Re: IPv6 Destination Option Requested for IPv6GEO

[Brian Haberman <[e-mail]>]

All,
     I am making these statements with my AD hat off...

We have been here before.  Several times.  I am quite sure that the
issues raised during the previous discussions have not (and cannot) be
addressed.  The history of trying to add geographic information to the
network layer includes:

1. VANET-related research
(http://www.researchgate.net/publication/224108335_Geographical_information_extension_for_IPv6_Application_to_VANET)

2. Tony Hain's geo-based IPv6 addressing scheme
(http://tools.ietf.org/html/draft-hain-ipv6-geo-addr-02)

3. A different proposal to add geographic information to the IPv6 header
(http://datatracker.ietf.org/doc/draft-add-location-to-ipv6-header/)

4. A patent on carrying geographic information in an IPv6 extension
header (https://www.google.com/patents/US7330726).

I am sure I have forgotten other instances, but that is beside the
point.  In this day and age, there needs to be strong justification to
add this type of information to the network layer, especially in light
of RFC 7258.

Regards,
Brian

On 9/16/14 7:00 AM, Bless, Roland (TM) wrote:
> Hi,
> 
> Am 16.09.2014 um 00:03 schrieb Ronald Bonica:
>> Fred,
>>
>> Why does this information belong in the IP layer, as opposed to the transport or application layers?
> 
> This was also my question when I read this. Currently, I don't see any
> reason that this information is actually useful at the IP / routing layer.
> 
>>> Our reason is this. In our industry, there are many examples of nodes that
>>> willingly submit themselves for active surveillance. If the last thing ever heard
>>> from such a node is a single isolated IPv6 packet or fragment, the surveillance
>>> authority will want to know where it came from.
> 
> Without having read the draft I don't understand why an application
> layer protocol couldn't provide this in the very first bits if this GEO
> information is so important for you. Fragmentation by the IP layer
> should be avoided anyway and better be done at the application layer
> (following the ALF principle, or maybe at the transport layer).
> Moreover, I think that the GEO location information - even if
> transported via an IP extension header - will _not_ be evaluated within
> the IP layer, but maybe at the application layer.
> So IMHO it is not good to couple such things that belong to normally
> totally disjunct functions (at different layers), and which
> are separated for a good reason.
> 
> Regards,
>  Roland
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
>