[LinuxPPS] Considerations on pps_event()

[gnu not unix]

Hi Hi--

In message <20070220085204.GC30997 at enneenne.com> you write:
>Hello,

(a design case...)

>I'm asking to myself if this could improve the PPS resolution...

The Mogul/Windl et. al. pps implementation do use a header define 
as the  mechanism for the timestamp. So that is the right direction,
I think.

It still puzzles me about the "square wave" offset my data
plots were showing. I restarted ntpd and its now smoothed
out the square wave effect. The offset is comparable to the
freebsd 5.4 system now, about +/- 1 to 2 microseconds. Nice.
But why does it go away sometimes? I posted to the list
about this before. And now its smoothed out again.

../Steven