Does the message go straight from one end of the mesh network to the other, or can it store things for later forwarding?
Can it really support a mesh network with thousands of clients? In a mesh network with thousands of clients, how much bandwidth does each client get?
How do you divide time between different clients - is it contended like wifi so you have to keep the network uncongested?
How does roaming work? What triggers provoke roaming events?
If I want to communicate with one device 10cm away and another 10km away at the same time, is that possible? How do you avoid near-far problems?
I think we probably can sell in all of North America without changes, I'm checking on it. Europe is also pretty easy design wise, I just need to consult with our radio expert about certification, it was something I just did not look into, with all the things it takes to make this happen! But I should update the FAQ to say "we're" looking into it, since we absolutely now are!
Disappointed that this appears to be US only. would love to use a few of these for monitoring our Hackerspace!
Actually I'm also involved in mobile ad hoc security research so might try and 'acquire' a few of these for testing out some collaborative network theories on our quadcopters... but can't do that as current delivery status stands!
Good luck with certification.
Also, I don't feel comfortable donating before a finalized hardware design is reached. I feel the potential for bait-and-switch is too high and no words will convince me otherwise due to the nature of crowd-sourcing EULA, sorry.
The more backers we get, the easier delivering this hardware will be. I won't try to convince you of anything you can't be convinced of, but I will tell you that if you are interested I would really appreciate your support, and as per the kickstarter terms of service I can at least promise that if what we are delivering does not meet your needs, you can cancel your order and we will refund the full amount.
The EULA does require me to either deliver or refund you, and I absolutely will comply with that.
I'd also suggest waiting to see? You could back us at $25 so you get the updates, and then cancel before it finishes if you would prefer not to spend any money at all.
And yeah, the radio chip is pretty simple. It's basically a cheap general purpose radio. The other proprietary chip is the Atmel ATSHA204, which you don't even need to use, but provides some excellent cryptographic functions.
Either way, I look forward to winning you over at some point, even if that's not till the products ship and the reviews are in! :)
This (Flutter) is using a Cortex-M ARM chip, which is entirely different. These are used in small, embedded devices. Like wireless headphones, or robotics. Cortex-M chips don't have TrustZone.
So, a few answers:
Yes, you can do encrypted mesh networks over long range. I'm a hardcore robotics guy at heart - the name "Flutter" was chosen exactly because this protocol is designed to be as lightweight as possible, and though the mesh protocol itself is still being developed (basically, that's why we need kickstarter, aside from the fact that I'm tired of ramen), I really intend for the protocol to stay out of the way. You'll also have full control over what happens, so you can always turn meshing off and use hard addressing. It's still really simple to use that way and cheaper than everything else I've seen that compares.
I absolutely will not sell a product as "cryptographically secure" unless I believe it is, and I understand that the "smallest" flaw in a cryptographic system can render the entire system worthless. I am learning everything I can about cryptography, starting with some amazing e-books I have found on the subject, and just like wireless
Our data rate is proportional to our range (this is generally true for any radio system), so we get 1km range at low data rates (1200baud tested, but we physically ran out of test room and had a strong signal). Theoretically this radio can do 3km at 1200 baud and it looks like we should be able to do 1km at 30kbps at least. Also I think the new radio chip doubles all the data rates in these cases (I would have to confirm), and I do know the new chip has higher transmit power and higher sensitivity, so range should be even better.
There is a cheap radio amplifier (PA and LNA) called the CC1190, and texas instruments has all kinds of datasheets on using either of the two radios we are working with to give us 10km range, or much better data rates at 1km, etc. We're going to design the protocol with expansion in that direction in mind, if people really dig the range thing (it's a bit much for a beer brew setup, but if you want dark meshnets, lets do it!)
I'm a hacker at heart. I first learned lockpicking at the old Noisebridge location (the original one). I love 3D printers and open source, did a rant/thread about open hardware on the Ultimaker mailing list a year back (I was one of the first in the bay area to get an Ultimaker too, and designed a new mount for it that I have on thingiverse). I got on hackaday for some 3d printed helicopter blades, and have been building robots since I was 15. There's a longer bio about me on our site, at flutterwireless.com
See also me ten years ago talking about robots (and I feel silly for the "few people do" comment, but I was 15 and trying to keep it simple). http://www.youtube.com/watch?v=8FJu1eL_dYs
Please ask me anything!