History and structure

• Originated from a single XML working group
• The groups has been splitted in multiple working groups last September
• An XML Coordination Group coordinates the various XML activities
• Groups also explicitely provide Liaison Statements to express dependancies.
• The collection of all the members of the XML groups are the Plenary Group
• Some groups are not chartered within the XML activity but are strongly tied to XML.

Classification

• Some groups are finishing the work started during the initial working group: Schema, Linking, Infoset.
• Some groups have been scheduled to handle new topics: Fragment, DOM, XSL
• Some new working groups are to be created: Packaging and Query

Descriptions

• Schema:
  • Dsign an alternative mechanism to DTDs
  • Provide a type system for XML content
• Linking:
  • Xpointer: adressing mechanism for subparts of XML documents
  • XLink: extended hypertext link mechanisms more elaborate than the simple HTML ones
• Infoset:
  • Describe the data model for XML documents

    => canonical XML

Descriptions II

• Fragment:
  • extract a subpart of an XML document and associate metadata for independant analysis

    => signature for the fragment or the original document could be added

• DOM:
  • Document object model, set of API for navigating in document and modify them
• XSL:
  • Transformation language

    => useful for client side processing

  • Stylesheet properties

New groups

• Packaging:
  • how to package an XML document with the associated data, like stylesheet, dtd, external entities ...

    => adding signature for the package and individual files within it

• XML Query:
  • normalization of interfaces for querying and searching XML documents
• XML revision:
  • only minor revision expected in the short term

Questions

Forwarded from Tim Bray:

1. Are people from the Signed XML workshop are interested in signing "bits on the wire" or a precomputed signature done on a canonalization form ?
2. One of the biggest troubles of the canonalisation process is about normalizing combining characters. It is extremely resource consuming, but this step may be required for a 100% unique canonical form. Do people care about this level in canonalization.