[FSFLA] Fwd: wireless geolocation

[willi uebelherr]

Dear friends,

based on this discussion in the "IETF discussion list" you can see, that 
Google is working for geo-location IP addresses. Not in a strong form, 
clear. But more and more they understand this big nonsense with the 
virtual addressing in telecommunication.

Read the "draft-google-self-published-geofeeds-02". The link is part of 
this email. I have used only the first 3 emails from this discussion thread.

many greetings, willi
Asuncion, Paraguay


-------- Weitergeleitete Nachricht --------
Betreff: Re: wireless geolocation
Datum: Tue, 6 Jun 2017 12:48:04 -0400
Von: Warren Kumari <warren at kumari.net>
An: Ted Lemon <mellon at fugue.com>
Kopie (CC): IETF Rinse Repeat <ietf at ietf.org>

On Tue, Jun 6, 2017 at 12:35 PM, Ted Lemon <mellon at fugue.com> wrote:
> Why do we even have a stable IP address range anymore?   It seems like an unnecessary lift.   If we just take what the ISP offers (IPv4 and IPv6) and use that, why is that not good enough?

Because we use a *large* amount of address space - we (currently) give
everyone a public IP address, we have multiple SSIDs / networks, etc.
ISPs would be quite unlikely to be willing to give us a big enough
block, we (often) also have multiple providers, etc. We would also
need to renumber all of our infrastructure, redo DNS, etc.

We currently publish geo-location info in
https://tools.ietf.org/html/draft-google-self-published-geofeeds-02
format -- noc.ietf.org/geo/google.csv. This gets updated (assuming I /
we don't forget :-)) before each meeting, and should be imported by
google in advance of the meeting...

But, much of the issue is location being tied to the MAC address of
the APs, not just the source IP. There are lots of geo providers, and
they all need to be updated, etc.

W


-------- Weitergeleitete Nachricht --------
Betreff: Re: wireless geolocation
Datum: Tue, 6 Jun 2017 12:35:30 -0400
Von: Ted Lemon <mellon at fugue.com>
An: Randy Bush <randy at psg.com>
Kopie (CC): IETF Rinse Repeat <ietf at ietf.org>

Why do we even have a stable IP address range anymore?   It seems like 
an unnecessary lift.   If we just take what the ISP offers (IPv4 and 
IPv6) and use that, why is that not good enough?


-------- Weitergeleitete Nachricht --------
Betreff: wireless geolocation
Datum: Tue, 06 Jun 2017 21:36:07 +0600
Von: Randy Bush <randy at psg.com>
An: IETF Rinse Repeat <ietf at ietf.org>

you may have noticed, when you get to an ietf meeting and get on the
wireless, you will often geolocate to the last city on the eternal tour.
this is because access point bssids are recorded by a number of
geolocation providers and those location data are then used by
application providers.

the noc tries to deal with this, but it is a manual mess involving
trying to convince geoloc providers to do complex things such as answer
their frelling email.  the pain is not congruent with the fix.

the ietf is far from the only event with this problem.  but it is kind
of in our purview to automate it.  the noc is not the place to work on
protocols and databases; the noc is just the customer/user.  a bof or
some existing wg might be.

oh, and warning.  there is a significant chance we will see the issue in
praha as we have been unable to get a response from one major geoloc
provider (not google; they were wonderfully responsive).

randy