Re: [gpsd-dev] SemPiTernal - Bounding PPS uncertainty

[Eric S. Raymond]

Phil Salkie <address@hidden>:
> If PPS signals are always less than 500ms wide, it would be pretty
> straightforward for the kernel driver to determine the leading edge by
> looking at the width of a pulse, and inverting its logic if it sees
> the pulse width is >500ms.  This obviously wouldn't work for an exact
> square wave output, but that would seem like a weird thing for a
> vendor to build, just because it's harder to determine which edge one
> should be looking at

Yeah, I figured out both these things last night when thinking about you and
I could check the Adafruit HAT's behavior with a scope.

Everybody, please forgive my occasionally fumbling and ignorant questions.
I'm trying to get the HOWTO *really right*, and I don't have all the
knowledge I should about signal-level behavior of electronics like this.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>