- From: Jan Hidders <jan.hidders@ua.ac.be>
- Date: Tue, 23 Sep 2003 19:02:18 +0200
- To: www-ql@w3.org
Bas de Bakker wrote: >> I don't think that answers my question. This says something about >> the order of the nodes but nothing about their order in document >> order: it is still quite possible that Value2 precedes Value1 in >> document order. It all depends on where in the universe of all >> created fragment newly created nodes are placed when they are >> created. > > Section 3.2 of the data model says: > > The relative order of free-standing nodes (elements, attributes, and > other nodes created outside the context of a particular document) is > also implementation-dependent but stable. > > That seems to answer your question. Yes, it does. So thank you, Bas, Torsten and Michael for your answers. However, now that I read this section again, I'm slightly confused. Is the relative order between element nodes inside a fragment really implementation-dependent? So the semantics of (<x><y id="1"/><y id="2"/></x>)/y could be either <y id="1"/><y id="2"/> or <y id="2"/><y id="1"/> ? Or should I read "document" in the above Section 3.2 quote as "fragment"? Is it also correct that the ordering can "mix" nodes of different documents? So a node of document d1 can be between two nodes of document d2? Finally I'm also a bit confused by the first sentence that says that the document order is defined on all the nodes in a document. Does that mean that it is not defined on nodes outside documents? What then is the name of the ordering over *all* the nodes? I thought *that* what was called the document order. -- Jan Hidders PS. I'm getting often two versions of each reply in my mailbox, one directed to the mailing list and one directed to my personal e-mail address. Is this intentional, or is the mailing list badly configured?
Received on Tuesday, 23 September 2003 13:01:46 UTC