On Thu, Aug 9, 2012 at 7:51 AM, Andrew Sullivan <[email protected]> wrote:
> On Thu, Aug 09, 2012 at 08:17:22AM +0100, Brian E Carpenter wrote:
>>
>> ALQDN - ambiguous, or potentially ambiguous, names (like printer.local)
>
> That doesn't need a new name.  We have a name for that: "relative".
> See RFC 1034 section 3.1.
>
>> ULQDN - unambiguous, or almost certainly unambiguous, names (like 
>> printer.<gibberish>.sitelocal)
>>
>
Change the leftmost '.' to '-' and I'm with you.

> That example is strictly speaking a relative name, too, since it isn't
> qualified by the root label.
>
> But my point upthread was not about bikeshedding the things we call
> this, but to observe that there are different naming technologies (we
> concentrated on DNS and mDNS names, but we also have NetBIOS and llmnr
> names.  The latter is very similar to but not exactly the same as
> mDNS.  We can only wish that NetBIOS names are all gone).  I think the
> impulse to try to get these to work together from the user's
> perspective is right, but for the purposes of protocol development we
> need to attend to the fact that they're all just different
> technologies.
>
We all agree we shouldn't call them FQDNs, so that's progress.

The distinction I was trying to draw was not which protocol is used
for resolution, nor the degree of ambiguity in the name, but that these
names are defined within a local scope (e.g. without recourse to
recursive DNS lookups outside the site boundary).  I feel that names
such as those in RFC 6303 would also qualify by my definition, so
maybe we should call them Locally Served DNS Names or similar.

I also want to draw attention to terminology Erik Kline used in his
discussion of http://tools.ietf.org/html/draft-kline-default-perimeter,
the notion of "inside the home" versus "The Internet".  I think this is
a very useful model to enforce with users as it can be applied
equivalently to naming, discovery, addressing, and security.

>> I will be unhappy if ALQDNs are allowed except as legacy.
>
That's exactly the point.  mDNS isn't going away, so let's fix the issues
and leverage it as a basic capability.

-K-

> Me too, but since they're built right into RFC 1034 (which is still
> STD 13), I think we're dreaming.  We might as well wish that the world
> be freed from NAT.
>
> Best,
>
> A
>
> --
> Andrew Sullivan
> [email protected]
> _______________________________________________
> homenet mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/homenet
_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to