Nokia N900

Joshua Judson Rosen rozzin at geekspace.com
Wed May 12 13:46:39 EDT 2010


Stephen Ryan <stephen at sryanfamily.info> writes:
> On Tue, 2010-05-11 at 22:12 -0400, David Rysdam wrote:
> > On 05/11/2010 09:31 PM, Joshua Judson Rosen wrote:
> > >
> > > How's the GPS? I heard from someone that the N810's GPS was lacking,
> > > though I'm still somewhat suspicious of his specific evaluation....
> > 
> > Google will back me up.  Although Google will also claim various fixes,
> > particularly "A-GPS" which requires an internet connection and is
> > therefore useless in (my) normal use.
> > 
> > Other solutions include an external GPS, which I don't really consider
> > to be a fix so much as a separate product. And in any case that would
> > still leave me using the N810's map software which is also pretty bad,
> > at least for what I do.
> 
> My N810 takes 15 minutes+ to lock on to the GPS satellites, and usually
> takes a lot longer than that (a couple of hours, which is the same thing
> as useless IMO). I don't bother to use it because of that.  On the few
> occasions I've managed to actually get a connection[...]

But is that with or without having the almanac and ephemeris data loaded?

The thing to note, here, is that GPS receivers that go longer between
use-periods actually take longer to get a fix: the ephemeris (fine-grained)
data is valid for something on the order of an hour, and takes ~30 seconds
to download if you are able to maintain the downlink continuously.
The almanac (course-grained) data is valid for ~months, and takes 12+ minutes
to download in full. Straight from the factory, or after months of disuse,
a GPS receiver will need to download both the almanac and ephemeris in full;
that basically amounts to `falling back to a brute-force approach', and
would easily account for `15+ minutes or even hours' of time to first fix
(TTFF). Note that, because there's no *uplink*, the way that you resolve
having missed any part of the transmission... is just to wait for it to
repeat....

If you go for a while without using the thing, and then try to start it
in a sufficiently far-off location (`it'd be great if my GPS worked now,
I guess I might as well try it'), then that's probably even worse.

What I do with my FreeRunner is that I have it configured to turn its GPS
on whenever it's plugge into an external power-source, so it'll download
updated data from the satellites all while I'm asleep (and both of us
are recharging), and whenever I have it plugged-in in my car.
I suspect that the car-based units have misleadingly quick TTFF because
they're also able to use this trick.

This is why I'm dubious of these `N810's GPS receiver is slow' claims--
because coupling them with `... so I never use it' is actually a
vicious cycle. If that *is* actually the issue, then the fact that
the N900 can be Internet-connected all the time, though, could lessen
the other `N-series GPS' failings to the point where use-Hz increases
and eliminates the `slow TTFF due to expired data' problem. Heck, if the
`also being a phone' part means that it gets plugged into a wall-charger
more frequently, that could also contribute to a solution--never mind
being able to download assists from outside the scope of GPS per se.
Assuming that this is your problem, of course....

-- 
"Don't be afraid to ask (λf.((λx.xx) (λr.f(rr))))."



More information about the gnhlug-discuss mailing list