Arm9 sbc
I never heard of them. They give you some documentation on DVD-ROM, but what's the quality of the text? They're also shipping from China. What happens if you have a problem? Who can you call or email? Don't get me wrong, the board has nice specs. It's your money so it's your call.
Take a look at these folks: Technologic Systems PC/104 Single Board Computers and Peripherals All docs and data sheets are available. There's also an active mailing list.
I want to order this board in the near future: TS-7550 Ultra-Compact 250MHz ARM9 Computer with XNAND
Disclaimer: I have no affiliation with TS. I just think they have interesting boards that are economical for the hobbyist.
I never heard of them. They give you some documentation on DVD-ROM, but what's the quality of the text? They're also shipping from China. What happens if you have a problem? Who can you call or email? Don't get me wrong, the board has nice specs. It's your money so it's your call.
Take a look at these folks: Technologic Systems PC/104 Single Board Computers and Peripherals All docs and data sheets are available. There's also an active mailing list.
I want to order this board in the near future: TS-7550 Ultra-Compact 250MHz ARM9 Computer with XNAND
Disclaimer: I have no affiliation with TS. I just think they have interesting boards that are economical for the hobbyist.
nyc_paramedic,
the only COM I have been able to find without any on-board SMPS is probably the old Asus WL500b http://files.wl500g.info/asus/wl500g/img/rev2.40/HPIM1736.JPG . I happen to have one at home so I will check. My current audio project (apart of all the other non-audio ones with higher priority) is to build a miniPCI-PCI converter to hook a regular PCI sound card (e.g. Juli) to this little guy, while all the power supply pins would be fed from dedicated linear PSUs. Unfortunately the half-finished converter is being covered by dust in my workshop for many many months.
the only COM I have been able to find without any on-board SMPS is probably the old Asus WL500b http://files.wl500g.info/asus/wl500g/img/rev2.40/HPIM1736.JPG . I happen to have one at home so I will check. My current audio project (apart of all the other non-audio ones with higher priority) is to build a miniPCI-PCI converter to hook a regular PCI sound card (e.g. Juli) to this little guy, while all the power supply pins would be fed from dedicated linear PSUs. Unfortunately the half-finished converter is being covered by dust in my workshop for many many months.
Well, you are right, a headless DAC unit with wireless remote control using the already existing sw components makes more sense.
See. You're talking about a SB Touch without monitor in principle.
The basic Touch concept is IMO as good as it can get. ( It is even running a Linux. )
The processor spits out an I2S stack ( see also Beagleboard). That one even gets pretty well reclocked afterwards.
If you bypass the output caps of the internal DAC, you'll be able to experience a pretty good sounding device.
I think for many people this will be a real nice and easy entry into the streaming world.
I really think if anybody wants to start a DIY ethernet streaming project, the Touch shows the way to go. Why bother with USB and all its negative side-effects again if I have an I2S bus at hand!?!?
If I'd be looking for a nice little PC+USB out, I'd still opt for FitPc ( I stlll have one up'n running) or if I want to go very tiny I'd go for Guru Plug ( 100$). For both of them you'll find Ubuntu Linuxes.
There's IMO no need for those awful micro-Linuxes a la Voyage. Just change
the Ubuntu startup scripts and you're done.
And if you're looking for Touch related Linux challenges -- you'll find plenty of DIY potential.
Cheers
alix/mpd/voyage wins Fight Club against modified SB Touch
For DIY or for the non-techie masses?
You are contradicting yourself. For a DIY ethernet streaming project, the Touch requires a lot of work. Even on this DIY site, not "many" people have the know how or the basic tools to start stripping down a SB Touch.
A much better DIY ethernet streaming project is using an Alix board with Voyage Linux and MPD.
And what's so bad about USB audio? We have a bunch of new DAC supporting Asynchrous mode; more being release every month. The Ayre QB-9 has high speed opto-isolators between the DAC board and the USB board, effectively isolating the computer from the DAC. It could be argued that the SB Touch DAC and analog section are being polluted but the high frequency ARM CPU. With a USB DAC, you can create some distance between your CPU and your DAC.
Also, your SB touch is...
limited to 24/96
DAC/analog section can't be separated from the CPU, like a USB DAC can
subject to the whims of Logitech who can issue firmware updates in the future, thus locking you out from further DIY hacks.
has been known to have very slow and buggy software
has a useless touch screen when the unit is placed at the hi-fi rig and one sits away from the speakers.
Voyage is awful? And you recommend Ubuntu for a player? You are comparing a full GUI desktop OS to a version of Debian that runs off a 128MB compact flash card and specifically designed for embedded devices? I know that English is not your native language, but you are quite opinionated *and* you really don't know what you are talking about.
Some people have a little DIY spirit, but don't have thousands of hours to waste on questionable mods. They just wanna hear a little music. I can assure you that many more people have experimented with an alix/voyage/mpd combo than anything else. The bar is lower, the quality is better, and you don't have to sit at an electronics bench for hours on end.
Cheers!
See. You're talking about a SB Touch without monitor in principle.
The basic Touch concept is IMO as good as it can get. ( It is even running a Linux. )
For DIY or for the non-techie masses?
You are contradicting yourself. For a DIY ethernet streaming project, the Touch requires a lot of work. Even on this DIY site, not "many" people have the know how or the basic tools to start stripping down a SB Touch.
A much better DIY ethernet streaming project is using an Alix board with Voyage Linux and MPD.
And what's so bad about USB audio? We have a bunch of new DAC supporting Asynchrous mode; more being release every month. The Ayre QB-9 has high speed opto-isolators between the DAC board and the USB board, effectively isolating the computer from the DAC. It could be argued that the SB Touch DAC and analog section are being polluted but the high frequency ARM CPU. With a USB DAC, you can create some distance between your CPU and your DAC.
Also, your SB touch is...
limited to 24/96
DAC/analog section can't be separated from the CPU, like a USB DAC can
subject to the whims of Logitech who can issue firmware updates in the future, thus locking you out from further DIY hacks.
has been known to have very slow and buggy software
has a useless touch screen when the unit is placed at the hi-fi rig and one sits away from the speakers.
Voyage is awful? And you recommend Ubuntu for a player? You are comparing a full GUI desktop OS to a version of Debian that runs off a 128MB compact flash card and specifically designed for embedded devices? I know that English is not your native language, but you are quite opinionated *and* you really don't know what you are talking about.
Some people have a little DIY spirit, but don't have thousands of hours to waste on questionable mods. They just wanna hear a little music. I can assure you that many more people have experimented with an alix/voyage/mpd combo than anything else. The bar is lower, the quality is better, and you don't have to sit at an electronics bench for hours on end.
Cheers!
Last edited:
phof,
Have you seen this? Link: IM300 (Mini PCI Type III to PCI Adapter Card) : Interface Masters, Inc., Innovative Network Solutions
Web page here: IM300 (Mini PCI Type III to PCI Adapter Card) [IM300] - $88.00 : Interface Masters, Inc., Innovative Network Solutions
It's only $80 USD.
nyc_paramedic,
a few years ago I emailed those guys about availability of their card in central Europe in single quantities and they did not even respond. I would like to use a passive solution. There is a chinese seller of passive version but those guys did not respond either. I do not think it was due to my english
But your post may make me reconsider my project and I may try to purchase the ready-made solution again. The DIY version is extremely labour-intensive. I had to design the insert card, had it produced, soldered IDE 80pin flat cable, removed PCI slots from an old MB with a heat gun, and I still have around 30 wires to solder out of those 100 or so PCI slot pins (many pins are wired together). And when finished, it will very probably not work due to some soldering bug or major design flaw
a few years ago I emailed those guys about availability of their card in central Europe in single quantities and they did not even respond. I would like to use a passive solution. There is a chinese seller of passive version but those guys did not respond either. I do not think it was due to my english
But your post may make me reconsider my project and I may try to purchase the ready-made solution again. The DIY version is extremely labour-intensive. I had to design the insert card, had it produced, soldered IDE 80pin flat cable, removed PCI slots from an old MB with a heat gun, and I still have around 30 wires to solder out of those 100 or so PCI slot pins (many pins are wired together). And when finished, it will very probably not work due to some soldering bug or major design flaw
Soekris
Soekris should have their new net6501 board out in a few months: Products
net6501, a faster and more advanced mainboard, up to 1.66 Ghz CPU, 2 Gbyte DRAM, 4 Gigabit Ethernet ports and PCI Express expansion, production availability in Q4 '10
And this from Soekris list:
Soren Kristensen soren at soekris.com
Tue Dec 15 20:19:31 UTC 2009
* Previous message: [Soekris] New models?
* Next message: [Soekris] New models?
* Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hi Everybody,
Maybe it's time to tell a little more.... The net6501 is moving forward
and I do expect to have hardware ready in Q1 2010, although I am known
to be an optimist
The net6501 will basically be like the other boards, just faster, with
more memory and PCI Express expansion.
There will be both a 2 and 4 port gigabit ethernet version, using Intel
controllers, where the 2 ports version will be targeting small servers,
with up to 4 of them in a 1U case....
And yes, it will be based on the next generation Intel Atom processor,
the Pineview platform, in both single and dual core versions and with up
to 2 Gbyte DDR2-SDRAM soldered on.
And it will still be low power and high reliability, with passive
cooling. Ok, a tiny server with two 2.5" 10K rpm SATA drives will need a
small fan....
A new goodie will be onboard NiMH batteri charger/controller.
Best Regards,
Soren Kristensen
CEO & Chief Engineer
Soekris Engineering, Inc.
Nice, eh?
Could make a nice experimenters board. You could compare and see if there is any difference between alix/ethernet and soekris/ssd.
Soekris should have their new net6501 board out in a few months: Products
net6501, a faster and more advanced mainboard, up to 1.66 Ghz CPU, 2 Gbyte DRAM, 4 Gigabit Ethernet ports and PCI Express expansion, production availability in Q4 '10
And this from Soekris list:
Soren Kristensen soren at soekris.com
Tue Dec 15 20:19:31 UTC 2009
* Previous message: [Soekris] New models?
* Next message: [Soekris] New models?
* Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hi Everybody,
Maybe it's time to tell a little more.... The net6501 is moving forward
and I do expect to have hardware ready in Q1 2010, although I am known
to be an optimist
The net6501 will basically be like the other boards, just faster, with
more memory and PCI Express expansion.
There will be both a 2 and 4 port gigabit ethernet version, using Intel
controllers, where the 2 ports version will be targeting small servers,
with up to 4 of them in a 1U case....
And yes, it will be based on the next generation Intel Atom processor,
the Pineview platform, in both single and dual core versions and with up
to 2 Gbyte DDR2-SDRAM soldered on.
And it will still be low power and high reliability, with passive
cooling. Ok, a tiny server with two 2.5" 10K rpm SATA drives will need a
small fan....
A new goodie will be onboard NiMH batteri charger/controller.
Best Regards,
Soren Kristensen
CEO & Chief Engineer
Soekris Engineering, Inc.
Nice, eh?
Could make a nice experimenters board. You could compare and see if there is any difference between alix/ethernet and soekris/ssd.
Last edited:
I am talking about a processor with direct Ethernet stack and I2S out,
such as the beagleboard for the DIY enthusiasts.
It just needs an additional I2S reclocker to get closer to the Touch approach.
People, like me, who'd like to get things done, get things done timely and with lowest effort go for a Touch.
The SW mods I propose can be applied without any in depth knowledge.
There is not much to strip down on the HW side. Just open up the box to remove the screen.
By opeinng the box you actually automatically remove the screen
You can build your own linear 5V supply or you buy an "audiophile" PS.
Your solution sounds to me like a basic PC with an OS and some off-the-shelve apps . No idea what you call DIY here!?!?
This you can do on any other PC. I consider a Guru plug a much better
base.
Wrong forum. Wrong thread. Ayre products you should discuss at Computer Audiophile.
Beside that many people, including me, claiming since years -- because we're using it -- that USB isolation gets you a big step forward. Charles Hansen just picked it up at AA and applied it to his DACs with great success.
Agreed. You'd probably have the same with SPDIF out from the Touch.
I'd say - give it a try to make up your own mind.
After applying the tweaks the Touch IMO sounds great. At 300$ I couldn't be more then happy with that device. Yes I could. A Duet II at <199$ would be even better.
More and more USB-DACs do have PICs or DSPs or FPGAs onboard. You'll face similar issues on those devices.
John from EC-designs has shown on his SD-card player that the only way to get around the interference/intermodulation/noise problems is to use one master clock for your processor and DAC.
I2S isolators are slowly but surely popping up here and there in discussions.
Let see how this develops.
Anybody out there who need s more then that? Do you know Dan Lavrys article about Hirez >96kkhz.
That's a risk.
I am fine with the performance. I don't experience more or less Bugs then with other SW. MPD plus Ubuntu/Alsa/Pulseaudio was ten times worse at the time of using it. The MPD feature list won't come close to the SB feature list.
The SB stuff - the audio relevant part - runs pretty much independent from your/any OS, what I consider a major advantage. However with the Touch
they also entered the embedded Linux and Alsa environment. This I consider a major disadvantage. The DUET firmware approach was IMO the much better choice.
That's why I take it out, which sounds even ways better.
Please read what I am writing.
Installing Ubuntu is a no-brainer. Disk-space is not a problem either.
I changed my startup scripts and boot my audio client up in network-recovery mode. I start only the processes I need. In fact I don't run the
Ubuntu overhead at all that way. I get along with a couple of MB ram utilization.
Something like that you can implement within 10 minutes and you can use the infrastructure, tools and support of Ubuntu. (I should write an article about that)
I tried Voyage, when you came up with it. You know my opinion.
Sorry for my poor english.Though I guess you understand what I am saying. I am sure I know more about what I am talking about then you do. *AND* I do consider myself pretty open minded. And I'd try to avoid putting an "opinion" out into public, which I couldn't prove. Usually I verify what I am claiming.
Have you ever had the Touch in your hands?
This is really complete nonsense and disqualifies you for any further discussions.
Poor guys.
Cheers
Major streamlining of full ubuntu desktop installation is no simple task, even pro seasoned linux users. I do doubt you can do it in 10 minutes. Modern linux desktop is VERY complicated and ubuntu is not renowned for correct and minimal dependencies among packages. Even stopping some services after login, easily accessible to any user, ruins key desktop functionalities without any warning. Such as disabling the start of gnome power manager (I do not want any power-saving features on a desktop PC ) leads to quiet disfunction of the shutdown button in gnome menu. Ubuntu desktop is definitely not easy to streamline.
For a streamlined audio client without GUI it is much less work to install a debian netinstall and add the remaining packages, or even use the specialized embedded distros (basically a predefined set and configuration of debian packages).
For a streamlined audio client without GUI it is much less work to install a debian netinstall and add the remaining packages, or even use the specialized embedded distros (basically a predefined set and configuration of debian packages).
To start just NFS, SSH and e.g. MPD it doesn't need more then three more lines in the recovery mode startup-scripts. This won't have any impact on the actual Ubuntu installation.
There is no GUI with my setup or any other fancy service started.
There is just the kernel+my apps of choice up'n running.
This way I strip the system up, not down. This approach I consider very efficient.
My headless FitPc is up'n running in a few seconds.
Again. I am not telling fairy-tales here. I am running my system that way since almost 2 years by now.
If you go back in the thread you'll find discussions about those micro Linuxes quite a while back. I was spending more times tailoring those systems and fixing problems, compiling apps. and making kernels etc. then playing music.
One of the major advantages of my setup is that I can boot up the system in normal mode too. I can use a complete Ubuntu system with all its gadgets if I want to. And of course I take advantage of a rock solid and quick installation process.
If I'd tell you how I done it. You'd be up'n running in 5 minutes (assuming an Ubuntu would have been installed beforehand) - I stick to that statement
Cheers
OK, recovery mode (i.e. single mode, runlevel 1) in ubuntu is just as plain as in any other linux with SysV runlevels. BUT using a single mode as a regular operating mode is a BAD practice and bad example for newcoming linux users. The single mode has its purpose (troubleshooting, recovery) and adding new daemons to it produces a functioning system for your needs, but definitely is not the proper way. A clean way would be using runlevels 3, 4 or 5, keeping only required services, and leaving runlevel 1 intact.
Last edited:
phofman, sorry, but I don't buy your arguments. I don't see your problems.
As an example a Squeezebox Touch as well as other low level embedded systems are running in exactly this mode - those devices don't have different runlevels. This runlevel is all what's needed to start a diskless audio client running apps such as MPD.
There is no security risk associated to that runlevel, since login is still required.
Cheers
As an example a Squeezebox Touch as well as other low level embedded systems are running in exactly this mode - those devices don't have different runlevels. This runlevel is all what's needed to start a diskless audio client running apps such as MPD.
There is no security risk associated to that runlevel, since login is still required.
Cheers
Soundcheck:
Thanks for prompting the quick lesson on run levels. The wikipedia article Runlevel - Wikipedia, the free encyclopedia seems to be a good intro into it. if I'm not confused what you are doing is changing run level 1 (single user no network) into a different run level with network access. Since the whole thing is a convention and especially with Linux you are always free to make it your way its not really important. However the security types would point out you are running in a less secure mode if the network is active and you are running as an admin.
I prefer using Voyage for a bunch of reasons but I'm not about to run down any other choices. I had a friend that for years ran Windows 3.1 in recovery mode since it was the only way he could get the system to run. For the less skilled its better to use a system as designed if you can than to bend it in a non-standard way and hope for the best. Something a complex as a linux distro (even the stripped down ones) have so many moving parts that need to be tested and sorted using one in a very different way could lead to unpleasant surprises.
Thanks for prompting the quick lesson on run levels. The wikipedia article Runlevel - Wikipedia, the free encyclopedia seems to be a good intro into it. if I'm not confused what you are doing is changing run level 1 (single user no network) into a different run level with network access. Since the whole thing is a convention and especially with Linux you are always free to make it your way its not really important. However the security types would point out you are running in a less secure mode if the network is active and you are running as an admin.
I prefer using Voyage for a bunch of reasons but I'm not about to run down any other choices. I had a friend that for years ran Windows 3.1 in recovery mode since it was the only way he could get the system to run. For the less skilled its better to use a system as designed if you can than to bend it in a non-standard way and hope for the best. Something a complex as a linux distro (even the stripped down ones) have so many moving parts that need to be tested and sorted using one in a very different way could lead to unpleasant surprises.
@ 1Audio
As you say. When changing your Linux, you should know what you're doing.
I can tell you, the changes I applied are pretty trivial. If you'd have had a look at those startup scripts, you'd know what I mean.
I do have quite some professional sysadmin and system integration experience on Unix systems. I wouldn't come up with an unsecure solution proposal.
Ubuntu's recovery mode comes with two options:with and without network access and is protected with a password.
As I said. There is not more or not less security as you'd have it with a SB Touch.
There are Linux systems which boot right into the root account without password, if booted into single-user mode. That's not the case with Ubuntu.
Cheers
As you say. When changing your Linux, you should know what you're doing.
I can tell you, the changes I applied are pretty trivial. If you'd have had a look at those startup scripts, you'd know what I mean.
I do have quite some professional sysadmin and system integration experience on Unix systems. I wouldn't come up with an unsecure solution proposal.
Ubuntu's recovery mode comes with two options:with and without network access and is protected with a password.
As I said. There is not more or not less security as you'd have it with a SB Touch.
There are Linux systems which boot right into the root account without password, if booted into single-user mode. That's not the case with Ubuntu.
Cheers
Soundcheck,
whatever you do, is your own responsibility. Just please do not present your solution to linux beginners as a proper way to go and admit that it is a hack to make your life easier.
Runlevels have been designed for a reason. Tweaking perhaps the most important one - single, to serve your purpose, is a bad practice and no amount of justification will change it. Plus there is no reason why on such a powerful system, capable of installation and running ubuntu, the buggy audio clients should run under root. That is a bad security practice and if you really have "some professional sysadmin and system integration experience on Unix systems", you should not even propose such solution, let alone defend it. And if you perhaps run them under another user which I very doubt, it still denies the principle of runlevel 1 where only root should be logged in. It does make sense as e.g. /home can be mounted via nfs while network is disabled in the single runlevel. And running the unreliable network manager, the default networking daemon in ubuntu (and again I doubt you have changed it to the legacy networking since it is not that simple), in the single mode, as a regular operation, absolutely wrecks the principle of the single mode as a secure and always available minimum working system in case of troubles.
whatever you do, is your own responsibility. Just please do not present your solution to linux beginners as a proper way to go and admit that it is a hack to make your life easier.
Runlevels have been designed for a reason. Tweaking perhaps the most important one - single, to serve your purpose, is a bad practice and no amount of justification will change it. Plus there is no reason why on such a powerful system, capable of installation and running ubuntu, the buggy audio clients should run under root. That is a bad security practice and if you really have "some professional sysadmin and system integration experience on Unix systems", you should not even propose such solution, let alone defend it. And if you perhaps run them under another user which I very doubt, it still denies the principle of runlevel 1 where only root should be logged in. It does make sense as e.g. /home can be mounted via nfs while network is disabled in the single runlevel. And running the unreliable network manager, the default networking daemon in ubuntu (and again I doubt you have changed it to the legacy networking since it is not that simple), in the single mode, as a regular operation, absolutely wrecks the principle of the single mode as a secure and always available minimum working system in case of troubles.
It depends on your distro, but "runlevels" are mostly just a way of saying which "snapshot" of services you want running at boot. Ubuntu has switched a new init manager so it's configuration is slightly different to other unixes, but for most distros there is literally a directory with a bunch of symlinks in it and when you request to be at a certain runlevel the system simply runs all the scripts in that directory. This means you can easily compare run levels and change around their contents
I realise that doesn't actually help anyone do anything practical, but hopefully that gives you an overview of what we are actually talking about? I guess to put it in Windows terms it's basically a way of saying "run this particular bunch of services and give it a name of 'level 3' " - you are completely free to choose what services are included in a runlevel (but there is a weak historical precedent which hints at some common conventions)
Good luck
I realise that doesn't actually help anyone do anything practical, but hopefully that gives you an overview of what we are actually talking about? I guess to put it in Windows terms it's basically a way of saying "run this particular bunch of services and give it a name of 'level 3' " - you are completely free to choose what services are included in a runlevel (but there is a weak historical precedent which hints at some common conventions)
Good luck
...Further I don't in principle see any obvious reason why a machine in run level 1 is likely "insecure"? Run levels are largely just a way of quickly switching to a given bunch of services running. Personally I would use something like runlevel 2 or 5 if I were doing this, but that's only because I would want to keep a working level 1 in case I break something and need a very plain rescue system - no other reason though (just make runlevel 2 a clone of runlevel 1 and proceed from there...)
Security is going to be largely a function of how you setup the machine and in general linux machines are provided in a manner which is secure against intrusion from the network almost no matter what you do with them.
Good luck
Security is going to be largely a function of how you setup the machine and in general linux machines are provided in a manner which is secure against intrusion from the network almost no matter what you do with them.
Good luck
ewildgoose,
I agree that the technical implementation of runlevels is very simple and their names/numbers are just a convention. Still, having a minimum bootable system (the purpose of single mode) is pretty useful, e.g. if some daemon started automatically in runlevel 2 triggers kernel panic caused by a buggy driver. Not so uncommon, unfortunately. And the option to boot from live medium (CD, USB) is not always readily available.
Until recently, RT kernel in combination with ICE1724 soundcards was causing kernel deadlock. I would not want my single/recovery mode to initialize and use such HW automatically from init scripts.
I agree that the technical implementation of runlevels is very simple and their names/numbers are just a convention. Still, having a minimum bootable system (the purpose of single mode) is pretty useful, e.g. if some daemon started automatically in runlevel 2 triggers kernel panic caused by a buggy driver. Not so uncommon, unfortunately. And the option to boot from live medium (CD, USB) is not always readily available.
Until recently, RT kernel in combination with ICE1724 soundcards was causing kernel deadlock. I would not want my single/recovery mode to initialize and use such HW automatically from init scripts.
Last edited:
phofman:
If we were talking about an appliance (router, media player, etc.) you would not need run levels since your users would never have a way to use them. but the root access would be very locked down and the system would run at a user level.
For an audio machine built by a DIY'er who is more interested in playing content I would shy away from a desktop system kludged to operate like an appliance. Voyage has its solid read only OS so its really hard to screw up and it comes without the desktop audio stuff so you don't need to hack/remove it. There is a few hours of learning to get ALSA and MPD working well. And I would not suggest that accomplishing a running system qualifies someone to cross compile and bring up the same on an arm core. But unlike Win or Mac based systems one can do pretty much anything in Linux with time and Google.
If we were talking about an appliance (router, media player, etc.) you would not need run levels since your users would never have a way to use them. but the root access would be very locked down and the system would run at a user level.
For an audio machine built by a DIY'er who is more interested in playing content I would shy away from a desktop system kludged to operate like an appliance. Voyage has its solid read only OS so its really hard to screw up and it comes without the desktop audio stuff so you don't need to hack/remove it. There is a few hours of learning to get ALSA and MPD working well. And I would not suggest that accomplishing a running system qualifies someone to cross compile and bring up the same on an arm core. But unlike Win or Mac based systems one can do pretty much anything in Linux with time and Google.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.