RE: Identity transformation for XSLT 2.0

> With XSLT 2.0 that amputated I am wondering why saxon:parse() and
> saxon:serialize() didn't make it to the spec?

They were never proposed, and never considered by the Working Group. I only
recognized the potential of these functions long after the time had passed
for adding new features to the spec.

I'm not sure whether the WG would have warmed to them anyway. Some members
might have felt they were a step too far. Hard to tell - committees can be
unpredictable; but there were several members adamantly opposed to
dyn:evaluate() which had a longer track record of implementation and
adoption in XSLT 1.0 processors.

Michael Kay
http://www.saxonica.com/