- From: Ray Denenberg <rden@loc.gov>
- Date: Wed, 18 Jun 2003 11:57:28 -0400
- To: www-zig@w3.org
Mike Taylor wrote: > I think that determining whether an ESN "takes the form of a URI" is > non-trivial; I wouldn't be comfortable coding it off the top of my > head, for example. No I certainly didn't intend that the server would dynamically make the distinction, or even know the difference, but rather just look up the string in an esn table. What I'm really trying to say is that we can use globally unique identifiers for esns, "globally unique" implies a naming authority, and that naming authority can define that identifier to mean whatever it wants it to mean and publish that definition, and because it's globally unique there is no risk of ambiguity. > Wouldn't it be simpler to say: > > Whenever the XML record syntax is specified, the ESN is > interpreted as a URI. No, for two reasons: (1) we don't want to rule out for example "XML" and "full". Even if we don't know what that means, and we don't want people doing it, I don't want to try to push an implementor agreement that covers this area quite so comprehensively. (2) what would "the ESN is interpreted as a URI" mean, anyway? We 're not going to prescribe dynamic behavior upon encountering a URI. --Ray
Received on Wednesday, 18 June 2003 11:57:30 UTC