Comment 7 for bug 696562

Happened again when trying to demonstrate at the AEGIS UK Plenary :-(, pretty quickly - after only two or three presses. I think my physical switch h/w was sticking, which correlates with the idea that switch noise is what brings it on. Disconnects then reconnects after short gap, every 15s or so, with no switch events being delivered. Lights on the board were the same as when it's working: only one light, from the green PCB, just underneath the top edge of the red "sparkfun.com" thing. (Top edge = when reading sparkfun.com / BlueSMiRF the right way up, which is the opposite way up to the green board it's stuck on).