Try going outside. Lock times seem really shaky, sometimes under 15 seconds, other time up to 5 minutes. But its not as bad as before, wit the huge 15 minute waits and stuff. Oh, this is with or without an active internet connection.

Obviously with WLAN or a phone, it picks up the Sat positions quickly, but I have found that this only slightly improves the lock time. I believe this is because the stupid GUI for the AGPS app lets you pick a point on an extremely zoomed out map, making its accuracy very bad on a cold start.

Same setup as yesterday and I just connected in under 5 minutes. I did shutdown for the night so maybe the reboot helped. Maybe something to do with the daemon? (I've got cloudy skies this morning as opposed to last night's clear ones...)

I just blew off some dust from my 770, turned on and tethered to my BT GPS (also from Geeks.com), by the time I lifted my stylus from the screen after entering BT pairing code, the thing already locked to the satellite, split second, not even under 5 seconds! Oh yah, it is a 770. I really do not know what to do with this expensive N810 ****** gps.....

bun

I was sat in a 737 the other day, my N810 got a cold lock (I had previously used gps-ui near that location, but no list of satellites was shown - does that mean cold?).

Anyway, I got a lock in ~5min with the N810 by the window, then it dropped out, and I couldn't get a lock again (even though the satellites were all still shown). Is this a signal strength problem - my SNRs (I presume that's what's displayed above each satellite) were ~26 and lower.

Does anyone have some info on required SNRs? This is presumably down to the antenna quality, but the internal processing will also have an effect with a fixed (low) SNR to work from.

Anyway, it was just painful to sit there with lots of satellites in view (between 9 and 11) but still not be able to get a lock.

Same was true when we were flying, but then it is doubtless going to be more difficult to get a lock when you're not stationary (but ~ 35k' and ~500mph) (still less than the ITAR limits of 1000 knots & 60k' though, so it ought to work eventually - I tried for 45min)

I switched on my Holux M1200 in a 747 at intercontinental altitude and speed.. it took a few minutes to lock but after that I had a good signal and solid tracking (10-11 satellites). The ground speed agreed with the on-board flight display but the altitude was a bit off (as it always is).

As part of my job I write reviews of GPS-capable Nokia phones, and they are all much faster than the N810. The N810 must have the worst built-in GPS of any device Nokia has ever made.

The worst, not that worst, little worst, maybe worst.........cannot help but wondering what is the plan? This is the top most bug and the most talked about, and exactly like that, just most talked about.

Anybody know what Nokia has in mind? AA-gps? AAA-gps?

bun

Not yet, but this is a reasonable request. There's some talk of this (that is having a way to interface with the supld daemon) in bug #2878. This may need a separate bug specifically asking for information about how to talk to the SUPL server (so that we could use a GeoClue backend to provide the location info, rather than a coarse map). Manual lat/long or post/zip-codes are one of the GeoClue methods.

From some quick tests, if the satellite chipset is up and running and spitting out NMEA data (i.e. I'm using maemo-mapper to look at the satellite strengths), using agps-ui either doesn't work, or is delayed until you restart the chipset (i.e. close maemo-mapper, wait for the satellite symbol on the top bar to disappear, then restart maemo-mapper and wait for it to bring the satellite chipset back up).

I should add that using agps-ui while the chipset is up is variable in what happens when you restart; it always goes to no satellites displayed when you restart, but will either sit there and find a real satellite or will eventually reload the new almanac data and say that is has the satellites in view.

Sounds like a bug, but as I'm not quite sure how the agps-ui should be used I'm not too sure.