CircleID

I was more than a little surprised to see Paul Vixie's recent vitriolic attack against alternative roots here. Surely there was no need to gloat, that battle had long been won with the growth of alt/complementary roots having more or less stalled since the ICANN November 2000 meeting to remain used only by hobbyists (we're not all scammers Paul), and faux alternatives like new.net never having gained any traction. I wondered if he was concerned about growing faux IDN alternatives like i-dns, but surely they pose no real threat to the security and stability of the internet.

So either Vixie is prescient or he had some forewarning about this USG announcement and was firing an opening shot in what is likely to become a renewed debate regarding one or more viable alternatives to the USG/ICANN root, what the late Peter de Blanc of the ccTLD community called the nuclear option.

Most of the press seems to think that paragraph two of the USG announcement will be seen as reassuring to ccTLDs as it enshrines their right to exist. I don't. There is no legitimate reason that a ccTLD should be co-managed by itself and the USG. If you think the argument that the USG has done it historically and (they say) done it well is a legitimate reason, then you shouldn't have skipped Philosophy 101 to hang out in the computer lab.

The USG is engaging in a game of chicken with the ccTLDs (to say nothing of what is probably a large majority of internet users) and we'll see who blinks. One could wait to see how the root server operators respond but surely they must have had some prior knowledge of this, it would be a political disaster if they just said no. Paul Vixie said here in an exchange with Karl Auerbach that they did have such powers. What say you now Paul? -g

I'm not gloating.  I don't like the current situation at all, and the fact that this had to be a "battle" with winners/losers demonstrates some weaknesses in the design of DNS.  The article you're referring to is not my defense of these weaknesses nor of any/all of the world powers now battling for the soul of the Internet's naming system.

All I'm trying to explain is that DNS has exactly one root, and this design requirement is completely unresponsive to whims.  If you want a naming system that supports multiple roots, or doesn't have a root at all, or whatever, then feel free to try to design such, and get the world to switch over to it.  (Ref: W3C's Semantic Web, et al.)

And FWIW, I am a root nameserver operator and I had no prior knowledge of the recent USG announcement (But, Note Well: anyone who hasn't been expecting USG's announcement for years is having a "same planet, different worlds" problem and might want to review their medications.)

But with or without prior knowledge, ISC as a root nameserver operator wasn't asked, and won't be asked, to take sides in any USG/WSIS battle.  ISOC/IAB define what IANA is and recognize who IANA is, and ISC as a root nameserver operator publishes IANA's data faithfully. USG and WSIS are free to lobby IANA, and ICANN, and ISOC/IAB, but when that dust settles, ISC's f-root server will still be right here, publishing IANA's data.

And if China/Brazil decide on the "nuclear option" (DNS balkanization), I'd expect them to ask ISC to operate one of their new root nameservers—we're well qualified for this kind of work, and as secretariat for DNS-OARC, we'd like to measure their DNS traffic. Obviously this would call for separate infrastructure from our f-root, and there's no way we'd mix IANA and non-IANA data.  But, if the U.N. asks ISC for help, then speaking for the record, ISC will help.