[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Feedback on draft-nottingham-http-link-header-04.txt
Eran, Mark, here is my feedback on: http://www.ietf.org/internet-drafts/draft-nottingham-http-link-header-04.txt A) I very much like the term "context IRI" in section 3 - I think that's a great name for it. (On that score, it makes me think that the best name for what's now /host-meta is actually /context-meta, but I'll save that for feedback on that spec.) B) This is purely an editorial suggestion. You say: A link can be viewed as a statement of the form "(context IRI) has a (relation type) resource at (target IRI)." Given that the context URI, relation type (string), and target URI essentially form an RDF triple (the only difference being that the relation type may not be an absolute URI), I was thinking: A link can be viewed as a statement of the form "(context IRI) has a relation of (relation type) to the resource at (target IRI)." C) In section 4: As such, relation types are not format-specific, and MUST NOT specify a particular format or media type that they are to be used with. Likewise, the context IRI for a given link is usually determined by the serialisation of the link (e.g., the Link header, a HTML document, etc.); a relation type SHOULD NOT specify the context IRI. This confuses me because I can't figure our how the relation type even _could_ specify the context URI. It's like saying an RDF predicate should not be the RDF subject. It's nonsensical to begin with -- unless I'm missing something (which I probably am). D) In section 5: Each link-value conveys one target IRI as a URI-Reference (after conversion, if necessary) inside angle brackets ("<>"). If the URI- Reference is relative, it MUST be resolved as per [RFC3986]. Note that any base IRI from the body's content is not applied. This para is too dense for me to clearly understand (or maybe I'm just being too dense). It says, "...after conversion, if necessary..." -- which I assume means conversion of the IRI to a URI-Reference -- however it's unclear when such conversion is necessary. Why not just spell it out (see also my final comment below about IRI ==> URI transformation in general). Secondly, plenty of technical folks I know don't understand the concept of "resolving" a relative URI-Reference - that could use another sentence of explanation. The third sentence, "Note that any base IRI from the body's content is not applied" also needs more explanation as I'm not even sure what its referring to. (Again, maybe I'm dense.) E) Second 5: Normally, the relation type of a link is conveyed in the "rel" parameter's value. The "rev" parameter has also been used for this purpose historically by some formats, and is included here for compatibility with those uses, but its use is not encouraged nor defined by this specification. I know this whole "rev" thing is a huge can of worms, but don't you think it deserves a little more explanation as to why it is confusing? Or at least a reference to the slightly longer explanation you provide in Appendix A? And maybe this deserves a SHOULD NOT rather than just "...is not encouraged..."? F) Section 5: Note that link-values may convey multiple links between the same target and context IRIs; for example Link: <http://example.org/>; rel=index; rel="start http://example.net/relation/other" Here, the link to "http://example.org/" has the registered relation types "index" and "start", and the extension relation type "http://example.net/relation/other". This particular example seems unnecessarily tricky if it is going to be the only example, i.e., two rels with the second one having two values. I would strongly recommend inserting a new section between 5 and 6 that is only examples. Include 5-6 examples, starting with several very simple and straightforward (the 80% case) and then including a few more advanced examples such as the above to illustrate the extreme cases. G) Appendix A: In HTML4, the link element can be mapped to links as specified here by using the "href" attribute for the target URI, and "rel" to convey both the relation type, as in the Link header. The context of the link is the URI associated with the entire HTML document. The word "both" before "the relation type" does not make sense - something appears to be missing. H) Overall I have found that the importance of the relationship and transformation rules between IRIs and URIs are not well understood even by many of whom you would expect it. So I would recommend this aspect of the spec be beefed up, i.e., perhaps even given its own subsection. Hope this helps, =Drummond
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]