- From: Wilson, Kirk D <Kirk.Wilson@ca.com>
- Date: Tue, 3 Mar 2009 13:57:41 -0500
- To: <ht@w3.org>
- Cc: <public-sml@w3.org>
- Message-ID: <F9576E62032243419E097FED5F0E75F306108700@USILMS12.ca.com>
Hello Henry, As you are aware, The SML team has been reviewing the Technical Notes that will be published in conjunction with the SML specifications. Monday we reviewed the EPR note: please see comment #18 to issue 5341: http://www.w3.org/Bugs/Public/show_bug.cgi?id=5341#c18, specifically the accompanying attachment (in PDF: http://www.w3.org/Bugs/Public/attachment.cgi?id=649). One of the points focused on the following phrase (in section 2.1, bullet 3): Since the URI (IRI) in the wsa:Address element of the EPR identifies only an endpoint of a service. . . The "URI (IRI)" needs further editing. The general question is, How should the note deal with the relationship between URIs and IRIs? As of the call on Monday, the thought was to delete the parenthetical reference to IRIs, since it is the only place in the Note (and spec !) where IRIs are referred to. But we wanted to check with you before making any change. Indeed, after further investigation, I believe dropping the IRI here is the wrong thing to do. According to the WS-A spec, the value of the abstract property [address] is defined to be an IRI. In fact, it seems that the WS-A moved from talking about URIs (in the Member Submission) in the abstract information model of an EPR to talking about IRIs in the Recommendation. A more concrete proposal, on which I would appreciate your thoughts, is: 1. Refer to IRIs for things pertaining to EPRs, which include values for the wsa:address and wsa:action elements. The sentence above would be "Since the IRI in the wsa:Address element...." 2. Use "URI" (1) where an actual example of a URI is provided, e.g., in the context, "the URI http://www....", and (2) for things pertaining to the SML specs (reference scheme URIs, aliases). 3. Text would need to be added in the section on Interoperability to recognize that IRIs in EPRs must be mapped to URIs in generating aliases and/or target-complete identifiers You might also recall a sentence from the EPR note, which was discussed on one of the previous calls (you engaged in that discussion). The sentence, as currently revised, is: However, not all documents or document fragments can be retrieved simply by means of a URI that may be dereferenced. Since this sentence deals with general addressability, I believe it should also refer to an IRI rather than a URI. Would you agree? Thank for your assistance in this mater. Kirk Wilson, Ph.D. CA, Inc. Research Staff Member, CA Labs Council for Technical Excellence Tele: 603 823 7146 (preferred) Cell: 603 991 8873 Fax: 603 823 7148 kirk.wilson@ca.com <mailto:kirk.wilson@ca.com> This e-mail may contain information that is privileged or confidential. If you are not the intended recipient of this communication, please delete this e-mail and notify the sender immediately. ________________________________
Received on Tuesday, 3 March 2009 18:59:05 UTC