No I'm not moaning about the OverDue, honest.
I was thinking today about all the things I could to with an ARM-based Arduino, and to be honest the list was pretty short (actually I couldn't really think of anything in my areas of interest) so now I'm wondering what people will do with all that horse power.
Real-time spectrum analyser? Various audio and video apps? Blinking LED?
What will be the killer app(s)?
Rob
Same thing all my other 32 bit micros do
sit in the bin and collect dust
~or~
the same stuff except faster and with more options, that I usually dont need.
less smartass answer
items that need a pile of I/O, connectivity (USB and whatnot on chip not soft emulation) and horsepower. What ... I dunno, as far as arduino goes, its abit piggy on ram and storage, and its speed does lack, so in most arduino projects you really could get away with a 8 bitter and just learn how to do it "closer to the metal", though I do not doubt there will be some surprising examples.
IO and memory can currently be obtained by using a Mega I guess but horsepower can't.
We do get a lot of posts about working with video and such, the response to them may in future be "Get a Due" instead of the "forget it" we currently have to say.
I think also that getting "closer to the metal" will be less likely to be needed which should allow less experienced people to do stuff that would have been beyond them before.
Rob
ARM-based Arduino
Which ones you got ,sir?, OLIMEXINO/MAPLE or which , im loving them want to get one myself.
IO and memory can currently be obtained by using a Mega I guess but horsepower can't.
Pretty True.
I'm wondering what people will do with all that horse power
I would like to load a Linux distribution on it and if i get a system which looks like maple/olimexino(its very much like arduino) i would give me a feel and look of arduino (though highly cluttered) and fulfil what did like to do with it.
Which ones you got ,sir?, OLIMEXINO/MAPLE or which
Neither, I'm waiting for the Due.
I would like to load a Linux distribution on it and if i get a system which looks like maple/olimexino
Then I think you should get a maple/olimexino, I doubt the Due will have enough RAM for Linux and anyway I don't think it should go in that direction, there are already a lot of boards that do that.
There aren't many designed for lower-level control, boards like the LPCXpresso do that but they don't have the Arduino tool chain.
Rob
I would like to load a Linux distribution on it
Don't, if you have to do that then the Raspberry Pi is a better bet if it ever comes out.
I don't think there will be a killer app. That implies that someone will produce something that everyone will want to copy.
For me a cheap web cam is like having a vast array of light sensors for not very much money and wiring. There are several projects I would like to do with that. For example large touch screens to make a gigantic monome.
Also the possibility of a stand alone connection to a Kinect to do proximity sensing in a much more reliable way.
Then there is the possibility of a real USB host, not the low speed version we have now. This opens up possibilities of using cheap blue tooth dongles.
Also digital signal processing will allow real time audio effects to be created. One simple project I want to do is just an audio delay. This is because for some sports like football and cricket there is coverage of the event on both radio and TV. In the analogue days you just turned the sound down on the TV and listened to the radio. You can't do this with digital because the radio audio is about 3 to 4 seconds in advance of the pictures, making it look like the radio commentators have a strange ability to see into the future. A simple audio delay controlled by a pot would be a simple enough project on a Due.
Don't, if you have to do that then the Raspberry Pi is a better bet if it ever comes out.
yes but then even Maple/Olimexino would have shortcomings? and if yes then why(factor's)?
I don't think there will be a killer app.
Yeah, maybe just a lot of smaller things that will become easier/possible.
I like the idea of a delay line.
a strange ability to see into the future.
'Er indoors and I usually watch TV on our respective laptops but for some reason there's often a time difference. Sometimes we watch the same show and she laughs for no apparent reason. 2-3 seconds later I get the joke.
Rob
its not strange its the difference in buffering
Sometimes we watch the same show and she laughs for no apparent reason. 2-3 seconds later I get the joke.
But i think it's FUN! , the moment you think something's wrong with her, she's backed up by the actual reason she was laughing and that will make you laugh even more.
its the difference in buffering
That and the fact that the TV signal has been bounced up to geostationary orbit a few more times than the radio signal.
Another thing you will be able to use a Due for is driving a display panel. Currently the memory requirements and the grunt it takes to refresh one is too much.
That's what I can't wait for, able to display my simple stuff on more than a 2 inch screen without a full computer and way too complex code to just put out readable size characters and maybe fit more than 5 touch buttons on a screen
My need for speed is because of this project. It evolves toggling 1 pin at 408KHz and a second pin at 8.5Khz with specific timing and 50% duty cycle. I plan to add digital pots, LEDs, pulse width measurement, LCD display, buttons for pulse speed incrementing, and serial output. Possibly more.
The problem I ran into was that the pin toggling would only work using Fastwrite code, and then if I did an analogread it upset my 50% duty cycle because the analogread took too long.
I purchased a Olimexino 32 and so far it looks to be up to the tasks that I need to do. However, the digital pots do not have a ready to use library so, I have to make my own code to operate them. Then, I am going to have to work a little harder to use the 3.3v limit with my hardware.
If the Arduino Due was out "with working Arduino code" and it performed like the STM32 chips I would be very happy.
digital pots do not have a ready to use library so, I have to make my own code to operate them. Then, I am going to have to work a little harder to use the 3.3v limit with my hardware.
Unfortunately that will be the same with the Due I would think. The software issue will eventually go away as (hopefully) authors migrate their code but the 3v3 is here to stay.
Rob
Sorry for my lack of knowledge but why is the 3.3 volt there to stay?
I don't see much problems in powering a pin with 5 volt.
And Arduino already has a "comparator pin"
So setting 5 volt on the 5volt pin and 5 volt on the comparator pin would do, wouldn't it? (-not doubting you skills and knowing my lack of electronics knowledge- I guess it won't but: "Why?")
Best regards
Jantje
There's been a push for years to lower the voltage (and therefore the power) used by chips, 3v3 was the first then 2v5, 1v8 etc. Before long they'll be running on 0v 
Most of the ARM chips are only available in 3v3 (actually the core normally runs at 1v8) and it's becoming common to only get modules and chips in 3v3 versions, for example the Xbees.
I'm more and more finding when I look at designing a board that it's easier to go 3v3 and upshift on the odd occasion I need to.
That will I think become more common until 3v3 is the norm.
Rob
It evolves toggling 1 pin at 408KHz and a second pin at 8.5Khz with specific timing and 50% duty cycle.
By the proper use of timers you can do this on an arduino, no need to go to an ARM chip.
I don't see much problems in powering a pin with 5 volt.
The point is that as the chip geometry gets smaller then a 5V system generates too much heat. The solution is to drop the voltage in order to not generate as much and stop the chip melting.
If you put 5V into a pin of a 3V3 processor then you will destroy it.
So setting 5 volt on the 5volt pin and 5 volt on the comparator pin would do,
Not 100% sure about what you are saying but no it would not do. The current arduino only has one comparator input the ARM chips do not have any.
So if I understand correctly you say that it may be perfectly viable to make the due 100% hardware compatible but it is not advisable.
For a non electronics arduino user it comes to me that I'll be waiting for the due to arrive, waiting for tests of the shields and then maybe find a fix, or waiting a shield that fixes incompatibilities.
I guess that all this will take some time after the release of the due.
So waiting for the due, waiting for new shields and/or waiting for shield compatibility.
Looks like this is not going to be short term ]
Now I understand better why the Arduino guys prefer a good first shot to a "to early release". There is lots at stake. Look at raspberriePi 
Best regards
Jantje
So if I understand correctly you say that it may be perfectly viable to make the due 100% hardware compatible but it is not advisable.
No it is not viable to make the Due 100% hardware compatible. You will only be able to make it 0% compatible.
For the Due you will need a different sort of shield.