Porting the TV server to Linux (1 Viewer)

[Marcusb]

I think when InfinityLoop said Linux wasn't better than windows he meant in terms of trouble free installations. As it is we get a lot of people with trouble installing MP due to OS issues. Imagine if we double OS's.

Since MP is still only at beta 0.2 verysion, why don't we wait for Frodo and the gang to fully release version 1 at least, before we all start calling for radical changes to everything about the way the system works? Or should we tell Frodo and the rest of the devs now that they should stop everything, learn all there is to know about programming Linux and then start from the ground up?

Personally, I like the idea of MP on Linux, but I'm a realist. It will take a lot of work (despite all the people who pop up on these forums from time to time saying "I'm no programmer, but I'm sure it would be easy") and my main priority and I'm sure the overwhelming majority here would rather just have MP released and as good as possible before we branch off on other tangents.

Simply, we don't have enough devs to work on something like this as well, so unless someone new comes along with a lot of time to do this, it probably isn't going to happen any time soon.

 

[HappyTalk]

I've not read up much on the new MP client server architecture, but surely if the recording engine & streaming/conversion plugin were constructed as web services then it could be accessed via a web server supplying a web front end to any machine or via a full blown app talking directly to the web service running on any machine. Such an engine would be fairly o/s independant (as it does no gfx output/user input) and so would compile under mono I'd have thought. O/S (and language) independance was why dotnet used managed code.

I guess if all the plugins were web services doing most of the complex stuff with lightweight front ends to handle user in and video out it would account for most permutations and offer real flexibility. Instead of chossing different skins, we could choose totally different front ends, even simple ones for our k800i's!

 

[piranha]

Webservice for streaming media? LOL

 

[HappyTalk]

Webservice for streaming media? LOL

No for offering an o/s independant, simple, reliable method for client and server to communicate that readily lends itself to mp being as its developed in c#. The streaming of actual binary data would use whatever binary transmission technology is best suited to it.

 

[kerpow]

Ok, how about this. I know that this will result in me being shot down in flames, but that's how I learn stuff!

What are the specs of the TV Server. what does it actually do?

> SQL Database. Ok lets get the easy bit out of the way, it has a SQL database. (Possibly the easiest bit to get on Linux).

> RTSP It streams video via RTSP. Well doesn't VLC do that too?

> It records and time shifts

> Control It takes commands from clients.

As I see it, those are the basic 4 jobs of the TV Server. So we have 2 parts in Linux anyway, surely adding these other two parts is not that difficult. The hardest bit might be to do recordings (but VLC can help there too).

What am I trying to say? Is is possible to bolt together a few bits and bobs and get close to what we are after, a Linux TV server?

Maybe I am over simplifying, but if we are serious about doing this, then we at least need to come up with a real spec of what we are trying to achieve. And that goes for making some sort of connector for any other TV server too.

Fire away.

 

[rtv]

okay, for all pseudo conclusion makers here:
1. Frodo suggested to use MythTV as a server for Linux why bothering about a new competitor instead of helping out in that project?

2. Open up monodevelop (If you have no idea at all install Edgy Eft before) do 3-4 sample projects; go back into your windows and open up Visual Studio. Now tell me what you think about mono's status..

3. A "control" plugin for TVE3 would be very easy to write for anyone reading the source of it - you'll see...

4. A "control" plugin for MythTV was already linked here - it's even .NET! What are you all just discussing this - better start working!

 

[j1nx]

rtv, says it all.

Conclusion: If your OS is Linux, use MythTV. If your OS is windows, use MP

It has no point trying to port the server to linux, as MythTV is way ahead on linux.

If you like MP as client, but want to use MythTV as backend, there is a gap and i think it is OK to port the client to connect to a MythTV-backend, using WinMyth as example.

Any other "porting-option" sounds idiot to me.

 

[leo212]

rtv, says it all.

Conclusion: If your OS is Linux, use MythTV. If your OS is windows, use MP

It has no point trying to port the server to linux, as MythTV is way ahead on linux.

If you like MP as client, but want to use MythTV as backend, there is a gap and i think it is OK to port the client to connect to a MythTV-backend, using WinMyth as example.

Any other "porting-option" sounds idiot to me.

And that's of course can be useful only if you have more then one computer in the house....
The ironic thing is that the hardware requirements for the Linux myth-server is much lower then the requirements for the Windows MP-Client…

 

[jawbroken]

That is not ironic. It is usually a lot less work to stream an incoming broadcast to disk than it is to decode, scale, process and display that video.

 

[James]

This whole discussion is pointless and has already be discussed many times over.

It seems every time there is some new development in MP, out comes the cry "lets port it to Linux". MediaPortal is a windows, DirectX and .NET based application. If you cannot accept that then please stop wasting everyones time and find/join a product/project that suits your needs and tastes.

All this "I hate Windows/C#/etc. Lets port all the code to MY favorite
OS/Language", needs to stop, because that is what MP is. Take it or leave it.