I'm not familiar with the Moko's implementation, be it hard or software. Possibly it was using something that most dedicated GPS hardware uses today: offline A-GPS. To this day I wonder why the N900 doesn't have this implemented. Probably something to do with the 'always on' mentality of the designers and developers.

Hmm. I simply cannot confirm such problems -- at least using GPSJinni, which just invokes the location API as well -- unless conditions are really bad, in which case it simply is not surprising. And you'd be surprised how often conditions are bad. And I am not talking weather (which, contrary to popular belief, doesn't have a huge impact on GPS) but satellite constellations and obstruction of reception due to buildings and terrain. Using something like Trimble's Planning Software may give you some idea.

(Incidentally, something like that piece of software would be very nice to have on the N900, come to think of it -- and it actually might provide the needed approximation data for an offline A-GPS. Of course I currently have _no_ idea if such data could even be fed to the GPS unit via location API or any other way short of hacking the cellmodem firmware. What a stroke of genius by the guy who decided to hook it up there instead of directly to the main processor's I/O. )

Oh, I'm not disputing that at all. An application using GPS that essentially folds after 30 seconds without a fix is ridiculous -- and that's putting it mildly. I would use stronger words, but that would go against my upbringing.

Acknowledged. But the thread had (once again, I might add, as this has happened in other threads before) drifted off into wild speculation about the N900's GPS in general, so I thought I might as well pick up on that. Mea culpa!

Regards,
Chris.