Per,

The reason I was thinking about one of the LINUX-based boards simply is that y ou then have a low/no-cost toolset and a low-cost hardware environment that can execute the code in native mode.

This would appear to me to be the least-convoluted way to do the job.

Is that not true? If not, why? Many of these boards have the higher-end MCU's, as you've stated, but they do directly execute the code in question, and any task-specific peripherals, whether external or on-chip, will require native code to drive them. This way, I can (a) use the on-chip peripherals, or (b) attach external peripherals and drive them with native code. Timing (execution speed) and addressing differences can be worked out and dealt-with in header files, can't they?

RE