Re: [gpsd-users] epx, epy values never change

[Eric S. Raymond]

Jeff Loughlin <address@hidden>:
> On Fri, Oct 21, 2016 at 10:54 AM, Eric S. Raymond <address@hidden> wrote:
> >
> > The error estimates are computed from satellite geometry and a scalar
> > multiplier called UERE, an empirical constant ralated to the physics
> > of delay jitter in the ionosphere.  They tell you your uncertainty
> > of location *if you have sat lock*.
> >
> > Local consitions like tree cover don't uncrease or decrease this 
> > uncertainty,
> > they control whether or not you can get sat lock.  One minor exception: if
> > you can see three sats but not four, you get a 2D lock that doesn't give
> > you altitude.
> 
> 
> Thanks for the explanation.  But now I'm wondering what epx and epy
> were showing me before (prior to the fccb36 commit), because those
> values seemed to make sense and reflect the local reception conditions
> - I could start gpsd with my GPS sitting on the window sill and see
> epx/epy values of 5-10m, then when I moved the GPS away from the
> window (but still have satellite lock) those values would immediately
> jump to 30-40m. Now they don't change, which makes sense given your
> explanation above, but that seems less useful.  I guess you're saying
> the values I was seeing before were wrong?  But if wrong, why did they
> seem to make sense, and change appropriately with the local conditions
> at the receiver?  It seems to me there was useful information there
> before that has been lost - unless you're saying that information
> wasn't really useful and only *seemed* to make sense...

That's what I think was going on.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>