Knowing more information would be helpful. I just tested the Mozilla database against google maps (sent 2 APs) and got back a radius of 40km centred on the city (I assume this to be using the IP then). I would have thought the 2 APs should have been stumbled. Is it possible to add a way to request additional [debugging] information to the request? Or maybe to include what was used (WiFi, Cell, IP) along with the response?
On 2014-03-17 11:03, Robert Kaiser wrote: > Hanno Schlichting schrieb: >> As of Friday we have added GeoIP as an additional source of location >> information. This is based on a city-level commercial database. The city >> level database doesn’t contain a range or accuracy field, so we currently >> hardcode the value to 40km as an estimate of a typical upper bound for a >> city size. We are looking at ways to improve that, as the actual result is >> sometimes only accurate to the country level and should report a larger >> accuracy value in those cases. > > Hrm, yes, 40km is optimistic for most of such data, it's too bad we > don't have accuracy data there as it can vary really strongly. > >> Right now the accuracy field is also hardcoded to 500m for WiFi-based >> results and 35km for cell-based results. We are working on improving those >> numbers as well. > > There at least we should have some chance to estimate accuracy well from > our data. Of course, I have no idea how that would be done in actual > code or formulae, but it should be possible. :) > > KaiRo > _______________________________________________ dev-geolocation mailing list [email protected] https://lists.mozilla.org/listinfo/dev-geolocation
