Cart
[Home] [By Thread] [By Date] [Recent Entries]
On Tuesday 13 November 2001 02:10 pm, John Cowan wrote: > Gavin Thomas Nicol wrote: > > Well, we've heard that, depending on the system, the resource identified > > by the URI might not be unique... so I guess we have varying degrees of > > uniqueness. > > URIs and resources are 1-1 and onto: one resource, one URI, and vice > versa, by definition of "resource". > > However, multiple resources can fetch the same *entity body* > (sequence of octets plus a media type), and a given resource may > have multiple entity bodies (via negotation, or varying with time). > Some resources, like namespaces, don't have to have corresponding > entity bodies at all. > > For more on this, see my postings to the xml-uri archive. I've been through all this, and to be honest, I simply find it hard to care enough to be precise when it comes to URI's and HTTP... all the theory flies in the face of everyday reality. In the above scenario, what you're left with is the fact that in order to uniquely, and reliably address a given octet sequence (entity body), you have to provide a URI, a complete set of headers (Accept*), a server, the server configuration, and all of the "entity bodies" as well, and even then, you might not have it. I don't think that's what people expect.... and I'm sure the web services folk will be none too happy about it.
|
