- From: Mark Baker <distobj@acm.org>
- Date: Fri, 16 Jan 2004 07:48:47 -0500
- To: "Williams, Stuart" <skw@hp.com>
- Cc: www-tag@w3.org
Hi Stuart, On Fri, Jan 16, 2004 at 09:48:31AM -0000, Williams, Stuart wrote: > Hi Mark, > > FWIW I'll offer my personal perspective. Thanks, I think that's useful at this point. > In order to answer your question you have to accept the premise that the > XMLP-WG has changed which resource is referenced by the URI. That may be > open to question, but I don't think you can take it as a given - an > personally I don't see that they have. Good point. I might have been clearer if I described the issue as one of disagreement between a URI publisher, and the world at large, about what a URI identifies; that until recently, the representations returned from that URI were consistent with the different identity interpretations of "both" parties. > IMO, the clarification from the XMLP-WG is entirely consistent with URI > policy for TR page URIs as articulated in section 1 of the W3C pubrules [1] > (I encountered these working on the metaDataInUri-31 draft finding - looking > for an example of an articulation of URI assignment policy). Yes, I agree. But I don't think the W3C/XMLP did what they could have to make this policy clear. Specifically, prior to a few weeks ago when the redirect to SOAP 1.2 occurred, the content of the SOAP 1.1 spec was returned directly in a response with only the Content-Location header set to point to the /TR/2000/NOTE-SOAP-20000508 (IIRC). IMO, what would have better supported the policy would have been a transient redirect from /TR/SOAP. That way, the second URI would have been made visible to agents and humans (i.e. show up in the browser URI bar). > I'd be a little troubled by specifications making latest version references > rather than specific version references (I don't know whether the former is > a widespread practice), just as I'd be a little wary of writing a blank > cheque - or writing a cheque in pencil. Sure. But when basically the whole world decides that it isn't a latest version reference, then IMO, it isn't. FWIW, if you consult the Google Oracle for backlinks ... http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe=utf-8&safe=off&c2coff=1&q=link%3Ahttp%3A%2F%2Fwww.w3.org%2FTR%2FSOAP&btnG=Google+Search You'll see that most of those pages use the URI to refer to SOAP 1.1, rather than "SOAP in general". Mark. -- Mark Baker. Ottawa, Ontario, CANADA. http://www.markbaker.ca
Received on Friday, 16 January 2004 07:53:06 UTC