[Tom Line]

I think you'll find that many of the top teams use a script or command based auto system.

We switched over to it for the first time this year. The flexibility it gives you is astounding.

In 2011, we had a 'standard' state machine autonomous, where changes required you to go in, change portions of the program, and recompile each time. That resulted in a horrible failure where we were on the practice field trying to complete 2-tube-auto but could not because we would only get one or two repetitions of program - deploy - run auto each time due to the field time constraints.

In 2012, we went to a script system. We took pointers from teams like 33 who were gracious enough to post their code.

As a result, changing autonomous is as easy as changing a comand or number in a text file, and clicking 'save'. The widget we built ftp's it to the robot, and the robot program re-reads the auton files each time it executes auto-mode so it updates any changes. Writing an entirely new autonomous routine literally takes a minute. No new programming, just a text file.

It's worth the time to spend an offseason working on it.