[yoyodyne]

Adman,

The points you make are appropriate and well measured. You can always pound a square peg into a round hole if you use a big enough hammer. The benefits this system brings are outweighed, by its cost, and size, weight, and power (SWaP). I see this abstraction as a step in the wrong direction in terms of teaching embedded system concepts and showing students that they have the technical capability and financial resources to build a micro-controller based subsystem. As an example, we had team members build an ATtiny2313 based project in the pre-season and as a result we were able to modify those to make a custom programmable Robocoach transmitter and multi-channel receiver. One of the good things about the IFI RC in my opinion was that teams were encouraged to build sensor co-processors and at least for larger teams, that was an excellent way to allow more students to be involved in both the HW and SW aspects of the control system. Somewhere in this thread the comment was made that a CAN bus interface was not possible with the old RC but I think they failed to see that a CAN to RC (maybe serial, or analog) converter board project would be a great learning opportunity.

I think that a team really needs three control systems; one for the competition bot, one for some type of test/integration platform/mobility base, and a third for control system development. The VeX controllers were a very good low cost option as extra development platforms for our software developers. The one extra cRIO is going to severely limit the time each developer has to unit test their code. Maybe the new system will be so abstracted and easy to use that little or no unit testing will be required (humor) but in all seriousness hard lessons learned about data types, protected sections, what the volatile keyword means etc. are likely to be lost.

From a teaching perspective what was really valuable about the FRC programming experience was that it was the first time most students worked with something besides FLL block diagram code or the school CS classes in JAVA, visual C/C++/C#.

I think that we will put even more emphasis on the use of micro-controller based co-processors as a way for students to learn how to develop powerful low cost, low power affordable solutions.