Re: May F&O: 15.2.1.1 Deep Equality of Nodes

David Carlisle wrote:

In comments on earlier drafts I commented that deep-equality ought to be
moved to the user-defined examples appendix as it will almost always be
the case that specific applications need to tweak this definition
to have different behaviour with respect to comments, white space,
in scope namespaces etc etc. 

However whether the definition is moved to the appendix or left as a
core function in this section, there are some XPath corrections.


At the recent F&O meeting we first discussed whether we wanted to retain
this function, and decided, on balance, that we should do so. It's worth
remembering that simple equality is no longer defined between elements with
element content, based on their string values, and this function is intended
to fill that gap in a more semantically rich way.

We accepted your technical corrections to the spec - thank you.

We also discussed the handling of comments. We felt that it is right to
ignore comments encountered as children of the nodes being compared, but
that if the user explicitly compares one comment with another, it would be
very confusing to return "true" when the comments are quite different from
each other. So we're going for usability in preference to consistency. I'm
sure this won't please everyone, but we feel it's the best we can do.

Thanks for your comments, as always.

Michael Kay

Your comment is numbered FO-LC1-0018
This note discharges action A-NIST-07

Received on Friday, 23 May 2003 18:45:33 UTC