Quote:The Linux NVidia binary drivers are not documented to support anything but MPEG2 offloading, so PureVideo is currently unsupported for Linux.

nate12o6 Wrote:I hate the fact that we are debating hardware specs this early so this is not a request for a port to the ps3 but only a question about the cell processor. Does anybody know how good the cell processor could decode or even encode video? It seams like it would be built for that. But i also here horrer stories about how hard it is to develope for.

nate12o6 Wrote:question #2

I know the xbmc team has talked with microsoft about getting a dev liscence and where denied. Would you say the chances with sony would also be 0?

szsori Wrote:I have a HTPC that I use in a different room instead of my Xbox. It actually runs quieter than the Xbox. Not saying it's silent, but it's far more easy to customize a PC to make it silent than customizing a console or STB.


A computer that's just sufficient to decode h264 content can be found for far less than $1000. Shuttle has really high end stuff for HTPC applications for $1000 that looks really nice, but I personally built mine in a nice Coolermaster case for under $700.


Once again, not all of them.

I do like the Apple TV device and think it looks pretty cool, but you're extremely limited compared to a PC or Mac Mini. The thing that the Mac Mini/Apple TV fans keep forgetting is that there are plenty of sufficient PC options as well that run considerably cheaper. They look as nice and are more flexible as far as which hardware you can use with them. I'm not saying that a PC is definitely the way to go... I'm just saying that you folks need to be more accurate with your comparisons and "facts".

I still think the way to go would be to have one of the XBMC devs talk to Shuttle (or another reputable company) and ask if they can develop a device with the exact specifications they want. Any company would be likely to jump on the opportunity since they'd be guaranteed to sell at least a few hundred of the computers just to the XBMC crowd. If the number of buyers was predetermined, perhaps we'd also be able to get additional discounts.

Gamester17 Wrote:I have been thinking about this during the last couple of days. I now like to argue that besides the 'good old' Xbox (and possible future ports to PS3 and Xbox 360 which I do not think we should discuss in this specific topic-thread), when the Linux port of XBMC becomes mature enough for end-users to fully use it will probebely be a very good idea to keep two (and only two) specific 'locked-down' x86-based computer platforms (or motherboard) with different specifications and differently priced as the 'locked-down' concrete reference hardware and the sole computer platform that Team-XBMC will officialy support. Smart would probebely be to have one relativly inexpensive ($500 USD or less) low-end computer hardware model and one expensive ($1500 USD or less) high-end computer hardware model, both with clearly specified minimum requirements and exactly detailed hardware parts, (that way the users can let their needs and size of their pockets decide which model to go for).

I think that low-end model should be fanless and at least capable to playback H.264 encoded video at 720p native resolution but also capable to upscale all video to 1080i, and have at least have optical-audio, component and DVI or HDMI ports. The high-end model should be capable to playback H.264 encoded video at 1080p native resolution and also capable to upscale all video to 1080p, and have at least have optical-audio, component and HDMI/HDCP ports. The low-end computer hardware model should not be required to have a DVD-ROM drive, while the high-end model on the other hand should be required required to have a DVD-ROM drive. Niether the low-end or the high-end computer hardware should nieter be required to required to have a harddisk-drive, but both should be required to be able to boot from a USB-key and completly run the operating-system and XBMC (plus any other applications) from there. The hardware's lifle-cycle should also be as long as possible, meaning new ones should be should in retail stores for years to come, (that is why a game-console hardware makes a perfect sense to use as a reference model).

Today Apple TV (from Apple) already fills all of those given requirements for the low-end computer hardware model, and on top of that it only costs $300 USD. I do think that would make the perfect reference platform, (the only 'issue' I see is that it does not have a LCD-display on the box itself but that could always be case modded if one wants it). The minimum requirements I given above for the high-end computer hardware model is however harder to meet today (especially the long life-cycle part), so I do not think that model should be set in stone just yet. I do however think that the upcoming update model of the Mac Mini from Apple will possibly make a great high-end model candidate, I think it will probebly come down to the retail price that will be set for it when it is released this fall, (and if it so happens to come with a HD DVD and/or Blu-ray disc-reader as an option that would not hurt either), ...yes I know that the PlayStation 3 (PS3) premium-edition game-console hardware also fills those requirements, but it might automaticly have to be ruled out because of the combining issues like it is not an full open platform; with non open source device-drivers and the PS3 hypervisor is limiting Linux on it to 2D graphics and low-level access to the hardware, plus the fact that processor (CPU) is not x86-based but instead PowerPC-based Cell processor which could make development applications for it harder and more time-consuming.



Know though that no matter what platform we choose as the locked-down computer hardware, since this XBMC Linux port will run as an application end-users and developers a like will of course be able to run it on any computer hardware they like, they will just not get any support with problems from Team-XBMC, nor will we fix bugs on other that only occur on the computer hardware other than the two specified locked-down reference models. If this XBMC Linux port will be distributed as an official Live CD (LiveDistro) bundled with the Linux operating-system to futher lock-down of the static constant system enviroment then only that Live CD distro will be officialy supported by Team-XBMC, and it will thus stripped down as much as possible to keep the size and potential compatibily issues down (by ripping out device drivers, etc. that will not be needed unless we would intend to support multiple hardware platforms).

PS! I'm have now moved this topic-thread from the XBMC Development Forum to the XBMC to the XBMC General Discussion Forum since the discussion har clearly spun away from code and programming, also only developers should really start new topic-threads in the developement. So feel free to continue this broad general questions/answers and discussion about the different aspects of what porting XBMC to Linux could/will bring.