- From: Rushforth, Peter <Peter.Rushforth@NRCan-RNCan.gc.ca>
- Date: Wed, 22 Aug 2012 12:13:56 +0000
- To: John Cowan <cowan@mercury.ccil.org>
- CC: "public-microxml@w3.org" <public-microxml@w3.org>
Hi John, > Rushforth, Peter scripsit: > > > I'll be particularly interested in how the public schemas are > > _referenced_, as well as what format they are in. > > I should really have said "public document types"; there's no > requirement that a formal schema be provided or even exist. > If we decide to have PIs, then the xml-model PI can be used > to refer to a schema from a document claiming to conform to it. > > > I see in your later email you refer to "a source and an _archmap_". > > Is the archmap component found in-line, or by reference? > > The relation of archmaps to source documents will often be > many-to-many. > That is, a source document may be mapped, using multiple > archmaps, to different target documents. Consequently, it's > unclear that any such direct linkage between documents and > archmaps should be defined. > Alternatively, the xml-style PI (if we have PIs) could be used. That would be a wasted opportunity to define global hypermedia affordance support, and thus improve the self-contained-ness of MicroXML. > > > > Few XML documents are on the web, though many documents > on the web > > > are generated from XML documents. > > > > Another way of putting that is: > > > > Many resources on the web should have an XML representation because > > they are generated from XML. > > "Should" is an open question. Some do, most don't; it's up > to the publisher of the document. Agree; if you choose to hide the semantics of your resources from applications, putting them in HTML-only is a good way. But, then again, there's a big movement of putting semantics in HTML via various means. The architecture of the web is designed such that the resource-representation model allows a clean separation. But, should you decide to make the semantics of your content available to applications, XML is a good way. I thought MicroXML was about helping off-load the semantic overloading of HTML. Perhaps I was wrong. > > > I'm a user of XML, and potentially MicroXML. What would be the > > intended client of a MicroXML document? > > I don't know. Someone who finds full XML overkill for their > purposes, whatever those purposes might be. I think the point here is: _what_ would be the application which consumed a MicroXML document, not who. Peter
Received on Wednesday, 22 August 2012 12:14:26 UTC