- From: Paul Derbyshire <derbyshire@globalserve.net>
- Date: Mon, 22 May 2000 14:41:01 -0400
- To: www-amaya@w3.org
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