GeoIP2 for retrieving city and region ?

[Νικόλας]

Στις 13/7/2013 9:19 μμ, ο/η Roy Smith έγραψε:
> In article <krs3jj$vnq$1 at news.grnet.gr>,
>   ÉΪÉ«É»όλας <nikos at superhost.gr> wrote:
>
>> But then how do you explain the fact that
>> http://www.maxmind.com/en/geoip_demo pinpointed Thessalon£ki and not
>> Athens and for 2 friends of mine that use the same ISP as me but live
>> in different cities also accurately identified their locations too?
>
> I just tried 24.136.109.105 on that demo.  It comes up with:
>
> US Englewood Cliffs,
> New Jersey,
> United States,
> North America
> 07632 40.8915,
> -73.9471
> Time Warner Cable
> Time Warner Cable
> rr.com
> 501
>
> Not bad.  Google street view shows that as a very pretty residential
> neighborhood in one of New York's fancier suburbs.  Unfortunately, it
> happens to be in run-down industrial building in a factory district
> about 20 km away.
>
> There are lots of interesting (and superior) ways to do geolocation
> other than looking up IP addresses.  Here's a few:
>
> 1) GPS.  Obviously, if you're on a device that has a GPS receiver and
> you have access to that data, and you've got a good signal, nothing is
> going to beat GPS.  Well, other than Glonass.  And Galileo and IRNSS,
> whenever they become operational.  And whatever the Chinese are calling
> theirs.
>
> 2) Cell (i.e. mobile) phone tower triangulation.  The phone systems know
> where all the towers are and know which towers your phone is receiving
> signal from.  Since they know the signal strengths from each of those
> towers, they can do a rough triangulation.  It's kind of messy since
> signal propagation depends terrain and obstructions which aren't well
> mapped.  But it's better than nothing.
>
> 3) WiFi triangulation.  Right now, I can see four WiFi networks (Worb,
> J24, MusicWiFi, and jcglinksys).  There are databases of WiFi network
> names and approximate locations (obtained by wardriving and other ways).
> If you can see enough networks, it's easy to look in the database and
> figure out where you must be.
>
> 4) Who knows what the future will bring.  I suppose some day, inertial
> nav will become cheap enough that we'll all be walking around with INS
> in our phones.
>
> In general, mobile operating systems control direct access to all of
> these signals and only allow applications to get the location data when
> the user agrees to such access.
>
Do you know a way of implementing anyone of these methods to a script?

-- 
What is now proved was at first only imagined!