[FOSS-GPS] GPSD, geoclue, and the geolocation service stack

[Eric Raymond]

Guilhem Bonnefille <guilhem.bonnefille at gmail.com>:
> You plan to use "libgps". Do you evaluated an higher layer? 
> I'm thinking about geoclue.

GPSD supplies geoclue with time/position/velocity reports, but geoclue
is intended to support providers of complementary information (such as
address lookup and reverse geocoding) as well.  From my position at
the bottom of the stack, I have to say geoclue's aggregation function
looks like a good idea to me and I encourage applications like
foxtrotgps to use it.

Ideally, geoclue would work side by side with the sort of map service layer
people have been discussing here.  I didn't have a lot to contribute to
that discussion, because I don't know a lot about the map end of things.
But I do think it would be silly to duplicate map-service code across 
a dozen different projects.  I'd prefer to see a map-service project with
well-defined APIs that concentrated expertise about getting and caching
tiles in one place.

Accordingly, I think the right direction for user-facing projects like
foxtrotgps, tangogps, viking, navit, etc, is to be doing less and 
less of their own low-level data gathering.  We should be looking for
cleavage planes in the stack underneath these where we can disaggregate
them.  I think map service is one such; there might be others. 
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>