April 30, 2003

The death of the "Global Uniform Name Space"

Remember the "Global Uniform Name Space" that ICANN's ex-president made such a noise about in his unilateral, never approved ICANN policy document, ICP-3?

I've always considered that policy to be brain-dead and detached from reality.  (See "Thoughts on Internet Naming Systems - http://www.cavebear.com/archive/rw/nrc_presentation_july_11_2001.ppt)"

I'm here at Networld+Interop and I noticed that there are products on the show floor that happily rewrite domain names and URL's - sort of like NAT but at the DNS and URL layer.

If these products fly - and some of 'em do look useful - then DNS names and URL's will be context sensitive just like IP addresses are context sensitive when used in the presence of NATs.  And I expect that context sensitive DNS/URL names will harm the connectivity of the internet no more than do context sensitive IP addresses - i.e. not much.  (Web servers and content management systems have been playing with URL's for years - these new products go much further.)

Despite great wailing, gnashing of teeth, and rending of garments, it is quite clear that the notion that there is a global uniform name space on the internet, if it ever existed, is now dying, if not utterly dead.

Posted by karl at April 30, 2003 11:16 AM