+- XBMC Community Forum (http://forum.xbmc.org)
+-- Forum: Help and Support (/forumdisplay.php?fid=33)
+--- Forum: XBMC General Help and 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)
Linux distros - dailydisco - 2007-05-31 15:32
Based on the news that XBMC is being ported to Linux, I rekindled that flame a few weeks ago and started messing around with various distributions. I tried this on a Dell 610 laptop as well as a Gateway P4 3.0 Ghz desktop. I played with Knoppix (laptop/desktop), Ubuntu 7.04 (laptop/desktop), PCLinuxOS 2007 (laptop/desktop), and OpenSuse LiveDVD (desktop only).
As a relatively computer savvy individual (took some C back in college and some VB afterward for fun, but don't remember much) who has never used Linux, my statistically not-so-significant findings lead me to the conclusion that whichever "plug-n-play" linux distro the developers choose will be helped greatly by some specifications set by them. I can see where the core system (graphics, sound, video capture if that gets included, etc) will need to be tight. Some of the other details do not necessarily need to be set in stone so long as the distro is thought out well. I'll use wi-fi as an example. I was able to get online using the laptop... Ubuntu and PCLinuxOS distros could all see and use the intel wireless chipset without any intervention (didn't try OpenSUSE on the laptop). The desktop was another story.
My biggest issue from my personal experience was that the three distros with the biggest hits on http://www.distrowatch.com (Ubunto, OpenSuse, PCLinuxOS) were not all thought out real well for the end-user. For this reason, PCLinuxOS is now installed on my system. It is the only distro that came with ndiswrapper on the LiveCD so that I could at least connect to the net with my trendnet PCI wireless card using the windows drivers. For whatever reason, Ubuntu left that as a package to install... which is odd since you need to be connected to the net to install it... paradoxical! While this is a little off-topic, it does illustrate the need to make the setup somewhat noob-proof if we would like to really see this takeoff. One word of advice from a Linux noob... please make sure the liveCD/USB includes the tools needed to give the stable platform a chance to survive in the wild (where someone might have everything except wireless card X)! This should be easy since you don't need to include all the other software (OpenOffice, etc) to attract people to the distro. This tends to lean towards some hardware specs... at least for a core solution from this team.
I also agree with one of the other posters... if this system takes off in Linux on a standard setup as well as a couple out-of-the-box solutions (e.g., AppleTV), then this will spawn support sites for those who wish to retool it to work on other Linux distros. Isn't this what makes open source so great? I think the whole point is that the developers of the core XBMC are willing to support the core setup(s) which may change or expand over time, right? This should provide a good starting point for ancillary XBMC-extreme support sites.
One thing to remember, OpenSUSE is owned by Novell, yet I was never able to get online with it...no ndiswrapper setup in the install process (PCLinuxOS was alone in that regard) and the tools they did provide did not see the card... and they have a lot of developers who actually get paid!
- raid517 - 2007-05-31 15:45
@ultrabrutal, as I said I have no objection to the developers coming up with some kind of baseline reference platform.
But I do have a problem with it being tied to yet another proprietary product - such as Apple TV, Mac Mini, PS3 or whatever.
The problem with these I have is that this then takes the element of choice (as a consumer) away from me. I am no longer able to make a choice if I want to spend $400 to $700 or more just to use XBMC (with HDTV support etc). I am genuinely concerned that I may be told that in order to continue to use XBMC I will have to support these companies whether I want to or not.
Nonetheless the advantage of building for a small specific reference platform from off the shelf PC parts are indeed significant.
First if you build a reference platform from standard PC parts, the chances are (despite the developers reservations about offering support for these) that this will also work on 95% of all other recent PC's out there too.
I can completely see where the developers are coming from in this regard, with them wanting perhaps one low end reference platform, maybe a mid range reference platform and perhaps a high end reference platform also. In this sense they want to be like Apple - and work within a known set of given parameters.
But tying XBMC to yet another proprietary platform and pinning XBMC future on this platform's success (or lack thereof) whatever this platform may be, seems to me to be extremely counter productive.
AppleTV, the PS3, Xbox 360 etc may come and go - but the PC is here to stay and is nowhere near so subject to the whims of the market as many of these products may be.
Also as much as I might like XBMC - I am not certain I would want to go out and spend $100's of dollars/pounds or whatever just to use it.
I got my XBoxs' only when the market made this cheap enough to do (I think I got my first Xbox for £80) so if I could simply do the extra work (which if XBMC was built on a standard PC reference platform should be quite trivial) to get XBMC working on one of my spare PC's and not have to spend any money at all in order to get it up and running, I would prefer this much more than any other option.
Also I do think you are being very idealistic and unrealistic. There is no way that enough people are going to rush out and buy an AppleTV or a Mac Mini (which also has an uncertain future) just to install XBMC on it. First again as I said, because of the expense - but also because there just aren't enough people who know about XBMC to make this viable. After all, one of the main attractions for installing XBMC on any PC like platform was that it would be both cheap and upgradable. Not closed or expensive like many proprietary platforms are already. In addition to this why limit your user base in this way, by tying it to a platform that not everyone will be able to have access to? Far more people have access to PC's than to any number of these proprietary platforms combined.
Anyway as I said, I do think a small reference platform (or number of platforms) is a fair enough idea (although other developers seem to worry much less about this). But I really do think tying XBMC to some form of closed/proprietary platform might prove a very big mistake and will prevent many people who would like to use XBMC from doing so, due to both the ethical and financial considerations involved.
GJ
- pike - 2007-05-31 15:46
@dailydisco
aren't you contradicting yourself a bit here ?
if we say we will have target platform Y, then naturally we will have a customized distribution for platform Y
keyword: will, as in eventually
- jonb2 - 2007-05-31 17:03
Ultrabrutal, I don't think anyone but the devs speak for the devs, and they do a fine job of speaking for themselves.
Distro and Question... - KillerDr3w - 2007-05-31 18:14
Hi Pike,
Nice news on the Linux port of XBMC. I am a long time user of XBMC, but haven't really been in the forums or on IRC.
The talk of a XBMC distro rang a few alarm bells
I am sure you know this, but it hasn't specifically been mentioned... building your own distro is as big a project as XBMC is. It might be wise to choose an existing distro, (Ubuntu OR Debian OR OpenSuSE etc) that has all the configuration tools and scripts in already, and then strip it down of apps and add the extra bits needed specifically for XBMC. Then you can concentrate on XBMC rather than hardware detection and configuration, scritps etc.
It would still be "the XBMC Distro" but other than the stripping and customization, you wouldn't have actually done much of the distro development.
Another quick question - does the lack of accelerated 3D on the PS3's hypervisor rule it out as a target platform. I only ask as its a nice looking system with a remote thats already under my TV and can already run Linux if needed.
Thanks - and good luck!
- pike - 2007-05-31 18:46
KillerDr3w Wrote:Hi Pike,
Nice news on the Linux port of XBMC. I am a long time user of XBMC, but haven't really been in the forums or on IRC.
The talk of a XBMC distro rang a few alarm bells
I am sure you know this, but it hasn't specifically been mentioned... building your own distro is as big a project as XBMC is. It might be wise to choose an existing distro, (Ubuntu OR Debian OR OpenSuSE etc) that has all the configuration tools and scripts in already, and then strip it down of apps and add the extra bits needed specifically for XBMC. Then you can concentrate on XBMC rather than hardware detection and configuration, scritps etc.
We will either mod an existing distro or make our own.
The distro will be customized for the hardware in question to eliminate the need for hwprobe and such (to speed up the booting).
With that said, it's way too early to decide stuff like this today!
Quote:Another quick question - does the lack of accelerated 3D on the PS3's hypervisor rule it out as a target platform. I only ask as its a nice looking system with a remote thats already under my TV and can already run Linux if needed.
Like it looks today, yes. Right now we're focused on porting the existing XBMC platform to x86 Linux (regular pc's)
- ultrabrutal - 2007-05-31 18:58
raid, you keep talking about tying? no one is tying anything! we are talking about a common platform for support/development. there is nothing tied that will make it not work on any pc except incompabilities in hardware and drivers on those pc's. appletv and macmini are just pc's. any incompability problems will ofcourse also be fixed by users and/or team members one way or the other - so fear not!
jonb2, well I think Gamester speaks for the devs on many occations
ok guys... come with a list of pc hardware that will suite the xbmc linux pc and let us see what you can come up with to compare with just for fun. I see no suggestions from you guys except bashing
- dailydisco - 2007-05-31 19:41
pike Wrote:@dailydisco
aren't you contradicting yourself a bit here ?
if we say we will have target platform Y, then naturally we will have a customized distribution for platform Y
keyword: will, as in eventually
Pike,
yeah, probably am contradicting myself a little... I guess I was just trying to get at the point of if you try to make a distro that is so utterly locked down (excluding a distro for a boxed platform such as AppleTV) it may be difficult expect that every last piece of hardware will be available. I assume you will lock down graphics, sound, motherboard, etc... there may be a few things which are necessary (ethernet or wireless card) but don't have to be locked down if you just look at how some good linux distros handle things like that (I use PCLinuxOS 2007 as my example since it was the only one designed in a way hat allowed me to get online without having to find another computer that could get online to download some packages). It is indeed way too early to think about that I guess... by the time this goes out as a distribution, the popular distros will be a release or two or three advanced as well.
As for the AppleTV idea, if it goes that route and they can get HD working and the hack is ready for the general public (and Apple doesn't close an exploit holes) I will most likely buy one.
If they don't, and the hardware setup is "in-stone" down to the type of keyboard we can use... what the heck... I'll put one together.
BTW, life if full of contradictions...:o
- bmfrosty - 2007-05-31 21:18
Given all the talk about hardware support, I was thinking that it might be worth talking to these guys:
Mythdora Homepage
When the time gets a little bit closer. They've built a modified Fedora Linux installation that's core feature is that it supports what's needed for MythTV, and not much else. I'd think that modifying that to support XBMC-Linux wouldn't be a huge feat of acrobatics, and given their pre-existing propensity for HTPC distribution, they may not only be capable, but willing as well.
Here is another long one for you all... - Gamester17 - 2007-05-31 22:57
KillerDr3w Wrote:I am sure you know this, but it hasn't specifically been mentioned... building your own distro is as big a project as XBMC is. It might be wise to choose an existing distro, (Ubuntu OR Debian OR OpenSuSE etc) that has all the configuration tools and scripts in already, and then strip it down of apps and add the extra bits needed specifically for XBMC. Then you can concentrate on XBMC rather than hardware detection and configuration, scritps etc.It is way to early to decide anything now, but I personally fully agree with you on this. There is no point in re-inventing the wheel or doing more work than is nessesary to get the job done, so to me the idea of building our own distro from scratch (so to speak) is out of the question. I too think the best thing is to take an existing distro and 'simply' strip away the 'stuff' that we do not need/want and add any extra 'stuff' we do need/want to optimize that forked distro for XBMC and our reference hardware platforms.
It would still be "the XBMC Distro" but other than the stripping and customization, you wouldn't have actually done much of the distro development.
I have no clue which existing Linux distro available today that would best suit all of our needs, and even less of an clue which future Linux distro will be best for our needs in the future when XBMC Linux port is mature enough to be distributed as its own 'Live CD' type Linux distro. However I did read that "Ubuntu Mobile and Embedded Edition" is comming this October and that might we worth taking a closer look at then as the description of what it will be match up well with our own goals for XBMC Linux port. Other Linux Distros that is probably also worth a second look is; KnoppMyth (Knoppix with MythTV), Mythdora (Fedora with MythTV), GeexBox (bootable mplayer with Freevo), and LinuxMCE (their own front'end with multiple back-end players). At least to borrow ideas about which modules/libraries/deamon/script packages to include and which ones to not include.
KillerDr3w Wrote:Another quick question - does the lack of accelerated 3D on the PS3's hypervisor rule it out as a target platform. I only ask as its a nice looking system with a remote thats already under my TV and can already run Linux if needed.That, and the fact that LibSDL has not yet been ported nor optimized for the Cell Processor, (Cell Processor support is also lacking in FFmpeg and MPlayer, and those are the audio/video codec suit and player core that XBMC uses).
jonb2 Wrote:bla, bla, bla....@jonb2, pike and I are the two Project Managers of The XBMC Project. Yes it is true that we do not always speak for all the developers, and we have/will never claim that we to do, however as active contributers to the project we do enjoy the ears of the projects developers which in most cases weighs more than someone that do not contribute at all. As with all open source hobby projects (which XBMC is) everyone working on the project do so for free, thus we can not bark orders at anyone. Eveyone have their own personal agenda for working on XBMC, some do it for fun, others for the challenge or experince, and some (if not most) do it because XBMC is a software that they themself use on a regular basis, (this is the reason why some features in XBMC no longer have an maintaner at all and are completly ignored, and why other existing and new features gain all the attention of some developers). Basicly if you want something done for sure in an open source software project then you have to do it yourself, so those who do nothing to contribute at all will most of the times just be ignored.
In any case, the fact remains that at least Pike and I agree that (at this time) restricting the official support to a couple of specified reference hardware platforms will be the right thing to do for The XBMC Project (once the XBMC Linux is mature enough). This might very well have changed in six-months. Nevertheless, users wanting to use XBMC Linux port on other hardware will have nothing to fear, we are sure that third-party websites and communities will be created that build and support their own Linux Distros with XBMC with their preferences, that is not a bad thing.
PS! What it is that make you jonb2 think that you have all the answers and eveyone should do what you say? If you have not noticed it you do not tend to use frases like "I think that...", "I feel this.." or "My personal opinion is...". Instead you speak as eveything you say is fact and should be blindly obeyed, and I think that is a very annoying trade in a person. Other people have been hanging around in these forums for years and they all managed to use common sense and good netiquette, but you we do not know so why should we respect your options, for all we know you could be a 12-year old shut-in with no social-interaction or real-world experience, who believes trolling and flaming is a cleaver way to get you heared online. I say that all you done so far is spread FUD (Fear, Uncertainty and Doubt).
By the way, what I am going here is called constructive criticism, and though my guess is that you will probably answer me in an immature way and do some more trolling, trying to spread more FUD, but how about if you instead try to take in what others and myself said, then continue this discussion in a mature way, respecting other peoples opinions and try to contribute to the XBMC project in a possitive and groove way? ...if not then this is probably the last time someone replies to you in this forum.