North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: MEDIA: ICANN rejects .xxx domain

  • From: Fred Baker
  • Date: Fri May 12 03:11:07 2006
  • Authentication-results:; [email protected]; dkim=pass (sig from verified; );
  • Dkim-signature: a=rsa-sha1; q=dns; l=2198; t=1147417756; x=1148281756;c=relaxed/simple; s=sjdkim4001; h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;; [email protected]; z=From:Fred=20Baker=20<[email protected]>|;; b=atqgMNxKdbzS7xXgLfoTT9jJ64X6+DGSgtgZMUy9Mibj9ODdxHsL/PF4E6a+ZZ0gAWx7AIeHae119nk13/c5qNawn2T+OmbL+i7C8HQ2HNR+m+h53ypnwcP4aNZeE5lI;

On May 11, 2006, at 11:28 PM, Martin Hannigan wrote:
Im having an offline discussion with a list member and I'll ask, why does it matter if you have a domain name if a directory can hold everything you need to know about them via key words and ip- addrs, NAT's and all?
I think there is a place for that discussion; a directory would allow for containment, which might allow the same character string to be used as a name by different groups if they have sufficiently low probability of needing to communicate. There are other ways to handle this as well. You might google some out-dated drafts by John Klensin that mention such a concept.

As someone else mentioned, there is this authority thing, though. So who manages this name directory? If there is a directory managed by a central agency of some sort that in turn hands LDAP queries (or whatever) off to local instances of directories managed by companies, how does that differ (apart from the use of a different transport) from what DNS does today? Is that central directory-managing authority someone we have to collectively agree to, and how do we do that? How do changes in that directory get made? And if there is no central directory, then basically we have the size and complexity of the .com, .net, .org, and other large namespaces to contend with - just how do we determine that www.renesys translates to and not to How do we distribute that information, and assure ourselves that it got distributed correctly?

I'm not saying it is impossible, or even difficult. I am, however, pointing out that the job DNS does today would have to be done in the new regime, and would have to be done at least as well, and would be fairly likely to have many of the same characteristics, at least when taken in the large.

Now, as to ccTLDs vs gTLDs, if anyone wants to eliminate one or the other they get my vote. I think that gTLDs mostly create a mess, and if I were King they would have been eliminated a long time ago. But that is the opinion of one person, and is probably worth what you paid to receive it.