Two more for someone to confirm or deny ..
1) When you average a waypoint, the estimated accuracy seems to be based
only on the EPE and the number of samples. The actual sample values are
not used. I.e. you could have 99 samples with a position 100km away
from the current point, and still have a reported accuracy of 2m.
Demo - just set it to 'averaging' and then keep walking. The 'estimated
accuracy' gets better and better even though the 'averaged' position is
moving at several m/sec, and never stabilises at all.
2) I'm almost convinced that the SIRFstar firmware won't lock on any
WAAS satellite if it isn't sending ranging data. For the last several
days EGNOS claims PRN120 has been sending the right correction messages,
but all I get is (briefly!) hollow bars.
Demo - hard to prove a negative, but has anyone EVER seen a 60CX/CSX
lock onto a WAAS?EGNOS satellite with not ranging data (hollow bars
only) and use the differential correction data?
GSV Three Minds in a Can
Google may be your friend, but groups.google.com posters definitely aren't.