[PLUG-TALK] Static IP address geolocation

[Rich Shepard]

On Sat, 3 Apr 2021, Russell Senior wrote:

> There is an amusing story about how swat teams and warrant servers keep
> showing up at some farm in the middle of nowhere Kansas or something,
> because that's the geographic center of the contiguous US, and because
> that's the default location for IP addresses that don't have a more
> specific geolocation.

Russell,

There were also the stories of a couple in north-central Nevada whose
vehicle GPS sent them on a forest road until they were completely lost. The
husband went for help; his body was found a year later by a couple of
hunters. The wife survived.

A similar event was the family traveling across the Coast Range one winter
whose vehicle GPS took them into a Forest Service road that was supposed to
be closed off by a gate. This time the father walked in the correct
direction and two days later all were rescued; hungray and cold but
otherwise okay.

Closer to home Google Earch will put a pointer within a few hundred meters
of the actual location. GPS receivers now have much higher accuracy than
that; they've improved tremendously since the mid-1990s when 5m was as good
as affordable systems were.

> Geolocating IPs is something for which people find 70% accuracy perfectly
> acceptable, because it is better (for them) than <70%. Personally (echoing
> Tomas) I consider some geographical ambiguity a nice feature. Of course, I
> want to switch everyone to UTC and do away with timezone proliferation.
> Your mileage may vary, but your influence on ip geolocation practices is
> very weak.

Today's GPS technology should allow higher accuracy.

Regards,

Rich