Re: [gpsd-users] [gpsd-dev] Getting the driver-switching logic right

[Eric S. Raymond]

Peter da Silva <address@hidden>:
> The only time you need to disable any driver is when you have a conflict,
> for example if there was some possibility that a non-gps serial device
> could be confused by a probe. Is that what people are concerned about?

This has never come up as a real issue. I've seen a complaint or two
about the theoretical possibility.

This is a list of every device we know about that needs a wakeup string:

Motorola OnCore GPSes
Ashtech/Thales/Magellan Professional GPSes
True North Technologies - Revolution 2X Digital compass
Jackson Labs Fury

These are all RS232, so we avoid the major can of worms by never
probing USB devices before they've IDed themselves by sending
checksummed packets of a known type.

Because people have to set up RS232 devices manually rather than relying
on hotplugging to notify the daemon, we presume that any device put on
an RS232 port anded to gpsd is a GPS.  This hasn;t gone wrong yet that
we know of.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>