[mythtv] Bringing it all together...

[Joseph A. Caputo]

On Thursday 02 October 2003 02:59 am, Trey Ethridge wrote:
> 
> Hello all,
> 
> I've been thinking to myself lately about mythtv and I wanted to see what 
other developers thought.  I mostly love everything about mythtv.  However, 
there is one thing that worries me.  The biggest problem I see is lack of 
integration.  I'm worried that this will limit the quality of mythtv in the 
future.  You see, I feel that it's a huge problem that each module doesn't 
really know about the other modules.  
> 
> For example, mythmusic maps the volume keys differently than mythvideo and 
when I leave a module, it quits whatever it's doing, whether that's playing a 
song or video.  Wouldn't it be cool if I could select some music and then 
switch over to mythgallery to show some pictures while my music continues to 
play.  I'm sure quite a few of you know that you can do this with other 
products.  The same thing could go for having a picture in picture type of 
effect with mythweather taking up most of the screen and having the tv in the 
smaller PIP box so you wouldn't miss any of your show while you checked your 
weather.
> 
> Of course I understand that this is because we are using (reusing?) existing 
products to quickly have a working product.  However, I wonder if our 
approach will ever reach the tight integration of something like M$ Media 
Center PC or Tivo's HMO. 

Exactly what do you mean by 'using (reusing?) existing products'?  You must 
have Myth confused with Freevo :-)  With the exception of MythVideo (which 
uses an external player) and XMLTV (which is not a runtime component, anyway, 
but a data feed/grabber), I can't think of any parts of Myth that are 
external to the code base.  Oh, sure, there are *libraries* that Myth depends 
on, but all of the GUI code and all of the video recording/playback, music 
ripping/playback, music visualizations and DVD ripping (well, the daemon, 
anyway) is all home-grown Myth code (though some of it uses other OSS 
software that has been 'captured' into the Myth source tree, and even that is 
pretty much limited to very low-level stuff like the actual encoding/decoding 
of video/audio data).  So, to answer your question, yes, we could achieve the 
level of integration you've suggested if (1) the Myth community (esp. 
developers) wants it, (2) we come up with an appropriate architecture/road 
map, and (3) we have the manpower to implement it.

-JAC