4 part domain names

[Peter Risdon]

Dick Davies wrote:
> * Peter Risdon <peter at circlesquared.com> [1140 15:40]:
> 
>>Hexren wrote:
>>
>>>JM> On Wed, Nov 24, 2004 at 04:08:06PM +0100, Hexren wrote:
>>>JM> : location. 510 could identify a rack or a datacenter so that
>>>JM> : us.510.mail.example.com means "a mail server in the datecenter with
>>>JM> : the id 510 which serves the United States".
>>>
>>>JM> So 'us.510.mail' is an atomic, arbitrary identifier.  All three as a 
>>>unit
>>>JM> identify a certain node, and are selected purely for convenience of 
>>>human
>>>JM> operators, right?
>>>
>>>I would say yes.
>>>
>>>
>>>JM> I'm just making sure that the network doesn't treat 'us.510.mail' any
>>>JM> different than it would treat 'foobar', right?
>>>
>>>I would say yes too.
>>
>>
>>How does this square with the fact, as I understand it, that I can 
>>delegate authority for mail.example.com to new nameservers which can 
>>then publish host information about this zone?
> 
> 
> That's got nothing to do with the network.
> For example, I can create a host in example.com called
>  
> us.510.mail
> 
> and you can't stop me (evil laughter).
> 
> 

Sent the RFC mail prematurely...

RFC 952 says:

<quote> A "name" (Net, Host, Gateway, or Domain name) is a text string 
up to 24 characters drawn from the alphabet (A-Z), digits (0-9), minus 
sign (-), and period (.). Note that periods are only allowed when they 
serve to delimit components of "domain style names". (See RFC-921, 
"Domain Name System Implementation Schedule", for background).
</quote>


So I guess you could, but it wouldn't be canonical. If authority ever 
gets delegated for mail.example.com, then for 510.mail.example.com, then 
a host called us is published, there's going to be a bit of a problem 
with your network.

Peter.

-- 

the circle squared

network systems and software

http://www.circlesquared.com