No, not Jesus. It's the spirit of Jon Postel we're invoking today, and while we're at it let's call up St. Jude, patron saint of desperate situations, and Jerry Springer, who needs no introduction.
Postel's the wizard who made possible much of the Net As We Know It. He's the guy who created, or cocreated, most of the technology that makes stuff like e-mail and the Web work. He's the guy you can thank when you click on a link, since he created the "root servers" (basically the big address book listing the location of every domain in the world). He helped set up the very first computer that routed data amongst machines on the old ARPAnet—the Internet's predecessor. He's the wizard architect of the Net's "plumbing." He da man.
He's also dead, and his timing couldn't be worse. Just before he died last year, Postel unveiled his vision for codifying and improving the plumbing. The idea was to move the duty of managing the Net's domain management away from the ragtag bunch of companies and committees that has sprung up over the years and to a multinational entity that represents all the Net's constituencies. That's where St. Jude comes in, and not a moment too soon: The Net is suffering some mighty growth pains.
Take Network Solutions Inc. (NSI), for instance. If you'd told most domain owners two years ago that the company with a near-monopoly on domain registration would ever be the underdog in things Nettish, we'd have laughed. With complaints against them ranging from arbitrary revocation of domain registration to bad service, the leading domain-name registrar has been widely hated for years. Simply put, they got swamped—under their watch (on a contract basically dropped in their lap by the National Science Foundation in 1993), registrations for .coms, .nets, and .orgs went from one or two a day to thousands and thousands.
Postel's proposal recommended a multifaceted body to handle the plumbing associated with domain registration. That was fine by the Department of Commerce, which was getting tired of taking flak for NSI's failings. Enter ICANN, the Internet Corporation for Assigned Names and Numbers. ICANN is supposed to take responsibility for shepherding the aforementioned plumbing into the hyper-connected 21st century. Problem is, ICANN has shown itself less interested in plumbing and more interested in policy issues, such as domain-name trademarks and committee funding. (And jetting to fabulous meeting locales. Why meet via e-mail when you can meet in Berlin in spring?)
Which is where Jerry Springer comes in. Last week, ICANN interim chair Esther Dyson started flinging digital chairs at NSI in a letter to Ralph Nader, who'd sent Dyson a note asking, in effect, what the heck's going on over there. He has reason to wonder: ICANN has decreed a $1-per-domain tax on all current owners, has refused to divulge information on what's going on in meetings, and has otherwise engaged in the kind of actions you don't like to see from an outfit due to take control of the heart of the Net. After all, grassroots Net development got us this far, and centralized Net decision-making doesn't bode well for the semi-anarchic system that made it great.
And when the committee chair, in response to such a note, accuses those who disagree with her of "funding and otherwise encouraging a variety of individuals and entities to throw sand in the gears whenever possible," you've got to ask yourself whether ICANN members are spending too much time jetting around to fabulous locales and too little getting down to the business for which they were gathered.
People are starting to talk. Government watchdogs (including some in the Department of Commerce) would like to know where ICANN, a non-elected body, gets off levying taxes. Rank-and-file domain owners want to know why they're not allowed to participate in the ICANN process while governments and big business are encouraged to do so; they're also uneasy about proposals that would give big businesses special rights to snatch up any domain name they, the big businesses, feel ought to be part of their corporate holdings.
Meanwhile, there's a real problem coming: The info-superhighway is running out of street addresses. The current system assigns every Net machine an IP address. There are 4.2 billion possible addresses, which seemed like a lot once. Not anymore; the burden of matching IP addresses to URLs has root servers working much, much harder than they're designed to, and sooner or later we're going to run out unless swift steps are taken to implement a new standard. Dr. Postel's beloved addressing system is as overburdened as Seattle's city streets. Is ICANN on the case? Or is plumbing less appealing than politics?