Flaming Mountainside

UN (United Nations) and ITU (International Telecommunications Union) commissioned a WSIS (World Summit on the Information Society) which in turn commissioned a Working Group on Internet Governance (WGIG). Their reports published in July 2005 will be cause for controversy...

The reports I'm referring to are online:

The timing and jockeying/positioning on this has been carefully fierce. This report came out during the Luxembourg ICANN meeting when virtually most of its authors and readers (and more than a few members of the press) will be under one roof. A week or so ago, the US Government announced that it would retain its oversight role of ICANN and of all changes to the DNS "root zone" for the foreseeable future. Is it all just sabre-rattling, or will there be actual hostilities?

Several things stand out from my reading of today's WGIG reports:

No formal relationship with root server operators.: We've all heard this one before many times. The thing that hasn't been explained is: "in either practical or theoretical terms why does this matter?" What harm has ever come from it, or could possibly come from it? How would the world be better if the relationships between the rootops and IANA were formalized? What's the risk and how would formality manage those risks? Speaking as a rootop, I'm ready at any time to memorialize the terms of the current relationship, or to explain them to anyone who asks about them. See also [my previous comments].
Four different proposed possible governance models.: This is probably a smokescreen designed to divert attention from a single proposal (which is to say, #4) while being able to claim, later on, that "all the options were on the table". The national governments who stand behind WSIS seem to have [ become impatient ] with USG control over the Internet and USG has pretty much reneged on its promise to set the Internet free, and I don't see any possibility of WSIS ratifying an ICANN-as-usual next regime. The Internet is just too important, culturally and economically, for the world to tolerate unbounded unilateral control over its development. No matter how well USG's unilateral control might have worked in earlier days, those days are over.
DNSSEC and zone-walking.: WGIG's report points out an IETF gaffe, which was to standardize a Secure DNS solution that was completely unworkable for any number of CCtld's due to national privacy laws and other considerations more or less related to sovereignty. Apparently, IETF should design the Internet's protocols for a larger audience than "whoever shows up at the meeting". This sounds absurd, but it's true. And so, any zone that deploys DNSSEC as specified gives up any subdomain naming privacy they thought they had, because DNSSEC exposes all the information needed to prove the nonexistence of nonexistant names, and that information indirectly highlights all of the existing names. While this lack of naming privacy is commercially uncomfortable for a number of Gtld's, when combined with WHOIS data, it's an [ actual violation of the law ] for some CCtld's. It'll be interesting to see how WSIS proposes to get Internet protocol design to leave the ivory tower.
Multilingualism or Internationalization.: Some of WSIS's concerns are strictly operational -- they want standardized multilingual domain names and they don't seem to think that the Internet's current government (such as it is) cares enough about internationalization of domain names to allow/require ICANN and others to undergo the difficult policy and technical work of admitting these names into the system, including into the root zone.
Peering.: WGIG seems to be concerned about the general lack of interconnection and the anticompetitive cost of entry for new ISPs. One can see the ITU influence here (ITU more or less regulates international telephone service today), but when I think of the way national sovereignty has been abused to turn telecommunications access fees into [ major GDP sources ], I already don't think I'm going to like the endgame if "regulation" occurs in the area of international Internet peering and interconnection.
Silly Errors.: Never fear, the GOV, MIL, and EDU zone files are not actually "part of the root zone file". Never fear, VeriSign does not have a contract to operate "the root servers". Just two root servers, A-root and J-root, are subject to VeriSign's contract with USG.
VeriSign.: The pink elephant in the Internet's living room, the one that nobody usually likes to talk about, goes by the name "VeriSign". This company, through its acquisition of the original Network Solutions years ago, has been able to stretch an old NSF contract for operating the COM and NET zones (which at that time were free for registrants, and very small) into a billion dollar business. Nothing happens in Internet naming without VeriSign's permission (could IDN or DNSSEC succeed if VeriSign doesn't implement them in COM?), and a lot of things that pretty much everybody thinks are bad get done anyway if VeriSign is in the mood to tussle (like domain waitlisting or SiteFinder).
Lack of Treaty.: I guess when we were busy building the Internet we forgot to ask the national governments of the world to get involved. (That could be because they were mostly fixated on the ISO/OSI protocols at the time and couldn't be bothered with our research-toy network, but I digress.) In any case, I'm no expert on treaties. I wish there were some. I don't see how that's going to happen given what's happened with the Kyoto Protocols, but if WSIS knows a way, I wish them only godspeed.
Equitable Distribution of Addresses.: Note: I am not speaking as a member of ARIN's Board of Trustees in this article. "Equitable" is a wild goose, not the holy grail, of IP address allocation. Anyone who needs IPv4 or IPv6 addresses, anywhere in the world, can get them. The policies governing address allocations have been set according to engineering needs not social needs, since there are a lot of ways to allocate addresses that might be somehow "equitable" but would result in a non-working Internet. Give every human being on Earth a /64 network, that's equitable, but then go ahead and try to connect all those networks together, it wouldn't work. But, tell every human who wants a /64 to get one from their ISP, and it'll work just fine.
Possible Period of Governance Reform.: What a lovely way to avoid saying "revolution".
Free/Libre/Open Source Software (FLOSS).: The authors and publishers of these reports could strengthen their recommendations in the area of FLOSS by following them! Why were several of today's documents released as MSWord files rather than OpenDocumentText (ODT) files? The F/L/OSS movement has been studying the freedom in cyberspace problem space for longer than the Internet has had a government (such as it is), and there's a natural alliance, if not outright overlap, to be found between people who want democratic government of cyberspace, and people who write software so they can give it away. So, don't just praise the recipe for this kool-aid -- drink it!

I think a lot of hard work went into this report, and considering the number and strength and diversity of views expressed during the WGIG process, the result has to be called herculean. I'm a bit concerned that it amounts to a generally agreed upon statement that "somebody ought to put a bell on that cat". Turning hegemony into democracy by peaceful means has been done only a few times in human history, and the outlook for this time isn't good.

Paul Vixie
La Honda



Some Comments on Working Group on Internet Governance (WGIG) 21-Jul-2005 (Thu) 18:21 UTC · permalink UN (United Nations) and ITU (International Telecommunications Union) commissioned a WSIS (World Summit on the Information Society) which in turn commissioned a Working Group on Internet Governance (WGIG). Their reports published in July 2005 will be cause for controversy... The reports I'm referring to are online: [Full Report] (PDF) [Background Report] (MSWord) [Press Release] (PDF) The timing and jockeying/positioning on this has been carefully fierce. This report came out during the Luxembourg ICANN meeting when virtually most of its authors and readers (and more than a few members of the press) will be under one roof. A week or so ago, the US Government announced that it would retain its oversight role of ICANN and of all changes to the DNS "root zone" for the foreseeable future. Is it all just sabre-rattling, or will there be actual hostilities? Several things stand out from my reading of today's WGIG reports: No formal relationship with root server operators. We've all heard this one before many times. The thing that hasn't been explained is: "in either practical or theoretical terms why does this matter?" What harm has ever come from it, or could possibly come from it? How would the world be better if the relationships between the rootops and IANA were formalized? What's the risk and how would formality manage those risks? Speaking as a rootop, I'm ready at any time to memorialize the terms of the current relationship, or to explain them to anyone who asks about them. See also [my previous comments]. Four different proposed possible governance models. This is probably a smokescreen designed to divert attention from a single proposal (which is to say, #4) while being able to claim, later on, that "all the options were on the table". The national governments who stand behind WSIS seem to have [ become impatient ] with USG control over the Internet and USG has pretty much reneged on its promise to set the Internet free, and I don't see any possibility of WSIS ratifying an ICANN-as-usual next regime. The Internet is just too important, culturally and economically, for the world to tolerate unbounded unilateral control over its development. No matter how well USG's unilateral control might have worked in earlier days, those days are over. DNSSEC and zone-walking. WGIG's report points out an IETF gaffe, which was to standardize a Secure DNS solution that was completely unworkable for any number of CCtld's due to national privacy laws and other considerations more or less related to sovereignty. Apparently, IETF should design the Internet's protocols for a larger audience than "whoever shows up at the meeting". This sounds absurd, but it's true. And so, any zone that deploys DNSSEC as specified gives up any subdomain naming privacy they thought they had, because DNSSEC exposes all the information needed to prove the nonexistence of nonexistant names, and that information indirectly highlights all of the existing names. While this lack of naming privacy is commercially uncomfortable for a number of Gtld's, when combined with WHOIS data, it's an [ actual violation of the law ] for some CCtld's. It'll be interesting to see how WSIS proposes to get Internet protocol design to leave the ivory tower. Multilingualism or Internationalization. Some of WSIS's concerns are strictly operational -- they want standardized multilingual domain names and they don't seem to think that the Internet's current government (such as it is) cares enough about internationalization of domain names to allow/require ICANN and others to undergo the difficult policy and technical work of admitting these names into the system, including into the root zone. Peering. WGIG seems to be concerned about the general lack of interconnection and the anticompetitive cost of entry for new ISPs. One can see the ITU influence here (ITU more or less regulates international telephone service today), but when I think of the way national sovereignty has been abused to turn telecommunications access fees into [ major GDP sources ], I already don't think I'm going to like the endgame if "regulation" occurs in the area of international Internet peering and interconnection. Silly Errors. Never fear, the GOV, MIL, and EDU zone files are not actually "part of the root zone file". Never fear, VeriSign does not have a contract to operate "the root servers". Just two root servers, A-root and J-root, are subject to VeriSign's contract with USG. VeriSign. The pink elephant in the Internet's living room, the one that nobody usually likes to talk about, goes by the name "VeriSign". This company, through its acquisition of the original Network Solutions years ago, has been able to stretch an old NSF contract for operating the COM and NET zones (which at that time were free for registrants, and very small) into a billion dollar business. Nothing happens in Internet naming without VeriSign's permission (could IDN or DNSSEC succeed if VeriSign doesn't implement them in COM?), and a lot of things that pretty much everybody thinks are bad get done anyway if VeriSign is in the mood to tussle (like domain waitlisting or SiteFinder). Lack of Treaty. I guess when we were busy building the Internet we forgot to ask the national governments of the world to get involved. (That could be because they were mostly fixated on the ISO/OSI protocols at the time and couldn't be bothered with our research-toy network, but I digress.) In any case, I'm no expert on treaties. I wish there were some. I don't see how that's going to happen given what's happened with the Kyoto Protocols, but if WSIS knows a way, I wish them only godspeed. Equitable Distribution of Addresses. Note: I am not speaking as a member of ARIN's Board of Trustees in this article. "Equitable" is a wild goose, not the holy grail, of IP address allocation. Anyone who needs IPv4 or IPv6 addresses, anywhere in the world, can get them. The policies governing address allocations have been set according to engineering needs not social needs, since there are a lot of ways to allocate addresses that might be somehow "equitable" but would result in a non-working Internet. Give every human being on Earth a /64 network, that's equitable, but then go ahead and try to connect all those networks together, it wouldn't work. But, tell every human who wants a /64 to get one from their ISP, and it'll work just fine. Possible Period of Governance Reform. What a lovely way to avoid saying "revolution". Free/Libre/Open Source Software (FLOSS). The authors and publishers of these reports could strengthen their recommendations in the area of FLOSS by following them! Why were several of today's documents released as MSWord files rather than OpenDocumentText (ODT) files? The F/L/OSS movement has been studying the freedom in cyberspace problem space for longer than the Internet has had a government (such as it is), and there's a natural alliance, if not outright overlap, to be found between people who want democratic government of cyberspace, and people who write software so they can give it away. So, don't just praise the recipe for this kool-aid -- drink it! I think a lot of hard work went into this report, and considering the number and strength and diversity of views expressed during the WGIG process, the result has to be called herculean. I'm a bit concerned that it amounts to a generally agreed upon statement that "somebody ought to put a bell on that cat". Turning hegemony into democracy by peaceful means has been done only a few times in human history, and the outlook for this time isn't good. Paul Vixie La Honda		Illegitimibus non carborundum · · · Active Categories: internet (6) · governance (3) · manners (1) · security (1) · technology (1) terra (3) · climate (1) · food (1) · rights (1)