- From: <noah_mendelsohn@us.ibm.com>
- Date: Wed, 19 Dec 2007 15:29:04 -0500
- To: "Mike Schinkel" <mikeschinkel@gmail.com>
- Cc: uri@w3.org
> I just spent some time re-reading the long series of email discussions about > httpRange-14 [1]. Wow, you're a patient man! > It seems they addressed at length what a URI points to, but did not address > what does point to a thing when one wants to be able to get an associate > representation about that thing. Not sure I'm parsing this right. I believe the TAG's resolution is: a URI identifies a resource. The "range" of such resources includes information resources, which can be represented with good fidelity in a computer message, and for which a status code 200 is appropriate if you are indeed returning such a representation of the resource. A 303 can be used with any resource, including an information resource, if you have other information to provide about it instead of a representation. In the case of a resource that's not an IR, that's your only option, in the sense that a 200 is inappropriate. What you can return from the resource identified by the URI to which you're redirected is more or less unconstrained. Not sure if that's what you were asking. > Further it seemed to me that most of the members in the discussion > reasonably saw the need for the HTTP URL to identify a thing and were okay > with some ambiguity, but that TimBL was most adamant that it behave certain > ways in order that it be consistent with his vision for RDF. Would you > concur or disagree? I don't think I'd care to comment on that. The httpRange-14 discussions were well along when I joined the TAG, and I prefer not to speak for other TAG members such as Tim. > BTW, my takeaway from the results of that discussion (thus far) is that > things might have been much different had RDF not been a central focus of > TimBL at that time. That seems to me to be a shame considering how RDF is > still only used on the periphery of the web and certainly not as part of the > mainstream web. And IMO, RDF will probably never make the mainstream > because it requires people to be too concise, and people in general are not > good at being concise (witness the percentage of HTML files on the web that > validate...) Well, we can all make our own bets as to which W3C technologies will be broadly adopted, but there's no question that the Semantic Web and RDF represent a major investment for the W3C as a whole, not just for Tim. Many, many people have worked hard on it. I confess that for some years I was unconvinced that I knew how to position it relative to XML. XML is a major investment for companies like IBM (my employer), and the use cases for XML and for RDF overlap quite a bit. Eight years ago I didn't see how we could make a major investment in both during the same period, and to a significant degree we didn't. Most of the focus went to XML. Speaking for myself, I think the industry is now starting to clearly see the ways that the semantic web adds value beyond what was in the traditional HTML-based Web and XML. In particular, I find things like dbpedia to represent a fascinating example of how data can be shared and re-aggregated on a global scale. In some sense, I'm saying that I find the goal of "linked data" more immediately helpful than the goal "semantic web", but they're closely related, and I'm in any case not against doing a careful job on the semantics in the cases where doing so adds value. In short, I personally think that the Semantic Web effort is going to start to pay off soon. Just how successful it will be, none of us knows, I guess. As to whether httpRange14 would have come up anyway, I'm not sure. There certainly seems to be value in having first class identifiers for tangible things as well as for information resources, and having all those identifiers come out of the same namespace makes a lot of sense. Being able to follow links to information about tangible things and to distinguish the resource that's tangible from the resource that's "about" the tangible one also makes sense. So, I think this discussion would have value even if the semantic web itself weren't the motivation. BTW: I'm not sure it makes sense to go to far with this discussion here. www-tag is a public list, and much of this ground has been covered there. If there's more to discuss on httpRange-14, wouldn't it make sense to do it on www-tag? Thanks. Noah -------------------------------------- Noah Mendelsohn IBM Corporation One Rogers Street Cambridge, MA 02142 1-617-693-4036 --------------------------------------
Received on Wednesday, 19 December 2007 20:28:51 UTC