This v-week in Packetistan
It is busy, busy, busy in packetistan.
Last week was the 64th IETF in Vancouver. Last week was also the ICANN meeting, also in Vancouver. Last week also "hosted" (temporal pun) a hearing of the Subcommittee on Telecommunications and the Internet on legislation to create a statutory framework for Internet Protocol and Broadband Services. This week its the World Summit on the Information Society in Tunis.
These are all related.
The WSIS is attempting to find a means of jurisdictionalizing a model that has no spatial, or temporal properties. Jon Postel used iso3166 labels, aka "country codes" (but Christmas Island is not a "country"), to delegate create/change/delete operations in a database. He could have used x.121 regions, aka "continents", or technical mechanisms that have no simple, and misleading, analogy. The WSIS is also attempting to allocate resources that pre-existed Jon, the set of all 32-bit network endpoint identifiers, aka "addresses". Superficial issues are symptoms of ajurisdictionalism, Nigerian 419s, phishing, and so on.
One can look at this as an attempt to upset a working neutral apple cart for mere "political interests". One can also look at this as attempting right an apple cart that was tipped on its side some time ago by mere "corporate interests". Note Bene: The ARPA or DARPA Net originally "worked" simply as a testbed survivable communication system for the C3I requirements of a nuclear weapons state committed to winning a global thermonuclear war.
Is a 501(c)(3) in California, which really is managed by Verisign, in the same sense that BigPharma manages most of Congress, on the issues that matter, the only, or the best answer to the question of what to do now that Jon is dead? Its been five years and ICANN is still the laughing stock of the technical community that trusted Jon, and created the model around that distributed, transitive, trust. Is the DoC the only, or the best answer to the question of what to do with ICANN?
I recommend reading the Human Rights Watch take on the Tunis round of the WSIS. Media oligarchies vs all the strengths and weaknesses of nation states.
The STI is attempting to find a means of jurisdictionalizing a model that has no link type or flow temporal properties. In the US, there is a regulatory framework for telecommunications, arising out of the predatory "Universal Service" bill of goods that AT&T sold to state and federal regulators to recover market share lost after the expiry of its patents in 1894, until "dual service" ment AT&T technology, and eventually, AT&T peering and pricing. There is also a regulatory framework for community access television (CATV).
Where a CATV or Baby Bell operator's subscriber (access) network attaches to a carrier or backbone the link type, whether over cable or over telco, packets exiting the access network looses the link type information (link encapsulation is removed). Similarly, packets entering the access network acquire the link type information (link encapsulation is added).
Panel 1 is the lobbists from SBC, Alcatel, Earthlink, XO1, Verizon2, Insight, Nat. Assn. of Broadcasters, Nat. Assn of Telecommunications Officers and Advisors and Microsoft. The "neutrality" that is being discussed is a "neutrality" that gets monopoly lobbiests out in force. Can they legally filter on link type? Can telco and CATV monopolies depricate each others' traffic, recreating walled gardens so beloved by the mobile operator market?
In Maine Verizon+Maine's Independent Telcos, Adelphia+TimeWarner, and (absurdly) Central Maine Power, all want "neutrality" and "broadband" to favor their subscriber access networks, their copper, their executives, shareholders and employees. Independent ISPs and ISPs that service dial-up (price points under $10/subscriber-month) are outside the scope of "neutrality". As goes Maine, so goes the nation.
One of my technical goals for the near future is to make Wampum available in low-bandwidth and high-bandwidth forms, and to finally take a look at Wampum from a visually handicapped acess point of ... view. Blogging has to be more than just laundry exchanged between digital yuptos provisioned with broadband at a $50/subscriber-month price point.
The ICANN meeting was dominated by the Verisign marketing triumph. Its just corruption on the march, and about as unlikely to catch anyone's attention outside of a small number of people than the tax on basic telephony service that funded (in a small part) the Viet Nam War.
IETF meetings are rarely externally significant. The 48th meeting in Pittsburg was a rare exception, one that my trip to Pittsburg last week (an interview, outcome unknown) brought to mind. At IETF-48 we adopted a position on cyphers that was contrary to the position of a nation state. The position we adopted was that "strong encryption" could be specified with "MUST" language. The position of the nation state was that no specification could require anything other than than "weak encryption". That was the meeting where, at least momentarily, the IETF became an International Standards Body. The nation state in question was the United States. The same issue was revisited at IETF-51, when another nation state insisted that only weakly encrypted data enter its jurisdiction. The IETF hasn't been back to London since.
I don't know how to characterize in 25 words or less IETF-64. The last time I was in Vancouver was IETF-18. The guy in the adjacent dorm room was also named "Brunner", from somewhere inside of the Bell system, and I was there on the DOD's nickle, via SRI. One thing is certain however, every meeting since IETF-44 (Minneapolis) has had ICANN Chair and Board members attempting to wrap themselves in the technical legitimacy of the IETF.
If anything surprising comes out of the WSIS I'll blog on it or post a link to someone who has. For the moment the WSIS hasn't decided to set a date to pull authority for the root zone out of suites overlooking beautiful Marina del Rey.
1 In the relatively recent past I've bought OC-3 transit data circuits from XO to the XO OC-192 mesh.
2 In the relatively recent past I've bought DS-3 backhaul (dialup) and OC-3 transit data circuits from Verizon.