more control options

[Jnadke]

Srawls and Nate demonstrated the point perfectly. PBASIC is a fairly simple programming language that seems high level, but it really isn't. I found it very simple to use (vs x86 assembly) and also somewhat robust once you start to understand what you are doing (...that means after reading the 300 page manual a few times).

Since FIRST is trying to make it easy to build a robot, these stamp controllers are pretty much as far as you would go.

[Dave Flowerday]

C and C++ are used to control hardware directly all the time. It's the operating system that abstracts the hardware, but in a situation like this an OS most likely wouldn't be used. Many processors use memory mapped I/O which allows access to input/output pins in much the same way that PBASIC does. By writing a value to a certain location in memory, you're actually setting the output pins to that value. Thus, in C or C++, you can access the input/output pins directly through the use of a pointer. This is done all the time in embedded systems. Realistically, for what we're doing here, microcontroller assembly wouldn't be much more difficult to learn than PBASIC. Also, x86 assembly isn't a very good comparison. If you look at the instruction set for something like an HC12 or similar, I think you'd find that it's a lot more simple than x86.

If such a microcontroller were to be used, FIRST could easily provide a library that initialized the processor, handled serial inputs/outputs, etc. All that would be left to the teams would be the traditional bit twiddling for input/output, and any calculations in between. However, you'd also be able to use interrupts and other features of the chip, not to mention the fact that it would run thousands of times faster than the BASIC stamp.

Personally, I'd like to see FIRST have some form of modular control system, which different user processors could be plugged into. Thus, new teams or teams without the knowledge required could just plug in the BASIC stamp module. But, for teams that wanted more power and don't mind assembly or C, there could be control modules with more powerful processors that would plug in.

[junkyarddawg]

Dave:

I agree 100%, for anyone that has programmed with more powerful languages, using PBasic is like going back to teletype from the internet. It is a painful memory to use basic when you know it would be so much nicer with something else. I'm guessing they picked the stamp chip because they knew it well enough to create the original robot controller and I'm guessing they are already discussing something better. I think a developement package similar to visio style logic diagrams would be a nice option for teams that don't want to work with text. Either way, adding external circuitry to make up for a controller that is too slow to take in a tachometer pickup is frustrating.

Just for the record, I think the Innovation FIRST people have done an excellent job, We just want more.

Tim Gates