Gauging interest in W10 universal box (xbox mostly)

[PLUCKYHD]

This is in no way a commitment but that being said I have been messing around in W10 universal app development. I switched to Tivo recently as many of you know, however with raising cable rates and Tivo only a short term solution I have began looking elsewhere for the future. OTA and OTT solutions are in my future more than likely. While the Tivo will do this it comes at a monthly or lump sum cost that isn't easy for me to swallow given their recent merger.

So back to the question at hand is there a big User base for a w10 universal app. It would be targeted to small (mini) pcs and xbox one users. It would be a client but not full blown like Stuckless's solution. In otherwords I would stick within the development suggestions of UI for W10 vs using a full blown stv. Java staying 32 bit on windows for the foreseeable future and my no interest in running Linux (sorry I am a MS guy on OS) make a client solution less than ideal for me. I hit the limit many times running heavy UI's on clients (miniclients extenders) before with sage and want to avoid that. I realize these could be accomplished using clients with sagetv but that's not as fun to me Also my Studio is very rusty as I have been away from it for so long.

What it would have

• Full SageDB support
• Use existing fanart from sage/bmt
• Use existing metadata from SageDB
• Allow watch/resume options
• Tv and movie support only (no music or pictures as I have no use case)
• Faster UI rendering than current extenders (in my heavy UI use case at least)

What it would lack mainly besides the UI difference is I don't see commercial skip becoming usable. Unless there I could come up with a solution to mark the commercial times and allow a channel up to skip (much like tivo works) so it wouldn't be hands free at best or won't be viable at all at worst.

Just looking for feedback. When I made the renamer for plex I got most of the backend stuff working in a universal app (reading/writing to the db etc) so at this point it would just take a front end to run it all.

[KarylFStein]

I run Placeshifter on all my W10 computers. It is not always stable with playback, though. I'd be interested in another Windows client, but no commercial skip would be a hard sell...

[stanger89]

Originally Posted by PLUCKYHD

So back to the question at hand is there a big User base for a w10 universal app. It would be targeted to small (mini) pcs and xbox one users. It would be a client but not full blown like Stuckless's solution. In otherwords I would stick within the development suggestions of UI for W10 vs using a full blown stv. Java staying 32 bit on windows for the foreseeable future and my no interest in running Linux (sorry I am a MS guy on OS) make a client solution less than ideal for me.

What's wrong with SageTV Client? There's no benefit to moving it to 64 bit, it doesn't have any sort of JVM issues like the server since it only has to deal with itself. And it doesn't impact the server's JVM in any significant way. But you get the full SageTV UI with all the plugins and the same interface everywhere.

Actually if someone updated Client to support madVR I'd be tempted to move back to a client for my frontend, in my HT at least.

To be clear I'm not talking Placeshifter, no offense to Jeff, but it's crap for a local network. The quality is just not there and it's slow compared to Client.

[Fuzzy]

Originally Posted by PLUCKYHD

This is in no way a commitment but that being said I have been messing around in W10 universal app development. I switched to Tivo recently as many of you know, however with raising cable rates and Tivo only a short term solution I have began looking elsewhere for the future. OTA and OTT solutions are in my future more than likely. While the Tivo will do this it comes at a monthly or lump sum cost that isn't easy for me to swallow given their recent merger.

So back to the question at hand is there a big User base for a w10 universal app. It would be targeted to small (mini) pcs and xbox one users. It would be a client but not full blown like Stuckless's solution. In otherwords I would stick within the development suggestions of UI for W10 vs using a full blown stv. Java staying 32 bit on windows for the foreseeable future and my no interest in running Linux (sorry I am a MS guy on OS) make a client solution less than ideal for me. I hit the limit many times running heavy UI's on clients (miniclients extenders) before with sage and want to avoid that. I realize these could be accomplished using clients with sagetv but that's not as fun to me Also my Studio is very rusty as I have been away from it for so long.

What it would have

• Full SageDB support
• Use existing fanart from sage/bmt
• Use existing metadata from SageDB
• Allow watch/resume options
• Tv and movie support only (no music or pictures as I have no use case)
• Faster UI rendering than current extenders (in my heavy UI use case at least)

What it would lack mainly besides the UI difference is I don't see commercial skip becoming usable. Unless there I could come up with a solution to mark the commercial times and allow a channel up to skip (much like tivo works) so it wouldn't be hands free at best or won't be viable at all at worst.

Just looking for feedback. When I made the renamer for plex I got most of the backend stuff working in a universal app (reading/writing to the db etc) so at this point it would just take a front end to run it all.

Honestly, if it wasn't a SageTV UI, I doubt I'd ever look into using it, even though I've started researching into what it would take to start rolling up an XBox One client myself (haven't found a decent 'here's how you make a hello world app and deploy if to the xbox one' document yet, though). I can already browse and playback my SageTV library via multiple playback apps already available on the XBoxOne - what I need on it is the SageTV UI.

Originally Posted by stanger89

To be clear I'm not talking Placeshifter, no offense to Jeff, but it's crap for a local network. The quality is just not there and it's slow compared to Client.

Placeshifter is not lower quality because of it being a miniclient (meaning UI composition and most interaction going on in the server process). It is lower quality because it has a much lighter weight and more portable video playback system. There's no reason the client's directshow based playback system couldn't also be used in placeshifter, it just was never implemented because the target for placeshifter was remote playback of shows transcoded down to SD or less quality.

The only way I'd be interested in a UniversalApp sagetv client would be if it genuinely was a SageTV Full Client, or a SageTV MiniClient. MiniClient is the way I'd go with it, simply because there's be less to develop for it, and nothing Java based needs to exist on the client end, especially if there is a media playback API that would work with the content streamed from the server. In the past, I would't have recommended a miniclient for local use (due to the JVM Heap limitations), but with the 64-bit server in the wild, that's far less of an issue.

[stanger89]

Originally Posted by Fuzzy

Placeshifter is not lower quality because of it being a miniclient (meaning UI composition and most interaction going on in the server process). It is lower quality because it has a much lighter weight and more portable video playback system. There's no reason the client's directshow based playback system couldn't also be used in placeshifter, it just was never implemented because the target for placeshifter was remote playback of shows transcoded down to SD or less quality.

True, my point was just if all I'd used was Placeshifter, I'd be ready to bag it to, the experience (at least on Windows) just isn't very good. It sounds to me like Plucky hasn't tried Client.

I generally prefer the miniclient model, it simplifies the whole SageTV system deployment, but for now, I think on Windows Client is a better option.

Now if someone were to update Palceshifter/miniclient with a modern playback system (even if it were something like Kodi's, it doesn't have to be Directshow) it would be much more interesting.

[Fuzzy]

Originally Posted by stanger89

True, my point was just if all I'd used was Placeshifter, I'd be ready to bag it to, the experience (at least on Windows) just isn't very good. It sounds to me like Plucky hasn't tried Client.

I generally prefer the miniclient model, it simplifies the whole SageTV system deployment, but for now, I think on Windows Client is a better option.

Now if someone were to update Palceshifter/miniclient with a modern playback system (even if it were something like Kodi's, it doesn't have to be Directshow) it would be much more interesting.

I'm pretty sure he's used Client - I think he is trying to target platforms that client won't necessarily work on (like the the XBox One and non-x86 W10 devices).

[PLUCKYHD]

So much to address

Yes I have used client but honestly I don't prefer it's quality to that of Microsofts built player for w10 universal. It's a preference thing. Also I don't want a computer at each tv Fuzzy was right I would do it more to target xbox one and none x86 w10 devices.

as for the miniclient you say 64 bit server is in the wild isn't that for linux only? Again I revert to my main post I don't run nor do I intend to run Linux. (mostly for lack of knowledge if I don't know it really good I don't want to run it). Miniclient is out of the question for me right now unless a 64bit comes to for wishing on the windows side. I agree it would take a lot less work to get a universal APP for windows based on what I have seen in Stuckless's code. Again if we have a 64bit server for windows coming or close please let me know I am not against this. I am also not for current UI's like gemstone or phoenix it's just not to my taste anymore so that would me more UI work in studio which again I use to know inside and out pretty well but I am rusty at best now.

It's not for everyone I get that but STV UI is also not for everyone It was pretty great when V7 was released but IMHO the UI capabilities and rendering fall behind what something like a W10 Universal App can handle. I think most say they wouldn't use it because of the custom-ability in the UI via and STV and I get that too but I think I can add enough to make it a good compromise for myself.

As for hello world examples there is lots of sample code out there in the wild for Universal Apps its just a matter of downloading them and reading the documentation for them to see how to do some stuff but I do agree it's not readily abundant like I wish it was especially on deploying to xboxone.

[Fuzzy]

Originally Posted by PLUCKYHD

So much to address

Yes I have used client but honestly I don't prefer it's quality to that of Microsofts built player for w10 universal. It's a preference thing. Also I don't want a computer at each tv Fuzzy was right I would do it more to target xbox one and none x86 w10 devices.

as for the miniclient you say 64 bit server is in the wild isn't that for linux only? Again I revert to my main post I don't run nor do I intend to run Linux. (mostly for lack of knowledge if I don't know it really good I don't want to run it). Miniclient is out of the question for me right now unless a 64bit comes to for wishing on the windows side. I agree it would take a lot less work to get a universal APP for windows based on what I have seen in Stuckless's code. Again if we have a 64bit server for windows coming or close please let me know I am not against this. I am also not for current UI's like gemstone or phoenix it's just not to my taste anymore so that would me more UI work in studio which again I use to know inside and out pretty well but I am rusty at best now.

It's not for everyone I get that but STV UI is also not for everyone It was pretty great when V7 was released but IMHO the UI capabilities and rendering fall behind what something like a W10 Universal App can handle. I think most say they wouldn't use it because of the custom-ability in the UI via and STV and I get that too but I think I can add enough to make it a good compromise for myself.

As for hello world examples there is lots of sample code out there in the wild for Universal Apps its just a matter of downloading them and reading the documentation for them to see how to do some stuff but I do agree it's not readily abundant like I wish it was especially on deploying to xboxone.

In my research so far - a 64-bit windows 'server' would not be difficult to develop for SageTV - if it was packaged as JUST a headless server. If we threw out the client portion, and threw out native tuner interfaces (sticking only to network encoders) we could probably have a 64-bit JVM server running without much work (though I don't know that anyone is actively working on it, and I don't even have a build environment at this time). But in short, the actual sage.jar would compile for a 64-bit JVM pretty easily, if the native bits were avoided.

Regarding the Linux 64-bit solution, I was thinking of trying to build a bootable image that could be run in VirtualBox - just wanted to come up with see if a simpler management interface could be made for it (for things like editing .properties files and such).

[Bagal]

I'd be interested for the xbox one especially and I'm of the opinion that the more clients we have the more choice everyone has in how they setup and run their system, which can only be good for the future of Sage.

[stuckless]

The challenge in creating a new client UI is not that hard, and as you noted, you you can use sagex apis, phoenix fanart, etc, and build up a beautiful UI.

The challenge is in the video playback, and to be honest, it's really where the miniclient client is quite impressive. SageTV invented it's own video protocols to completely manage the livetv, seeking, etc, and make it somewhat transparent on the player side.

For the sagetv miniclient player I didn't do anything special to manage livetv, just implemented the sagetv video protocols into the player. After that livetv, seeking, comskip, etc, all just worked with no extra effort.

The problem is that I don't think that sagetv video protocols could be used without creating a miniclient session connection, and even then, not sure the windows player would allow you to create a custom datasource for a video stream.

In the past the web ui, and my other android apps all sucked when it came to playing video, because trying to create a player from the sagetv stream that can support things like seeking and livetv is extremely difficult (not impossible, just really really hard).

That being said, you might be able to take advantage of an unused sagetv features which is a built in httpls server and maybe it will solve some of these video streaming issues.

[Fuzzy]

Originally Posted by stuckless

The problem is that I don't think that sagetv video protocols could be used without creating a miniclient session connection, and even then, not sure the windows player would allow you to create a custom datasource for a video stream.

Actually, I think IMediaSource can be inherited to make a custom datasource to feel content into the mediaplayer.

[Taddeusz]

I think for the Windows ecosystem it would be step in the right direction. At least on the client side I believe Java is holding it back. Java as a language and platform was great 15+ years ago but it's become more of an enterprise server thing now and a security nightmare on the client end. IMHO, Oracle has not done Java any favors.

That being said, I have mixed feelings about the UI. On one hand the SageTV interface is "tried and true", at least for SageTV. On the other hand I also feel that it's also bit tired and uninspiring. Clients should be able to have their own interface apart from the standard.

I believe it would be healthy for SageTV to integrate, not as a plugin as it is now, a full REST api and begin to totally divorce the server side from doing any major UI work. We live in a different world now than we did almost 15 years ago. Client hardware is now more powerful. At times more powerful than the server. I realize this could/would break compatibility with the MiniClient but I foresee SageTV being stagnate by being locked into this weird dual role as media data server/repository as well as UI compositor.

If it were just serving the interface, a la HTTP/HTML/Javascript and let the client do the UI heavy work I wouldn't necessarily have issue with it. However, if I understand how the MiniClient works, the SageTV server actually takes on the role of doing most of the UI heavy work and the client essentially acts as a canvas for the server. To me as a developer this feels counter intuitive and at least in it's current implementation doesn't scale well as can be seen by the need for 64-bit on the server to break free of the memory constraints of 32-bit.

Maybe I'm off base and this might not be the right place to air these thoughts but this topic just brought these thoughts to my mind.

[stuckless]

Originally Posted by Fuzzy

Actually, I think IMediaSource can be inherited to make a custom datasource to feel content into the mediaplayer.

I think this is more of what you'd be implementing... https://msdn.microsoft.com/en-us/lib...essstream.aspx

The IMediaSource appears to a be "source" for creating samples... ie, parsing specific file formats, whereas what you need is a RandomAccessStream implementation that is fed to a MediaSource. ExoPlayer and IJKPlayer have "DataSource" concepts that manage opening "something", reading bytes (not samples), seeking and closing. Not sure if MS has something like that or not.

Originally Posted by Taddeusz

I think for the Windows ecosystem it would be step in the right direction. At least on the client side I believe Java is holding it back. Java as a language and platform was great 15+ years ago but it's become more of an enterprise server thing now and a security nightmare on the client end. IMHO, Oracle has not done Java any favors.

That being said, I have mixed feelings about the UI. On one hand the SageTV interface is "tried and true", at least for SageTV. On the other hand I also feel that it's also bit tired and uninspiring. Clients should be able to have their own interface apart from the standard.

I believe it would be healthy for SageTV to integrate, not as a plugin as it is now, a full REST api and begin to totally divorce the server side from doing any major UI work. We live in a different world now than we did almost 15 years ago. Client hardware is now more powerful. At times more powerful than the server. I realize this could/would break compatibility with the MiniClient but I foresee SageTV being stagnate by being locked into this weird dual role as media data server/repository as well as UI compositor.

If it were just serving the interface, a la HTTP/HTML/Javascript and let the client do the UI heavy work I wouldn't necessarily have issue with it. However, if I understand how the MiniClient works, the SageTV server actually takes on the role of doing most of the UI heavy work and the client essentially acts as a canvas for the server. To me as a developer this feels counter intuitive and at least in it's current implementation doesn't scale well as can be seen by the need for 64-bit on the server to break free of the memory constraints of 32-bit.

Maybe I'm off base and this might not be the right place to air these thoughts but this topic just brought these thoughts to my mind.

Java is still the #1 programming language (and has been for several years). Not sure that is going away, especially with Android being so popular.

I don't see sagetv moving away from Java any time soon... and the 64bit issues have nothing to do with java, but rather the native parts that are written in c/c++ (second and 3rd most popular languages). (in fact on linux, running 32bit, I've never had an issue with needing more memory).

I think the SageTV ui (composed on the server) is the think that makes SageTV able to evolve quickly. Building a complete Media Center UI for a specific platform, covering all aspects, is HUGE... the fact that SageTV can extend to Linux, Windows, Mac, Android by only creating a UI player makes it so that SageTV can quickly extend to new platforms. ie, if someone were willing they could create extend the SageTV ui to the Roku, iOS, (there just doesn't appear to be any interest in those 2).

People could today, do what you say about about creating an HTML ui... It doesn't perform well, in most cases, and none of that would be fixed by replacing sagex rest apis with a set or core apis (of which I'm not the least opposed -- just not sure it's necessary to accomplish any of these goals).

[Narflex]

Originally Posted by stanger89

To be clear I'm not talking Placeshifter, no offense to Jeff, but it's crap for a local network. The quality is just not there and it's slow compared to Client.

Lol...no offense taken...I agree with you. It was never designed for a local network....just for remote playback..hence the name (we only made it work better on LANs after a lot of people starting using it for that; but of course it's MPlayer compared to the DShow player which is far superior on Windows).

[PLUCKYHD]

Taddeusz kind of hit the nail on the head with the why's. I realize some don't hit a wall but rendering on the server is less than ideal to me. It works but has it's limits on capabilities and speed. I realize some never hit that 32 bit wall but I did easily with graphic intense UI's. It also isn't so much the wall of 32 bit but the speed of rendering which is much slower over the miniclient then directly rendered at the server (ie client vs hd300).

But we all know I go overboard with the "flash" of UI's as well but it's my liking on a interface.

I agree the benefit of feeding the player is great for commercial skip and other things but I think there is still ways of getting around this to some extent but that's all theory until someone or I actually try it.

Where I am at now is the play around a bit phase and trying to learn what MS wants for universal apps and what it's limits are. Seeing how well Plex natively handles almost anything on the xbox one I have to assume the default player is more than capable. Unless they are rolling they own which I highly doubt. But right now it's still heavily concept and the UI rendering are trival for the most part as stuckless said playback will be the deciding factor for me.

Personally I love plex for movies and Love sagetv for OTA DVR content. they are the best to backends available IMHO. It's just how to peacefully merge the two. Tivo does a decent job but Plex still doesn't integrate with OnePass so it's still fragmented at that end. But the integration with neftlix,vudu, and amazon prime is well done on Tivo. I need to dig and see if that's an open API to search amazon/netflix and integrate with third parties like that. Even if it has to launch the netflix player once a show is selected integration is very nice.

[stanger89]

Originally Posted by Narflex

Lol...no offense taken...I agree with you. It was never designed for a local network....just for remote playback..hence the name (we only made it work better on LANs after a lot of people starting using it for that; but of course it's MPlayer compared to the DShow player which is far superior on Windows).

Actually I was playing with both yesterday, same machine dual monitors, Placeshifter is better than I remember, but there's a definite deinterlacing advantage to Dshow over Mplayer, especially Mplayer from 10 years ago

Originally Posted by PLUCKYHD

Taddeusz kind of hit the nail on the head with the why's. I realize some don't hit a wall but rendering on the server is less than ideal to me. It works but has it's limits on capabilities and speed. I realize some never hit that 32 bit wall but I did easily with graphic intense UI's. It also isn't so much the wall of 32 bit but the speed of rendering which is much slower over the miniclient then directly rendered at the server (ie client vs hd300).

I think that has more to do with the HD300 than the miniclient architecture. Playing with Client and Placeshifter side by side on the same machine, Placeshifter is basically as quick, possibly not quite as smooth in the animations.

I need to dig and see if that's an open API to search amazon/netflix and integrate with third parties like that. Even if it has to launch the netflix player once a show is selected integration is very nice.

Didn't Netflix cut off API access a few years ago?

[KryptoNyte]

Originally Posted by Fuzzy

In my research so far - a 64-bit windows 'server' would not be difficult to develop for SageTV - if it was packaged as JUST a headless server. If we threw out the client portion, and threw out native tuner interfaces (sticking only to network encoders) we could probably have a 64-bit JVM server running without much work (though I don't know that anyone is actively working on it, and I don't even have a build environment at this time). But in short, the actual sage.jar would compile for a 64-bit JVM pretty easily, if the native bits were avoided.

I know this wasn't intended to be a server thread, but to have a headless 64 bit, web based control server would be ... bliss.

[Fuzzy]

Originally Posted by KryptoNyte

I know this wasn't intended to be a server thread, but to have a headless 64 bit, web based control server would be ... bliss.

That's pretty much what the stuckless' docker container already available is.

[PLUCKYHD]

Originally Posted by stanger89

I think that has more to do with the HD300 than the miniclient architecture. Playing with Client and Placeshifter side by side on the same machine, Placeshifter is basically as quick, possibly not quite as smooth in the animations.



Didn't Netflix cut off API access a few years ago?

it has some to do with limit of ram when you are running several UI intense at once I many times capped out the heap In the past but also the hd300 does limit the speed and smoothness for sure. But it's an old processor no Adays.

i believe they killed their play API but not their search API I would be wrong I haven't looked

[KryptoNyte]

Originally Posted by Fuzzy

That's pretty much what the stuckless' docker container already available is.

That docker was for Windows 64 bit? Where is that?