[LinuxPPS] a fluke...?

[Scott Doty]

I'll be investigating plotting estimated error (if that's what you are
plotting)...pointers would be welcome. :)

So far, it (est. error) is alternating between 0 and 1 microsecond with linuxpps + ntp 4.2.4...

      _[/usr/src/kernel/git/git.enneenne.com/linuxpps]_(root at admin2.noc)_
      # git log | head -3
      commit 5f1bb8aa8d222dacc686abdb6c725468cb3eb7ec
      Author: Rodolfo Giometti <giometti at linux.it>
      Date:   Mon Jan 8 17:33:22 2007 +0100

This is a RH7.3 system w/self-hosted gcc-4.1.1, newtime.sr.sonic.net. 
Pentium III, 750MHz, 256M (An old Dell 1U) with a Garmin GPS18 ("bare wire")
with the PPS line wired to DCD on the serial interface.

 -Scott

p.s.:

_[/root]_(root at time2.sr)_
# ntpdc -pn
     remote           local      st poll reach  delay   offset    disp
=======================================================================
=127.127.1.0     127.0.0.1       10   64  377 0.00000  0.000000 0.03052
*127.127.20.0    127.0.0.1        0   16  203 0.00000 -0.000005 0.66179
=204.123.2.5     64.142.100.70    1   64  377 0.00604  0.000676 0.07005
=216.218.254.202 64.142.100.70    1   64  377 0.00513  0.000261 0.06412
_[/root]_(root at time2.sr)_
# ntptime
ntp_gettime() returns code 0 (OK)
  time c9696f68.d35a0000  Mon, Jan 29 2007 23:16:56.825, (.825592),
  maximum error 445204 us, estimated error 1 us
ntp_adjtime() returns code 0 (OK)
  modes 0x0 (),
  offset -4.000 us, frequency 12.924 ppm, interval 1 s,
  maximum error 445204 us, estimated error 1 us,
  status 0x1 (PLL),
  time constant 6, precision 1.000 us, tolerance 512 ppm,