rfdave

In theory, there's no difference between theory and practice.

My first job was at a large consumer electronics company, where I would ask for component quotes up to 10million/year, and not get laughed at... I'm now at a smaller SDR house, where 2k/year is pretty good volume. The level of customer support from the vendors and distributors is significantly different now, and not in a good direction.

We've also run into the "Sorry, we can't sell you those parts" a few times, so it's a for real thing.

There are turnkey manufacturers who will build all this for you, but the starting place is a manufacturable design. If you started with off the shelf modules, then one of the first steps would be to get there. So, schematic development and PCB layout to come up with a single PCB that incorporates everything, into an off the shelf plastic housing would be the first step. Then you've got something to build 5 or 10 of, and see if people want to buy them. Bunnie Huang's blog has some good material on this.

I think I tend to agree with you. In this context, I'm trying to make the point that the ability to "ship" software quickly and easily seems to lead people to the belief that shipping hardware is almost as easy, and it's not. I'm not trying to imply that actually writing the software is easy, just that the process of making a piece of software available to millions of people is substantially easier than shipping hardware to millions of people. There's a lot of tribal knowledge that's difficult to access via Google, although things are getting substantially better, as lots more people are going through small scale production ramps and blogging about it.

It's really amazing that people with domain knowledge in one field seem to think that everything they don't know about is easy.