Re: wireless geolocation

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 7 June 2017 at 23:27, Nico Williams <nico@xxxxxxxxxxxxxxxx> wrote:
> On Wed, Jun 07, 2017 at 10:09:39PM -0000, John Levine wrote:
>> In article <efebcd89-289d-2a3e-9246-6321c07bd255@xxxxxx> you write:
>> >So to sum up:
>> >
>> >"Here is a problem: ... "
>> >
>> >"No its not and even if it was, ...
>>
>> I'd say that summary is not even wrong.  Our problem is not technical.
>> Geolocation is handled by a handful of companies that have no
>> incentive to solve our problem.
>>
>> How many wifi networks are there in the world?  I expect upward of a
>> hundred million.
>>
>> How many wifi networks regularly move from one continent to another?
>> You can probably count them on your fingers.  We're not even a
>> rounding error, and we're not paying customers.  Why should Maxmind et
>> al. spend money to build a mechanism to deal with us and networks like
>> ours?
>
> See, now, this is why we need to put geoloc information into DNS and/or
> BGP, thus cutting out the secret sauce makers for all cases where
> services don't mind trusting DNS/BGP geoloc data.
>
> I mean that half-facetiously, half-seriously.
>

I'll take you seriously, if I may.

First, I'd note that several WiFi networks move around on one
continent on a frequent basis, and many fly through the air between
continents. While a network that is physically bundled into boxes and
moved might be an edge-case, there are an ever increasing number of
access points on the move these days.

Given that the geoloc problem fits into one of two cases, I think you
have half the issue sorted there - a DNS record in the reverse-lookup
domain would satisfy the case where a service wants to go from only an
IP address to a location.

But much of the discussion revolved around a device complicitly
sharing its location based off the BSSID; in this case a DHCP
extension would surely be useful? We're working on the assumption,
here, that the device - and therefore hopefully its owner - is
deliberately providing a location, after all.

On the assumption that administrators of access points are often not
going to set this up, it might also be useful to have an extension
which simply indicated whether the network was fixed or mobile.

Dave.




[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]