Okay, I'm being a smarta** here, but really 3 to the 2nd power equals 9 and that's just for 1 config.
Smarta** off now. I see what you are saying.
Wish I could have been of more help. I'm pretty sure that others more knowledgeable than myself will offer suggestions.
Thanks for playing, but bzzzt!
3! = 3*2*1 = 6. I have three boxes to cycle power on, with no repeats, and order is assumed to matter. So all permutations are considered distinct and I have three choices for the first, two choices for the second, and exactly one choice left for the third. Total possible sequences: 6. Now, you might have thought I was doing a different experiment, for which the answer is 9. That's fine, just not what I was describing. Really, Dan, I do appreciate your chiming in here. And if you have any ideas -- even crazy ones as long as they don't involve submerging the ooma in perfectly good Guinness -- I am interested.
Meanwhile, the Murphy Principle seems to be in full effect. Since Sunday afternoon about 14:30 MST, when I implemented the locked-MAC suggestion from murphy (coincidence? what's in a name...), the ooma hub has gone down exactly once, which was sometime between 01:30 MST and 07:00 MST Tuesday. I suppose the failure may be multimodal, and murphy's suggestion knocked out the first-order term. But that would fail Occam's Razor. If history repeats, the ooma will lock up again sometime overnight and I'll get another shot at it in the morning.