Nice discussion. You hosted a panel discussion at CES back in 2008 where I tried to voice this coming trend. The bottom line for chip makers is that it isn't enough to have the cheapest, fastest and lowest-power part---there also has to be an ecosystem that helps enable the shortest time to market and lowest risk. Attracting a community doesn't just mean that hobbyists are going to contribute software, but also that people build confidence in the platform and third parties have an audience to demonstrate their solutions on that platform.
I do think in general that chip maker dev boards do not develop a real community; after all, the biggest ecosystems aren't (Arduino, Beagle, etc).
However, sometimes it does happen, with the MSP430 LaunchPad being the best example (I'm sure TI's inintial <$5 pricing helped).
mbed is an intesting case; from a quick glance at Wikipedia, it looks like was started by two engineers on their own time, but it currently has a lot of chip maker support with dedicated developers to do the core development.
BTW, Squirrel was developed by Alberto Demichelis; it is also used by a number of games.
@AZskibum, I am glad you are making a clear distinction here. That's very helpful. So the friends help friends free" model might be happening in an open-source community, but so-called "community" organized by a chip vendor is more or less a marketing front. You need real professionals to help friends...eh...customers.
My personal opinion is that they view the "community" more from a Marketing perspective -- a potential source of extending functionality or enabling new applications. Some of that activity will have commercial value, but much of it -- maybe most of it -- will not. That is a far cry from the notion that the community will do a bunch of work for free, thus saving the company the expense of hiring developers to do the work internally.
this well designed sous-vide cooking machine. the machine is running squirrel - a scripting language used by electric imp which is easy for designers to work with , and not the usual c/c++ so common in embedded.
"Do you know of such examples and do you think that will ever succeed?"
I frequently hear of these, but can't name a specific one.
The very nature of the hobbyist/maker/midnight-tinkerer world is that you won't know who will succeed until they do. Look at Apple Computer - two guys in their garage. The same with Hewlett-Packard.
Also, one needs to be careful of the definition of "professional" vs. non-professional engineer. I know a number of very successful engineers who have no college degree, for example. To me, a professional engineer is someone who does engineering (design based on scientific knowledge) as opposed to tinkering, and who does it successfully.
I, for one, as a reporter who covered the news when Freescale announced WaRP earlier this year, have a great hope that it will succeed.
I do understand that such an undefined, and then, emerging market as wearable devices needs innovation and support that would help push that innovation.
In the end, it looks like this is going to be the race of who makes it easier -- in terms of building blocks, documentaion, development environment and support on the ground, rather than who has the fastest chip at the lowest power.
We shall see who will get the mindshare of the makers -- including weekend and daytime professional engineers!