[Arm-netbook] The Potential at Hand. (Was: device tree not the answer in the ARM world)

[Scott Sullivan]

On 05/10/2013 06:20 AM, Vladimir Pantelic wrote:
> On 05/10/2013 10:04 AM, luke.leighton wrote:
>
>>    ok, let's try this.
>>
>>    do you buy a product and then throw it away after 18 months when it
>> is obsolete?
>>
>>    please answer that question first and we will move on from there.
>
> no, my DVD player is now 10 years old

Then that is an example of a well built, long lived device and not the 
subject of what EOMA-68 is trying to address.

Devices that are 'smart' or have software that takes advantage of the 
internet need to be updated as standards move on. This means devices 
that have some kind of general purpose CPU in them.

Take for example web enabled desk phones, like this one.

http://www.amazon.ca/GrandStream-GXP2200-Enterprise-Multimedia-Phone/dp/B00A79G6CA/

Now step aside for moment and not think about if you would use one, but 
understand that it wouldn't be on the market unless it had value to someone.

Now that is a device where the peripherals don't need to change, you 
have your handset, audio, your screen, physical keys. But as software 
moves on the embedded process will not be able to keep up. Think of all 
the useless android bricks that are starting to fill the waste and 
recycle channels.

Now look back to late 90s,
http://www.tuxscreen.net/download/erire.com/eshan/

My the looks rather familiar now. And all of it still worked on the one 
I had until it was destroyed in a flood two years ago. You could get 
linux on it, make it really useful, but the 200 Mhz processor limited 
you. But all the peripherals were good. Now only if I could have just 
swapped out the highly integrated bits of the CPU and Memory for a more 
recent one. If only there was a common set of BUSes that made that easy.

The industry is going to keep making un-upgradeable junk until it's 
proven there is a better way. How to you prove there is a better way? By 
actually trying it out in real hardware.

So even is if this project fails, the learning opportunity will have 
been very valuable. If it doesn't fail, we'll have a chance at changing 
the world.

Already I'm looking at the Flying Squirrel Tablet and this is what I 
see. When the first CPU card becomes too dated, that screen will still 
be useful. It's not difficult to implement a bare card that takes HDMI 
or DVI and makes it just a screen. Those chips are a available, or It 
makes one heck of a nice FPGA project. Actually, as FPGA project you got 
the base for an excellent educational institutions by boot strapping and 
giving the students something usable after.

This is where I go off into lala land thinking of the possibilities. 
There are so many old computer cases (ie the Cobalt Qube) that I could 
slap together a 2 layer board design and have a functional 
computer/retro piece of art.

I have a friend in the Modular Synthesizer business and he loving the 
concept as he's been waiting for it for years. He can continue to do 
what he's good at, while also offering a better product with more 
powerful control interfaces using a real general purpose CPU. There is 
only so far you can push a micro controller, ability really far.

There is truly great potential, so don't say 'it won't work because of 
X'. It's more valuable to say, let do it and see what happens when the 
creative people start building on top of it.

-- 
Scott Sullivan