Re: [gpsd-dev] SiRF gpsmon client mode

[Eric S. Raymond]

Gary E. Miller <address@hidden>:
> With one twist in gpsmon.  Any PPS offset over a day is likely junk and
> is replaced with "> 1day".  So the larget string gpsmon will see is:
>       "-86399.123456789"
> 
> As sort of documented at the bottom of gpsutils.c.  I'll add this there
> to help.

Good.  I'll look at widening fields in the monitor panels to 21 chars 
appropriately.

I saw the constant 22 there and a bunch of other places.  As an
undocumented magic number!?! Bad hacker, no biscuit. :-)  And my error,
too, for not noticing and fixing it sooner.

I've defined a TIMESPEC_LEN macro appropriately.  I'll add an
objurgation to the Hacking Guide about magic numbers.

> > > Sort of.  I see the garbage is some JSON displayed after gpsmon
> > > starts and before the curses painted.  Annoying, but not fatal.
> > 
> > I just pushed a small patch that I think fixes this.  It does a clear
> > screen before each driver switch.
> 
> No joy here.  Same issue.

Alas.  Not a very important bug, but I'll poke at it some more.

This would probably be a good time to do sanity testing with the Garmin.
I'm not planning to ship immediately - good idea to let the code cool off
for a while first - but it would be nice to know we have all our ducks in a
row.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>