Topfs2 Wrote:The team is walking down the road of supporting TV in XBMC, currently playing mythtv is as easy as adding source myth://user:pass@ip if you have the server set up correctly. Though theres much missing ie schedualing recordings and manage recordings in a super great way (just a big list now), but plain play of recordings and play of livetv works.
The thing we want is the crossplattform way and support a number of different backends no matter the OS XBMC runs on, so mythtv is the starting point but on trac there is a patch for media portals tv-server and at some point Im sure an osx equavilant will pop up.
some read:
Incase you havent read this: http://wiki.xbmc.org/?title=GSoC_-_Unified_PVR_Frontend
http://forum.xbmc.org/showthread.php?p=2...post251506
http://forum.xbmc.org/showthread.php?tid=34186
Thanks for the reading tips and I must say it looks very interesting.
I have one concern and that is the zapping time when using streaming. You tell me cause I have never tried this...
I unsuccessfully tried out the Tubox client for my Dreambox 600 PVR but only got a very jerky picture as a final result, nevertheless the cashing time was a pain in the but to say the least. Satellite TV is is slow as it is and adding an extra delaying factor will have a bad impact on the user experience. THAT is my only concern.
Cause how I understand it is that you are streaming the picture/sound/metadata from the PVR server to the XBMC client. XBMC is parcing that information into a interface. That is the whole idea right? Cause you want the encoding to happen elsewhere and on not on cllient/XBMC side it self...?
I'm not a programmer but surly there must be a big difference between streaming from a PVR server and actually having built in hardware support for DVB cards in the actual XBMC. Basically you don't have to stream from one box to another and it is less hardware intensive...
On the other hand creating an API for easy PVR server integration makes sense. As I said my only concern is the zapping time if using streaming protocol between PVR server and the client. Have you ever been to a hotel that has IP TV and pay per view services? Sometimes it could take up to 4-5 seconds to zap between channels... different solution all together but worth thinking about. Otherwise we gonna have a big chunk of the XBMC community going heavy on the Losec (Ulcer medicine)
Tell me if I'm wrong, but what if you could integrate the MythTV logic/hardware support straight into XBMC instead? Wouldn't that make more sense? With that kind of solution, at least in the beginning, you might have to limit the hardware support to a number of devices but maybe it is worth the scarify..? You can still create the API... Some kind of MythTV light engine without the interface.
Cheers