RE: URNs?

Sounds interesting so far. I guess though you'll need to break the names
into different formats, identified first by a code specifying the naming
authority. So ISBN: will adapt to refer to the naming authority for book
ISBNs, and so on. Naming authorities for Internet resources will presumably
offer name-servers to turn URNs into URLs.

There is an issue though. Either there will be a monopolistic naming
authority for Internet resources (yuck), or browsers and other network
software will keep needing to be upgraded every time there's a new naming
authority or one changes URL (yuck), or Internet resource URNs will wind up
taking the form

URN:authorityURI:restOfURN

with authorityURI a URI for the naming authority that a browser can look
up. If it is a URN in turn, it must use a different naming authority. This
requires there be a fixed "root" naming authority identified with a URL
that must be persistent, and they will then have some sort of monopoly
because they'll be able to dictate terms to the other naming services --
and dictate prices. Plus the URNs would be very unwieldy even for simple
resources. And everyone will want in on the ground floor -- that is, the
root authority will be preferable to first-tier ones, and so forth. Ugly
indeed.

I see one possible solution:

1. Similarly to the DNS (or even worked into the existing DNS framework), URNs
   are resolved by a person's ISP's URN name server, which punts upstream
   toward an eventual root, and these guys are responsible for transparently
   updating to include new naming authorities. Users need not upgrade browsers
   incessantly, and URNs can be short and to the point. Drawbacks: Internic or
   some other monopoly at the root.
2. Pull the teeth of the inevitable monopoly. That is, a not-for-profit
   organization will take on the job of root name authority, and will provide
   their services gratis, possibly using advertising to recoup operation costs
   and provide worker salaries. Or operate on a volunteer basis so they need
   only recoup operating costs.

For item 2, I can think of no better organization to become root naming
authority than the W3C itself, which has shown itself on numerous occasions
to be neither greedy nor short-sighted, and instead, forward-thinking and
humanistic. Also, the large W3C Web site with its oft-referred-to
Recommendations and other specifications is worth a fortune in advertising
space, if a few banners were deployed discreetly at the tops of some of the
more oft-accessed pages. Not to mention money can be made selling books
about the W3C, the Web's internals, and explaining the various
specifications to the casual reader... books linked to a free and
widely-used computer product are always lucrative; witness how much money
Sun's made selling the Java Tutorial books even though there are free
online copies! :-)

-- 
   .*.  "Clouds are not spheres, mountains are not cones, coastlines are not
-()  <  circles, and bark is not smooth, nor does lightning travel in a
   `*'  straight line."    -------------------------------------------------
        -- B. Mandelbrot  |http://surf.to/pgd.net derbyshire@globalserve.net
_____________________ ____|________                          Paul Derbyshire
Programmer & Humanist|ICQ: 10423848|

Received on Monday, 22 May 2000 14:39:12 UTC