Jump to content

XBMB3C vs dedicated MB client vs MBT


sfnetwork
 Share

Recommended Posts

First - I really appreciate your positive comments towards XBMB3C - so please don't take any of this as an argument - just discussing :D

 

This is the bit that is really grey to me.  This part of it is very much a skin issue - not an issue with how it gets installed etc.  Do you have some concrete examples?  To me, Eminence is a LOT like some of the MB skins, and obviously Xenon is a lot like Aeon MQ5/Nox.

 

My family has been using XBMC for 10 years, so I always go for XBMC-like skins - even in WMC.

 

xnappo

Oh yeah, just discussing, no offence to anyone here (I really hope MBT devs won't be offended as well by this thread...)

 

Yes, first step, I think, to achieve all this is a new XBMC skin from scratch. I think MBT, in terms of layouts, etc... should be the template. This should be MB specific skin.

After having this skin, first, it would profit XBMB3C then, the "MBT2" project based on an XBMC fork could be started...

 

Again, I wouldn't change anything about XBMB3C addon (I mean besides optimizing it, etc..), Those hardcore XBMC users out there (there are a lot of them) wants to keep pure XBMC experience, stay up to date with XBMC releases but use the addon for MB.

MBT is my "problem", the target here.

 

Basically,

XBMB3C: pure XBMC experience with full MB support

"MBT2": pure MB experience but with XBMC core doing all the "behind the scenes" functions (UI engine, playback, etc...). And easier to get to support OE, different OS, etc... and no dependency with third party applications/codecs, etc... just relying on XBMC core for those kind of things...

***Plus, I think more chances to have more dev coming in the team to help out.

Edited by sfnetwork
Link to comment
Share on other sites

I think the reason that the MBT team will reject that is that the XBMC code base is old as the hills.  It doesn't use any modern programming features and has so much ugly legacy junk in it that just understanding it would be painful let alone improving it.  The Plex guys that split of had many man-years of experience with the code base.

 

Still either way we need an MBT-like skin apparently :)  I think a team should be put together to come up with a common look-and-feel across ALL the MB3 clients.  This is a LOT of work though and I am NOT an artistic type.

 

xnappo

  • Like 1
Link to comment
Share on other sites

I think the reason that the MBT team will reject that is that the XBMC code base is old as the hills.  It doesn't use any modern programming features and has so much ugly legacy junk in it that just understanding it would be painful let alone improving it.  The Plex guys that split of had many man-years of experience with the code base.

 

Still either way we need an MBT-like skin apparently :)  I think a team should be put together to come up with a common look-and-feel across ALL the MB3 clients.  This is a LOT of work though and I am NOT an artistic type.

 

xnappo

Good point... but still, whatever modern programming MBT is using (sorry, don't know much there) isn't enough (or the path to get there is too long...)

IMO, relying on XBMC dev for the core functions and working around it to maintain MBT is a much faster and promising path than handling everything independently. But I'm not the right person to make this argument with my lack knowledge so I just presume based on what I know and test.

 

The fact is that XBMC is the best HTPC software/player out there for a reason so why not using this as MBT core...

I think it's simply a question a ressources... 

 

I really think that this kind of MB client could attract a lot of incredible XBMC devs members.

 

***I know, A LOT of work there (skin, client using forked, etc...) but we never know... this could be the beginning of something  ;).  It's also fun to discuss it so..

  • Like 1
Link to comment
Share on other sites

Well.. I think what he is suggesting is that we pool the MBT team and XBMB3C team - maybe that is something we look into once MBT and XBMB3C projects are fairly complete at 1.0.

 

xnappo

Link to comment
Share on other sites

Ok, I don't see that happening because I can tell you the MBT guys are .net guys who would probably rather stay where they are. If MBT reaches a stage of maturity where they can move on, I think their natural progression would be to contribute to the xbox app development, more in their comfort zone.

 

My opinion, and this is just my opinion, but I really think given how slow Xbmc moves, there is great opportunity to recruit over in xbmc forums. I think people over there like having a project to contribute to. And this could be one that will further the cause for both xbmc and media browser. Obviously it would be easier to recruit if the linux server were released, but even without I still think we can.

Link to comment
Share on other sites

First - I really appreciate your positive comments towards XBMB3C - so please don't take any of this as an argument - just discussing :D

 

This is the bit that is really grey to me.  This part of it is very much a skin issue - not an issue with how it gets installed etc.  Do you have some concrete examples?  To me, Eminence is a LOT like some of the MB skins, and obviously Xenon is a lot like Aeon MQ5/Nox.

 

My family has been using XBMC for 10 years, so I always go for XBMC-like skins - even in WMC.

 

xnappo

 

I think what SF is really talking about is lowering the barrier to entry with one single application that immediately brings you into the media browser experience. Sure, you can put together as many wikis as you want but it still can't complete with something that walks you through the entire process. The lower the barrier to entry is, the more users you can attract.

Link to comment
Share on other sites

On that front, it feel like the timeline for XBMB3C 1.0 and Linux server are lining up fairly well.  We should plan a big push when we get to there.  I am thinking we will have 1.0 late June.  I am going to push 0.9.5 to the official repo this weekend.

 

xnappo

  • Like 1
Link to comment
Share on other sites

Yea the thing about the linux server is that at this point, there isn't really much development required. it's ready to go. creating install packages is where we need help.

  • Like 1
Link to comment
Share on other sites

A fork of XBMC with full MBT integration would be nice but obviously needs more work. However even a simple way to better integrate XBMB3C would be very welcome. XBMC has a robust playback and skinning engine that are well tested which should be reused. If upon installing an MBT addon, users could see a wizrd that would walk them through configuration, and then hide/disable parts of XBMC that are not applicable, then I think we are 99% there.

 

From what little I know the steps involved are simple :-

 

1. getting the server details (and I already see a FR for this based on my earlier post, which is fantastic!)

2. adding video nodes for Collections/Views in MB, and hiding other nodes in the main menu

3. syncing status flags/trakt etc?

 

I think work involved in this is much less than the other proposed solutions like writing a new skin to match MBT. I see a lot of XBMC skins have standardised support for things like Artwork downloader, TvTunes etc. Maybe XBMB3C could also expose properties etc in a way that adding support for it would be easy, and then we wouldn't need custom ports of skins.

 

End goal is for a user to spend the time setting up MB server, on Windows/Linux etc, then install XBMC standalone/OpenElec, and with a few clicks get their favorite skin and XBMB3C addon installed and working with minimal config needed.

Link to comment
Share on other sites

My opinion, and this is just my opinion, but I really think given how slow Xbmc moves, there is great opportunity to recruit over in xbmc forums. I think people over there like having a project to contribute to. And this could be one that will further the cause for both xbmc and media browser. Obviously it would be easier to recruit if the linux server were released, but even without I still think we can.

 

Challenge accepted. Thread here.

  • Like 2
Link to comment
Share on other sites

 

1. getting the server details (and I already see a FR for this based on my earlier post, which is fantastic!)

2. adding video nodes for Collections/Views in MB, and hiding other nodes in the main menu

3. syncing status flags/trakt etc?

 

4. I think work involved in this is much less than the other proposed solutions like writing a new skin to match MBT. I see a lot of XBMC skins have standardised support for things like Artwork downloader, TvTunes etc. Maybe XBMB3C could also expose properties etc in a way that adding support for it would be easy, and then we wouldn't need custom ports of skins.

 

 

 

1. Yes - in progress should be complete this weekend.

2. Most of our skins already do exactly that.

3. Trakt syncing is handled by the server, not sure how to do this.  But the XBMB3C watched, server watched, and Trakt watched all sync

 

4. No, not really possible - but we are implementing in a similar enough way that it is not a ton of work to add MB3 support to a skin and keep it reverse compatible with those other scripts.

 

xnappo

Edited by xnappo
Link to comment
Share on other sites

Yea the thing about the linux server is that at this point, there isn't really much development required. it's ready to go. creating install packages is where we need help.

 

Any volunteers?  

 

xnappo

Link to comment
Share on other sites

 

2. Most of our skins already do exactly that.

xnappo maybe I did something wrong, but I setup a simple MB server with only 1 media library containing videos, and then in XBMC I installed the addon and the custom Aeon MQ5 skin. After siwthcing the mode I still didn't see my video node in main menu and had to add it manually. Perhaps I should've used the Xperience1080++ skin? I saw that its author has stopped development (till Xperience is updated for Gotham) so I hadn't used it.

Link to comment
Share on other sites

Some skins are adding the default nodes (All Movies etc) - I believe Aeon MQ5 does that.  Others add both default and user nodes - I know Confluence does.  Either way you are going to end up modifying the root menu - because there is no way to guess which default nodes and which user nodes a user will want.

 

But all the skins should be operational after turning on MB3 mode, and the XBMC database nodes should be turned off.  Disclaimer: I have not used them all myself :)  

 

At any rate - it is possible - just ask in the skin thread you care about.

 

xnappo

Link to comment
Share on other sites

If you guys can find a skin developer, i can arrange the design resources. No problem there.

 

SF does have a valid point. The htpc apps are at the top of the food chain in terms of supported features, and as a result, they take a long time to build. So whatever man power is required to have our own xbmc build is still probably less than building mbt from scratch. And developers for that are easier to find than WPF developers. The reason why MBT is moving slowly is because as an htpc app, people expect the world out of it. They want the kitchen sink. If we had just focused on the basic ability to play content, it would probably be much more mature right now. But then again, the advantage of building an htpc for it would be lost and at that point, you might as well just go with a roku or android tv box.

 

I think what SF is suggesting here is a great idea, but obviously, I understand completely that we can't make you guys do it if you don't want to do it. So if you don't want to do it (which is fine), then it's a question of what is the next best thing. Maybe some kind of install script that gets you started with xbmc, the addon, and whatever our official skin is. I don't know, just thinking aloud here.

Link to comment
Share on other sites

I created the XBMB3C Confluence skin with a new user experience in mind, it adds all your sections automatically and now with the server discovery I added should need zero setup.

 

So for the client setup we are down to:

- Install XBMC

- Install XBMB3C addon

- Install XBMB3X Confluence skin (or some other skin but may need more setup steps)

 

Ideally a setup tool to install the addon and skin and set the default skin would simplify further.

This is great, thanks !! When you say sections does it mean all the media folders defined in MB server? Does it mean Views?

Link to comment
Share on other sites

 

 

 

Challenge accepted. Thread here.

 

 

Nicely done.

 

xnappo

 

Thanks, man. Lot of views so far, no responses quite yet. Might ping it on the OpenELEC forums while I'm at it. Hopefully we can drum up a bit of interest. There's definitely a market for this kind of thing (there're a few hacked-together solutions already - like this one; but a proper native solution like suggested would fulfill the requirement more elegantly... as would straight XBMB3C, on an appropriately powerful machine!

Edited by kreeturez
Link to comment
Share on other sites

crowclouds

Just stumbled onto Media Browser and it seems pretty impressive, also a long time XBMC user who prefers the client server approach.

 

However I think it very unlikely you will ever get an official version of XBMC with MB support

 

1) Requires interest from core XBMC team.

2) XBMC team have long term plans about a client/server system, a basic UPnP server that shares the library is already built into XBMC, then their are efforts to product a cut down version of XBMC to work as a server along with some automatic UPnP integration on XBMC clients (ala Plex) but neither of these are very far along.

3) Confusion, it's fine if their are plugins but trying to explain why XBMC has this other official build just for MB users, they wont go for that.

 

I think you'd have to go further, the dev team of MB reach out to XBMC team and see if they are interested in MB server becoming XBMC server, obviously both parties would have to be down with that and neither may be but both could benefit from it, MB is doing a lot of things XBMC is planning on the in the future and XBMC devs could help with all the cross platform stuff, not to mention XBMC's huge brand awareness.

 

It would be the optimal solution, you install XBMC and it asks a simple question standalone library (same as XBMC operates now) or client/server mode (looks for MB/XBMC server), no overlap of the media libraries, clean and simple, no special builds required.

 

All the other MB clients like Roku, iOS etc could be re-branded as XBMC Lite clients, there has been talk about doing these too once transcoding is built into the XBMC UPnP server, MB and XBMC are both treading the same path. Rather than both reinventing the wheel you could join forces but for that you'd have to talk to XBMC team directly either by contacting the XBMC foundation or hoping into their IRC channel, no point posting in their devs forums unless you have a proof of concept with a hacked XBMC showing off your vision outlying all the above.

Edited by crowclouds
  • Like 3
Link to comment
Share on other sites

@@crowclouds - I agree - THAT makes sense.  

 

But it will be a tough sell - MB3 is using .NET/Mono - which shouldn't be any different than saying it uses Python - but unfortunately it is a poison pill to some Open Source advocates.

 

xnappo

Edited by xnappo
Link to comment
Share on other sites

@@crowclouds - I agree - THAT makes sense.  

 

But it will be a tough sell - MB3 is using .NET/Mono - which shouldn't be any different than saying it uses Python - but unfortunately it is a poison pill to some Open Source advocates.

 

xnappo

Its funny, ironic and a little sad that the only platform that actually works across all Linux distros is .NET/Mono, and they refuse to use it due to irrational MS hatred.

  • Like 1
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

×
×
  • Create New...