Signs on the Sand

XPointer goes to Recommendation today. Remember XPointer? It's about "pointing", i.e. identifying of XML parts or fragments. Well, after monsntrous XPath2 specs XPointer one looks like a miniature work. (It reminds me XPath1, which was about 30 pages also).

XPointer (funny enough, no version defined for the language) consists of 3 quite small documents:

1. XPointer Framework, which defines basic semantics and syntax of XML addressing/fragment identifying.
2. element() scheme, funny XPath-like syntax for pointing to elements in XML, e.g. "element(intro/3/1)" - which identifies first child of third child of the element with "intro" ID.
3. xmlns() scheme, for dealing with namespaces in pointers.

Editors are just living legends: Norman Walsh, Jonathan Marsh, Eve Maler and Paul Grosso. They decided to leave the most powerful and (therefore?) contradictory xpointer() schema out of this release, it's still frozen at WD stage. Implemeters feels relief today probably.

Anyway, happy sailing in XML ocean, XPointer. There are many against you, but you can make it.

As a matter of fact main thing I wanted to say is that this event effectively means XInclude, which greatly depends on XPointer and still stays in Candidate Rec bed apparently may be also released very soon. So it's really right time to push my (our now) GotDotNet XInclude.NET project. In fact that involves XPointer implementation also, so there is a big pile of design and coding here. We desperately need more volunteers, I urge everybody interested in free implementation of XInclude for .NET platform to participate.

del.ico.us | digg it | reddit