XBMC Linux port questions and answers... - Printable Version
+- Kodi Community Forum (http://forum.kodi.tv)
+-- Forum: Support (/forumdisplay.php?fid=33)
+--- Forum: General Support (/forumdisplay.php?fid=111)
+---- Forum: Linux and Live support (/forumdisplay.php?fid=52)
+---- Thread: XBMC Linux port questions and answers... (/showthread.php?tid=26097)
- szsori - 2007-06-01 16:58
If "Joe Blo" had an issue like that, I'm sure he would get proper advice from a number of XBMC users. He just couldn't expect support from the devs themselves.
He may also be told that his best option for solving his problem will be to ask on "Linux Forum X" (whatever) if that's the case. Even now people are told to post on other forums if the issue is unrelated to XBMC (see TV show scrapers). Sometimes telling a person where to go for help is the best thing you can do.
We're just beating a dead horse now, though. On to better issues like TV card support, running other programs via shell, and a better web interface (now that it could run apache).
- bmfrosty - 2007-06-01 19:03
szsori Wrote:On to better issues like TV card support, running other programs via shell, and a better web interface (now that it could run apache).
Direct TV card support is a bit of a monster to support, and is probably way out of the scope of this project, but that doesn't mean a MythTV frontend isn't. An n-curses remote control would be drop dead sexy, and Apache is overkill for XBMC. A nice, single-purpose, light-weight web server is much better suited.
- szsori - 2007-06-01 19:44
Integrating TV recording and live viewing seems like it wouldn't be that hard in linux. Just run a program in the background like the MythTV backend or NVrec for recording and have XBMC create the entries in the recording schedule file. I'm sure there's even better programs than those to do straightforward recording.
For live TV, you just need to run VLS in the background and have XBMC grab the stream. Changing channels could be done by having XBMC pass through those commands to VLS in the background. That wouldn't give DVR capabilities like pausing live TV, but it'd be a decent stopgap until something better could be worked in.
webserver - Gamester17 - 2007-06-02 09:14
bmfrosty Wrote:Apache is overkill for XBMC. A nice, single-purpose, light-weight web server is much better suited.I think that you are probably right on this, since XBMC will be a cross-platform software using the very same code base (and not seperate source code trees), best would be be if we also used a small cross-platform webserver and a cross-platfrom web-site running on that webserver. That way even if you run XBMC nativly on an Xbox, or on a x86 Linux-computer, or a future new game-console port, both the webserver and the website for remote-control would still look and function the same (as they share the very same code on all platforms).
XBMC uses GoAhead WebServer as its current code for this, and it was the best webserver for Xbox version of XBMC at the time it was implemented. It is very small, designed for embedded systems, and is cross-platform already. Unfortunatly GoAhead Software Inc. who maintain the original branch of the GoAhead WebServer seems to have stopped development of the source code in December 2003, (and I at least can not seem to find any projects with forked version of the GoAhead WebServer on which development have continued since then, can anyone of you?).
Also nad have extended the HTTP-server support with his HTTP API which enables many more ways to control XBMC via HTTP:
PS! The webserver in XBMC is not meant to serve as a standard webserver, end-user are not suppose to use it to host their private website on, it is only ever meant to be a remote-control-interface for XBMC (and maybe at most also a way to access your local media-files remotly)
- gfmc - 2007-06-02 13:53
As a long-time XBMC user, I'm fine with official hardware support being limited to one or a few platforms. Really all I want is an XBMC box that can decode HD video streams, I don't care about Linux related politics or anything like that. I've tried all of the other PC based (Windows and Linux) media player front-ends and none of them are anywhere near as slick as XBMC in terms of great UI, number of formats supported, etc.
I'll be glad to get rid of the split I currently have between SD and "HR" HD content and full HD content, which I currently view via the Xbox 360's MCE functionality (which I find to be far inferior to XBMC). If this means I have to buy some specific supported platform for a few hundred dollars, I'm ok with that.
- szsori - 2007-06-02 20:16
With regards to the webserver, I think one that can at least handle php or perl would be ideal, since it would open up all sorts of neat plugin options.
For example, if XBMC-Linux does work as a frontend for a TV recording program, people might find it easier to schedule those programs in a web interface. I'm sure there would be scripts that could benefit from web configuration as well. Most of all, it would be REALLY cool to have a web interface to the library data, where people could manually edit entries if they needed to.
Finally, an "advanced" section could allow people to make changes to their AdvancedSettings.xml without ever having to touch the file itself. It would make things much easier on users without a lot of technical experience.
I'm not saying all of these things would come to fruition, and I'm not saying that Apache would be the way to go, but I think that if a more powerful web server is used it would give web developers like myself an easier way to contribute.
- raid517 - 2007-06-03 01:25
I haven't said 'support any hardware' at all - and you should know that this is a deliberate distortion of what I did say.
What I have said is that if you are going to support a limited set of hardware, it might be a good idea to choose that hardware from generic off the shelf parts, rather than tying it to any kind of proprietary platform.
Sure you could shoehorn it to run on Apple TV, or the PS3 or whatever, but despite what you have said these do have really quite a specific set of hardware components. Too specific in the sense that it does add more work to the task of then getting XBMC to run on more generic platforms.
If you choose off the shelf PC parts, there are a large number of advantages to this.
First you are not tied to the fortunes of a proprietary platform or company. If AppleTV is a flop, then what will this mean for XBMC? We will surely all at that point end up back at square 1.
Secondly by choosing off the shelf PC parts, this gives you the ability to respond quickly to new changes within the technology sector than would be possible with a proprietary platform. For example say the next new holohraphic disk drives came out in the next two years, or say a new kind of display technology was released? The PS3 and and Apple TV and other similarly proprietary platforms are not at all well placed to respond to these changes. So we might end up back in a few years at a point where we have everyone complaining (like they have about HDTV) that this or that technology isn't supported. While having a few reference designs may be a good idea, building these designs with off the shelf PC parts allows you to improve on these reference designs as and when new technologies emerge. (Just as Apple does, since while they do have a limited set of supported hardware, when new technologies that are desirable do emerge, since they too use fairly generic parts, it allows them the ability to respond quickly and efficiently to these changes).
Thirdly, if you do use generic PC parts and build a few specific reference designs, it makes it much easier to run XBMC on other similar platforms also - which might not exactly match the specifications of the reference designs. This is because in most cases, anything that runs on any relatively modern PC, will often run on 95% of all other PC's also, without very much issues. So it won't matter in most cases that it isn't supported, since 9 times out of ten it will probably work perfectly well anyway. This doesn't mean that you can't run XBMC on Apple TV, or the PS3 or whatever other platform you might like, only that this then becomes the task of the hackers who always do this kind of thing anyway. There is no reason if it runs on (a limited number of) generic PC parts that it can't be ported to AppleTV, or the PS3 or whatever you want - all it means is that you might have to go to a little extra effort to do this. (Which as I said is fair enough, as this always has been the job of the enthusiast hackers anyway to do this sort of thing).
From your perspective you seem to think the best way would be to develop for another proprietary platform - to only build XBMC (or at least the officially supported version) on Apple TV or the Sony PlayStation and then anyone who wanted to run XBMC on standard PC parts would have to do some extra work to get it running on their PC's. But I think this is the wrong way round - because this really badly limits the amount of people who will be able to use XBMC (with official support) from the outset - pretty much for all of the reasons I stated above.
PC's will never go away, but Apple TV and Sony PlayStation etc will probably come and go. PC's are also much cheaper in many cases than proprietary platforms (so you can keep costs down for users) and are much better able to respond to changes in technology. The number of people who also have access to PC's is also vastly more than all of the other possible proprietary platforms combined - so you instantly have the potential to vastly increase your user base.
These are of course all just suggestions, but as I said I think that the advantages for using off the shelf PC parts in the selected reference designs far outweigh any advantages of simply selecting and sticking to (yet another) single (or perhaps more than one) proprietary platform really probably ever could.
I don't think there is really such a good case for tying your hands - and at least if you use off the shelf PC parts, you won't really ever need to do this.
As I said, this is just my thoughts on this though. The ultimate choice will still unarguably be down to the developers themselves.
- raid517 - 2007-06-03 01:36
Sorry just a small adjunct since I can't seem to edit. But one example of what I'm saying is that what would happen if (for example) you wanted HDDVD and BlueRay disk support in XBMC? (I know these are officially supported in Linux yet, but it is likely this will be overcome, just as the issue of DVD playback was overcome). This would be virtually impossible to achieve on a proprietary platform (dose Apple TV even have a BluRay or a HDDVD drive?) but perfectly conceivable and achievable on a PC platform. This technology almost always seems to reach the PC first, before it reaches any other platforms.
The same is also true of any other new technology that might emerge. These proprietary platforms are simply not well placed at all in being able to adapt to these changes.
- ultrabrutal - 2007-06-03 01:47
off the shelf computer parts that are around in 5 years, heck say 5 months... does not exists... im not saying that xbmc should not support any hardware if it was possible, just that the common support platform must exists for more than a few months like the xbmc has... or do you want all xbmc users to keep switching hardware every few months? or support more and more hardware over time? I cannot see the logic. can't have it all. this is not a big company with 100's of devs you know
Keep in mind that it's currently being developed on regular pc's
No some people will make generic XBMC distros's that run on any PC (that are compatible of course) - these will just not be supported directly
- ultrabrutal - 2007-06-03 01:49
applytv has no drive no. but macmini has. there is a new version coming later this year. lets see what it has... anyways it might be exchangeable?
- bmfrosty - 2007-06-03 02:14
You don't get it. The AppleTV, for all the purposes that matter, *IS* a standard PC. It's big advantage is that it's cheap, and the different revisions that will exist over time, should be limited in number. Otherwise, it's slightly underpowered for H.264@720p, but if NVIDIA manages to release purevideo drivers for Linux anytime soon, it will be perfectly suited.
The other nice thing is, the AppleTV is *CHEAP* compared to just about any other solution because it is built as a single purpose media center from the start. No more, no less.
If another company, any company were to build something like this, at the same price, with the same expected product lifetime, and the same or better power, it would also make a good choice. At the moment, there isn't. The only other companies that I'd expect to do this type of thing are Sony and Microsoft. Microsoft may put out something useful, but you can bet that the potential Sony version would probably be as gimped as the PS3.
- szsori - 2007-06-03 07:49
Can someone rename this thread to "The beating of a dead horse" already?
I have an idea that might suit everyone, though. It's been mentioned that a few supported platforms from different price ranges might be selected as "officially supported platforms". Why not select one set of "official" PC components to support as well?
For example, you could support the AppleTV as your low-end platform and the Mac Mini as your mid-level platform. Then choose a set of PC hardware to support as your high-level platform. This could either be a system made by a PC manufacturer that doesn't change their hardware very often (http://us.shuttle.com/Home_MCE.aspx) or parts that users could assemble themselves.
By using the latest "basic" parts from well established manufacturers, it would be possible to come up with a standardized system that could be used for at least a few years. The most important of these is the video card, since it's going to be the pickiest component with the shortest life. The other stuff is easy, though... a motherboard from ASUS with on-board optical audio, a SATA hard drive, and a video capture device from Haupauge (if needed at some point). Including the drivers necessary for Bluray and HD-DVD (as well as DVD+-etc, CD, etc devices) shouldn't be that hard, since a handful of drivers cover all of them and rarely have compatibility issues.
- paperclipmonkey - 2007-06-03 15:03
I think it'd be a better idea to give the linux port it's own sub-forum so that all of the different ideas can be given threads of their own. So support/ideas/discussions can be held separately instead of being held in one massive thread.
Just an idea
- bmfrosty - 2007-06-04 02:06
Here's a manufacturer that may be worth talking to once the project is closer to it's second milestone:
They're currently building something that's designed to be a mythtv front end.
- Livin - 2007-06-04 04:33
This is a heated subject so I do not mean to put fuel on the flame... so if you are excitable, don't read further, skip this post.
To those pushing for an 'open' platform...
Members pushing for a 'closed' hardware platform like AppleTV, PS3, etc don't get it and they may not ever. (not trying to be mean, this just seems clear at this point). Most I get the feeling (yes, feeling - my opinion) do not know or understand the limitation of a 'closed' platform and do not understand that it will eventually run into the same issue we are facing now with the original XBOX.
Yes, 'closed' platforms are easier to develop/support for/on but are limited in longevity and ultimately usefulness... as we are seeing today.
To those making the decision...
What I and others here are pushing for is not a single/multiple platform standard but "reference hardware" ... like ... ATI and NVIDIA (the defacto standards in the industry) video, Intel chipset P945 and P965, etc. This is standard industry practice for decades now.
What this means is that you can use the 80/20 rule...
by officially supporting specific reference hardware and drivers (e.g. version XX of the ATI driver on chipset XYZ) then you allow a wider range of reasonable supportability for the majority of the users.
Anyway... I'm done explaining the point. To those who understand hardware and software support (like myself, 15 years professionally for Fortune 100 companies) this will make sense. For those that don't get it or just want to support the "easiest" and least effort, they will push AppleTV or the like.
That is my $0.02... coming from a simple dude that has worked for the largest software company in the world and the 2nd largest hardware company (the one that now supports Ubunto) -- Take it as you will.