> With regards to this, I think we came to agreement on new text for
> the section, something that would look like this:

I'm OK with this.

My comment was really more to do with this:

>   > I saw a privacy comment in the past (sorry, can't source the
>   > original author) that suggested that because of the procedure
>   > for address assignment where only one host allocates addresses
>   > within the prefix that there was no privacy benefit to
>   > regenerating interface identifier portion of the addresses since
>   > (for example) traffic analysis would just be done on prefix
>   > matching.

"no privacy benefit" might be a bit strong. But I would tend to agree
that the benefits are limited if in fact the /64 identifies the
handset.

>   > Does this (paraphrased) assessment seem correct? I wouldn't want
>   > 3GPP to mandate a behaviour that they would believe contributed
>   > to identity privacy but, based on some other procedure, did not.

AFAIK, 3GPP is not mandating the use of temporary addresses.
draft-ietf-ipv6-cellular-host-01.txt makes this pretty clear.

> => But the person tracking would have to know that the host is a
> 3GPP host.

This seems like it might well be straightforward information to determine.

> Otherwise, they won't know that it has a prefix for itself.  BTW,
> this is not specific to 3GPP, the same can be done with home
> networks. A house that gets a /48 can also be tracked by the same
> method.

Agreed. Temporary addresses are just one technique, that has some
benefits in some environments. But there also limitations.

Thomas
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to