If you're working with TXDocument, you're working with an ANCIENT version
of Xerces. TXDocument supported only DOM Level 1. Xalan currently requires
DOM Level 2. Three options:
1) Upgrade your parser.
2) (as you've discovered) Send the old DOM out as SAX and let us
reconstruct it in a namespace-aware model.
3) Work on developing a subclass/extension of DOM2DTM which understands how
to derive the additional information we need from a Level 1 DOM. That's
certainly possible. I don't think anyone is actively working on it right
now/. But this is an open source project... so feel free to develop and
submit a set of proposed patches to add DOM1 support. (Shouldn't be very
hard; I think the largest change would be extending DOM2DTM to bind
namespaces appropriately when a Level 1 DOM is passed in.)
