- From: Bijan Parsia <bparsia@isr.umd.edu>
- Date: Fri, 5 Mar 2004 23:19:32 +0100
- To: public-webarch-comments@w3.org
I don't see how to usefully split the discussion of this section, as the first line of 2.2 refers to 2.3. """The requirement for URIs to be unambiguous demands that different agents do not assign the same URI to different resources.""" But it's ok that the same agent does so? Sorry if that sounds snarky, but it is a genuine question. I would have thought that the primary case was the resource owner...thus far, in the document, no one else seems to have the power to *assign* the same URI to different resources. E.g., principle: URI Assignment, the first line of 2.1, Good Practice: URI aliases. (Hmm. In 2.1, right after Good Practice: URI Aliases, we get the notion of URI Producers. Who are they? Frankly, I don't find the arbitrary shifting of terms of art in a technical document to be helpful. None of these terms are defined in the glossary, either. Perhaps this isn't a technical document? If so, I'm going to feel a bit sheepish reading it this closely. Section 1.1.1 suggests that it is meant to be a guide for technical people like me (I do all of 1-4). Hmm. Section 1.1.2 says: """This document strikes a balance between brevity and precision while including illustrative examples.""" Ok, that gives me some standard. I think the above claim doesn't strike the right balance between brevity and precision.) There's another reading, to wit, that the *URI* can be unambiguous (say, because ambiguity of a URI is defined to be having being assigned to more than one resource, and assignment can literally only be done by the URI owner, and the URI *can* successfully only assign it to one resource (none of these are obvious!)) while various agents can *use* it ambiguously. On this reading, the requirment of URI ambiguity does *not* demand that different agents do not *use* the same URI to refer to different resources. (I don't know if you mean "assign" up there in the restricted sense I was using in the definition of ambiguity, or just to mean "refer", or something else.) Section 2.3 says that the ambiguous *use* of URIs is to be avoided (though, I'll point out, that the Good Practice is ambiguous between ambiguous URIs and ambiguous *use* of URIs). Of course, certain ambiguity doesn't matter, e.g., replicating Quine, I might use a URI to refer to me, the human being, and someone else to refer to the collection of undetatched people parts. As long as all our uses *align* in (all) our interactions, we're fine, ambiguous assignment or not. Sorry for the quick digression into philosophy of languages, but, really, at this time of night, I feel a little justified in turn around :) """Hierarchical delegation of authority. This approach, exemplified by the "http" and "mailto" schemes, allows the assignment of a part of URI space to one party, reassignment of a piece of that space to another, and so forth.""" First use of 'URI space', which is undefined. I see 'information space', 'uniform address space', and, of course, 'namespace'. As far as I can tell, only 'namespace' has a definition (and it's not in this doc, which is fine). Perhaps this is only editorial. A URI space seems clear (a set of URIs? why not say that then?), but I did spend some time wondering if it was the same as an infromation space or address space. *Are* you using unambiguous phrases here? Are they aliases? Is there a problem with either defining terms or using only one where there's only one concept? Some principles of the web apply well to technical prose. """Whatever the techniques used, except for the checksum case, the agent has a unique relationship with the URI, called URI ownership. """ Here is what I can find on what's an "agent", prior to this passage: """ Within each of these systems, agents (people and software) """ """strate typical behavior of Web agents — people or software (on behalf of a person, entity, or process) acting on this information space. Software agents include servers, proxies, spiders, browsers, and multimedia players."""" So, an agent is a person or a program. Thus, every http uri has, supposedly, one, and only one, person or program that is its owner. However, institutional ownership seems possible, as is joint ownership. """The social implications of URI ownership are not discussed here. However, the success or failure of these different approaches depends on the extent to which there is consensus in the Internet community on abiding by the defining specifications. """" First you say that the social implications of URI ownership are *not* discussed here, then go on to discuss some social implications. Don't do that. I don't believe the second statement of that quote, at least on many interpretations, and I've objected to its use in various technical arguments, some with TAG members. If this passage is to be a stick to beat me with in technical debate in W3C working groups, then I strenuously object to it, especially without substantial explication and clarification. So, I make the strong comment that I want this line struck. I object to it. Is anything in this document normative? I notice that there is some rejection of adding a conformance section, which is fine, but I have *NO* idea how to use this document in working groups, nor do I know how it may be used by others. I totally fail to see how this can be helpful. So, I would like some guidance about that. Cheers, Bijan Parsia.
Received on Friday, 5 March 2004 17:19:29 UTC