- From: Jonathan Marsh <jmarsh@microsoft.com>
- Date: Thu, 15 Sep 2005 10:41:44 -0700
- To: "Arthur Ryman" <ryman@ca.ibm.com>
- Cc: "Bijan Parsia" <bparsia@isr.umd.edu>, "Dan Connolly" <connolly@w3.org>, "Henry S. Thompson" <ht@inf.ed.ac.uk>, <public-ws-desc-comments@w3.org>
- Message-ID: <37D0366A39A9044286B2783EB4C3C4E81A2800@RED-MSG-10.redmond.corp.microsoft.com>
Yes, Dan's simple pointers point to XML elements rather than the description component, per the XPointer spec if we keep compatible with it for our media type (and we should!). If that is sufficient for his use he's free to do so, and even possibly layer on higher-level inferences of equivalence between the element and the component within a particular domain. But testing XPointer is not something I feel the WSDL WG should engage in. However. my job here is simply to bring a clear issue, hopefully with a proposal, before the WG for resolution. ________________________________ From: Arthur Ryman [mailto:ryman@ca.ibm.com] Sent: Tuesday, September 13, 2005 3:27 PM To: Jonathan Marsh Cc: Bijan Parsia; Dan Connolly; Henry S. Thompson; public-ws-desc-comments@w3.org; public-ws-desc-comments-request@w3.org Subject: RE: simple case of IRIs for Components in WSDL 2.0 Jonathan, We don't have any test cases for the frag ids yet. However, if we do add them, I don't think this implies we test Dan's proposal since it is very problematic and not part of the spec. Isn't it the case that if users want to use simple fragment ids, they can add an attribute of type ID to the element they want to reference? That is allowed by the WSDL 2.0 schema. Wouldn't that we the correct normative interpretation? The XPointer Framework calls these shorthand pointers [1]. [1] http://www.w3.org/TR/xptr-framework/#shorthand Arthur Ryman, IBM Software Group, Rational Division blog: http://ryman.eclipsedevelopersjournal.com/ phone: +1-905-413-3077, TL 969-3077 assistant: +1-905-413-2411, TL 969-2411 fax: +1-905-413-4920, TL 969-4920 mobile: +1-416-939-5063, text: 4169395063@fido.ca "Jonathan Marsh" <jmarsh@microsoft.com> Sent by: public-ws-desc-comments-request@w3.org 09/13/2005 05:23 PM To "Dan Connolly" <connolly@w3.org> cc <public-ws-desc-comments@w3.org>, "Bijan Parsia" <bparsia@isr.umd.edu>, "Henry S. Thompson" <ht@inf.ed.ac.uk> Subject RE: simple case of IRIs for Components in WSDL 2.0 Since component designators aren't used internally by WSDL I would guess we would not conduct tests of them, indeed I'm not sure what that would mean. But the WG hasn't discussed this yet. I'll try to bring it up when we move to CR. > -----Original Message----- > From: Dan Connolly [mailto:connolly@w3.org] > Sent: Monday, September 12, 2005 12:17 PM > To: Jonathan Marsh > Cc: public-ws-desc-comments@w3.org; Bijan Parsia; Henry S. Thompson > Subject: RE: simple case of IRIs for Components in WSDL 2.0 > > On Mon, 2005-09-12 at 11:49 -0700, Jonathan Marsh wrote: > > Would you be satisfied with adding a note along the lines of the > > following: > > > > "Note that the component designators given below are only one form > of > > identifier for WSDL components. Other forms can be used. For > instance, > > using targetNamespace#name is sufficient when out-of-band mechanisms > can > > be relied on to ensure no names are the same (across all symbol > spaces) > > within a WSDL component model. Such a mechanism cannot be relied > for > > general purpose use as is the one defined below." > > Yes, that's pretty much what I have in mind. > > That's not completely clear that targetNamespace#name has the > same normative status as the other format, but I'm not inspired > with better words. > > Do you have plans to do tests for component designators? I'd > be satisfied to see targetNamespace#name right next to the > others in a test suite. > > > Trying to cast this as a concrete initial proposal the WG could > > deliberate... > > -- > Dan Connolly, W3C http://www.w3.org/People/Connolly/ > D3C2 887B 0F92 6005 C541 0875 0F91 96DE 6E52 C29E
Received on Thursday, 15 September 2005 17:43:22 UTC