That's really neat to know. I really like that idea of defining a static service, similar to DNS, for looking up a user's "transient web" data. Do you have a link to your paper? I'd love to read it.