user-relative names - was:[Re: Yahoo and IPv6]

Steven Bellovin smb at
Wed May 18 21:10:39 UTC 2011

On May 17, 2011, at 10:30 13PM, Joel Jaeggli wrote:

> On May 17, 2011, at 6:09 PM, Scott Weeks wrote:
>> --- joelja at wrote:
>> From: Joel Jaeggli <joelja at>
>> On May 17, 2011, at 4:30 PM, Scott Brim wrote:
>>> On May 17, 2011 6:26 PM, <Valdis.Kletnieks at> wrote:
>>>> On Tue, 17 May 2011 15:04:19 PDT, Scott Weeks said:
>>>>> What about privacy concerns
>>>> "Privacy is dead.  Get used to it." -- Scott McNeely
>>> Forget that attitude, Valdis. Just because privacy is blown at one level
>>> doesn't mean you give it away at every other one. We establish the framework
>>> for recovering privacy and make progress step by step, wherever we can.
>>> Someday we'll get it all back under control.
>> if you put something in the dns you do so because you want to discovered. scoping the nameservers such that they only express certain certain resource records to queriers in a particular scope is fairly straight forward.
>> --------------------------------------------------------
>> The article was not about DNS.  It was about "Persistent Personal Names for Globally Connected Mobile Devices" where "Users normally create personal names by introducing devices locally, on a common WiFi network for example. Once created, these names remain persistently bound to their targets as devices move. Personal names are intended to supplement and not replace global DNS names."  
> you mean like mac addresses? those have a tendency to follow you around in ipv6...
This is why RFC 3041 (replaced by 4941) was written, 10+ years ago.  The problem
is that it's not enabled by default on many (possibly all) platforms, so I
have to have

# cat /etc/sysctl.conf

set on my Mac.

		--Steve Bellovin,

More information about the NANOG mailing list