Why I joined: the NanoNote

Werner Almesberger werner at almesberger.net
Mon Apr 20 03:03:03 UTC 2015

Ron K. Jeffries wrote:
> But I wonder if a niche remains (or not) for a similar but modern device.

I think for many people an inexpensive agenda type of computer with
a keyboard still holds some appeal.

Another market where I could imagine something Ben-like to be popular
would be the new "Maker" segment, i.e., where people play with Raspis
and Arduinos. Having a device that has strong low-level IO
capabilities (maybe 2 x UBB plus some simple "docking" interface,
nothing fancy - GPIOs and maybe USB host are all you typically need)
plus all it takes to have a UI may hit some spots there.

That's by the way how I use my Bens: one sits permanently on my lab
bench and talks through a collection of UBB-based adapters to my
other projects (one at a time, of course), e.g., to in-circuit
program some MCU.

That is a critical role, so I have a few more that act just like
backups for the one currently in use. Having a proper case and all
ensures that these Bens won't die of short-cirtuits or such in the
mess on my bench.  

Another one sometimes goes outdoors to act as wireless controller
via atben, talking to some other project.

> It did not have Wi-Fi or Ethernet so getting on the Internet was Not Easy.

BTLE may even be a more interesting choice than WiFi nowadays.
Still needs an all-open stack, though. But so does Anelok, so that
may stay in the family :)

> Anyway, yes, Werner, I am well aware the cost of development would be high.

For reference, I calculate my opportunity cost for developing Anelok
to be almost exactly EUR 100k until now, and it's not done yet.
Developing a Nanonote would be more complex and also more expensive
(more expensive parts, demanding mechanical engineering, can't just
make prototypes in the kitchen, etc.), so I'd say that you'd want
to have at least around EUR 300k at your disposition if you want to
start this.

> Well, maybe not quite so high if an off-the shelf dev board could be the
> computing guts...

Chances are that this would only make it bulky and/or expensive,
possibly also IO-limited. There is quite a number of open hardware
designs that use reasonably potent SoCs, so that part could be easier
than one may expect. Perhaps one would want to get a company with
suitable design experience - e.g., Olimex - aboard for such a

- Werner

More information about the discussion mailing list