PPS over USB

[Eric S. Raymond]

Dan Drown <dan-ntp at drown.org>:
> >Eric: Please plug your 701W into your Pi with HAT and collect some data.
> 
> Here's 7 days worth of data:
> 
> https://dan.drown.org/rpi/usb-vs-gpio-pps.html
> 
> The Rapsberry Pi 2's GPIO-PPS based local clock offsets at 0.1 percentile
> and 99.9 percentile were -3.61 microseconds and 4.13 microseconds.
> 
> The Raspberry Pi's USB-PPS based local clock offsets at 1 percentile and 99
> percentile were -582.97 microseconds and 539.08 microseconds.

Dan and Hal, how do you interpret those results? I would naively suppose
that the listed fudge should be the mean of each distribution (assuming
it looks Gaussian) and thius the arithmetic average of those pars
- am I missing anything?
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>