in the myth playback code, let's find it!
> Please?

I agree with Brian on this one; if root of the problem was the hardware,
other applications would similarly be affected when playing back the SAME
content. Myth seems to studder in some kind of buffering loop, like its
delaying the frames because the buffer is either too full or not full
enough.

For this reason, I've started looking over the player sources to find the
issue. So far, I'm seeing that the player needs to be split up into more
manageable classes. I don't care what programming experience you have,
5000-6000 lines of code in one class is insanely hard to read/debug/locate
errors in. I've even considered, on my machine, replacing the class with an
adapter class that uses the VLC API to handle playback. Only trouble is, I
now have two projects to understand (MythTV and VLC). Neither of the two
seem to like substantial developer docs.

I digress. For the features MythTV offers, the developers have done a superb
job making so much progress so quickly; I don't mean to throw their hard
work back in their faces. I just feel that it's at the point where, unless
some design changes occur, no more features can get added and no more bugs
can be fixed because things are so large and hard to maintain.

--Guy Paddock



More information about the mythtv-users mailing list