xml-dev - Re: Should we care about elements order ?

[Ken MacLeod <ken@bitsko.slc.ut.us>]

Tim Bray <tbray@textuality.com> writes:

> At 05:20 PM 20/10/00 +0200, Eric van der Vlist wrote:
> >1) Should we always pay attention to the relative order of the
> >elements within XML documents ?
> 
> No.  Tons of conventional business data is oblivious to the order of
> its fields, and one of the nice things about an RDBMS is that I can
> arrange for the rows of the table and the colums of a row to be in
> any order I like.
> 
> XML is differentiated in that it is *possible* (not necessary) to
> deal with order-significant data.  The most obvious examples where
> this is useful are in the publishing space, but there are others.
> In my opinion, this is the one aspect of XML that is the hardest to
> deal with when trying to apply traditional relational
> technology. -Tim

One of the underlying questions, which is in the original post but
hasn't come up in the followups, is in a situation where one *doesn't*
want to enforce element order, especially in which order groups of
elements from different namespaces occur, can one still perform
validation with any current/popular/recommended schema language?

A corollary is when you have a structure like this:

  <aa xmlns="uri:a"
      xmlns:b="uri:b"
      xmlns:c="uri:c">
   <b:b/>
   <c:c/>
  </aa>

and you have schemas available to you for uri:a, uri:b, but not uri:c.
Can you validate the structures for uri:a and uri:b, and still have
uri:a allow for elements, unvalidated internally, from other
namespaces, like uri:c?

It's not a big deal if the answer is "No", but it does mean that for
those applications the current/popular/recommended schema languages
won't/can't be used.

  -- Ken