=====================================================
= EXTENDED DOCUMENT OBJECT MODEL BUG AND TO-DO LIST =
=====================================================

KNOWN BUGS
----------
- XPath:
  - TdomNode.compareTreePosition does not support 
    XPathNamespace nodes.
  - The XPath id() function does return an unordered
    node-set.  In practice it is also unusable, 
    because validating of namespace-aware document
    trees is currently not supported, so building an
    ID list for such a document is not possible.
  - The implementation of the XPath number functions
    does not distinguish between positive and negative 
    null.
- The column number for an error in the first line of
  an internal subset of a DTD is incorrect.
- A few minor bugs and problems are indicated in the
  source code by 'xxx' accompanied by a brief comment.


RARELY TESTED FEATURES
----------------------
The following XDOM features need more testing.  Use
them with care:
- User data events.
- Document mutation events.
- Repositioning of a TdomNodeIterator object after
  document mutation.
- XPath evaluation.


TO DO
-----

- Add a 'userdata' property to TdomCustomNode.
- Consider the standalone status when parsing a 
  document.
- Implement asynchronous parsing.
- Add support for AS_STRICT_MIXED_CONTENTTYPE in
  TdomASModelCollection.validate() and
  TdomElement.validate2().
- Test for correct entity references when
  validating a DTD content model.
- Support resolvement of unparsed entities in
  TdomASEntityDecl.
- Add a Suspend flag to the TXmlDocBuilder class.
- Improve formatted output.
- Optimize validation.
- Optimize XPath evaluation speed.
- Improve ISO-639 / RFC 3066 support.
- Review the code for private methods to be moved
  to protected.
- Change type of RefNode parameter of some parser 
  methods from TdomNode into a more specific type.