Using a nook for xcsoar

On Sunday, October 21, 2012 9:57:29 AM UTC-6, Robert Dunning wrote:
There's a good reason fighter sticks and throttles have

over 20 buttons, switches, and pointing devices on them



I think if you try to design a good joystick system for a glider, you'll find it is a challenging (and rewarding) task. You need to balance the learning curve (a glider pilot may not spend hours each night memorizing the layout of a complex joystick, he may not be able to practice with a really complicated joystick outside of the cockpit) against the value of one button access to critical functions. (Maybe a good design includes a simulator joystick that the pilot can hook into Condor to practice with.)



The same challenge of designing a good joystick for a glide computer applies to designing a good touch screen for a glide computer. (1) what are the critical tasks that need to be one-touch (one button) access? Are there four such critical tasks, five, fifteen? Depending on how you analyze the needs of the pilot, the number of buttons on your joystick (or the number of one-click areas on a touch screen) will either be few or many. (2) after you've identified the critical functions, how do you account for pilot personal preferences? will you allow some sort of customization of the buttons (shortcuts to whatever the pilot thinks are his critical functions)? Maybe you have "themes" from which a pilot can select. This is another challenge -- a really flexible and customizable system adds complexity to the overall computer (regardless if a touch screen or a joystick).



So I "stick" to my original view that whether it's a joystick or a touch screen matters less than the effort that goes into the overall design. The goal is a flight computer that balances power with flexibility with ease of use. To the best of my knowledge, scant research has been accumulated about this aspect of glide computer design.



Rob Dunning

Possibly, I say possibly, the best approach is to have the glider computer do the switching itself. Most already can switch from glide to climb mode and back automatically. If you have input a task, it will know when to switch to final glide mode. It's likely a good algorithm could set "bugs" and M number better than most pilots.

With the computer doing most of this itself, a few over ride switches are all that is needed to deal with exceptions.