RE: References to a value of an element in another element?

> On the other hand, it might be that everyone ends up 
> implementing the whole of XPath 2.0 because it's too much of 
> a pain to implement a subset. That's certainly what I intend to do.


On this subject, I do hope that the WG won't end up deciding to require a
subset "for the convenience of implementors". That's almost always a bad
thing to do. The current spec has an example of it - the rules for defining
valid derivation of xs:all types require the processor to treat the order of
the xs:all particles as significant, and the spec is quite explicit that
this rule is there only for the convenience of implementors. Well, I found
it really inconvenient to impose this restriction, so what am I supposed to
do? Impose the restriction anyway, despite the fact that it means writing a
lot of extra code?

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