The discussion of this has probably been the longest thread on this board.
My understanding is that it boils down to two things:
1) The BB was designed and intended as a development board and for
hobbyists, not as a commercial product. It's reliability level is more than
adequate, but BeagleBoard.org can't stand behind, guarantee, or endorse it
for a use that was not intended.
Along this line, it comes down to us:
* not knowing the end commercial product use ---- so we have to make
sure you take responsibility for that and don't blame us when you use
it different than we anticipated and haven't documented every aspect
of what might matter to you
* not knowing we'll won't update the design to have a newer component
or extra feature that might not work in your exact application
* can't handle getting a boat load of returns because of one of the
above because we've priced it as a development board without all the
margin required to handle these potential issues
2) If you, as an oem, order thousands of boards rather than manufacturing
your own from the freely published design, you are restricting the
availability of boards to hobbyists and other developers. The hope was to
provide small numbers of boards to many people and organizations rather than
many boards to a small number of people and organizations. High volume
commercial use works against this.
This has become the more practical issue. It wasn't that we didn't
know the demand was going to be high, but our manufacturing partners
intended to have 16 week lead times and information from people
placing volume orders. When distributors had a bit of a stock position
last year, they hit the breaks a bit and that started a very ugly
domino effect. They asked to slow down shipments and left CircuitCo
scrambling to downsize their production without losing their shirt. At
the same time, the stock position triggered more people to try out the
boards and stock them in places like Radio Shack and Microcenter and
everyone started coming out of the woodwork looking for boards.
Back in April, when I wrote
I anticipated us being back in that in-stock position before the end
of summer. Well, it turns out it took a while for Element14 to start
to really get an idea of the size of the demand and they are still
scrambling to catch up with their own backlog.
We can't stop (and don't necessarily want to stop) people putting
these boards into products, but they need to:
1) take responsibility to determine its suitability themselves
2) take ownership of the supply chain by establishing sane lead-time
orders with our existing manufacturing partners or others and not
disrupting the existing supply
I believe this would have all been a lot better if we'd taken all the
risk (of building too many) in one place, but we've instead spread it
out across many distributors and multiple manufacturing partners
(CircuitCo being the official one, Element14 being an alternative and
now I've seen that folks like Special Computing are shipping "no logo"
boards). Still, the comments on the thread speak for themselves
regarding the quality and clearly the demand is there.