It means that you can't tell from xml syntax as specified whether
<x><a>z</a><b>z</b></x>
means the same as
<x>
<a>z</a>
<b>z</b>
</x>
you could assume it's always significant (which would never allow you
to re-indent or otherwise re-format some xml)
but in practice you need external knowledge of the vocabulary.
in (x)html
<p><b>z</b><i>z</i></p>
is not the same as
<p>
<b>z</b>
<i>z</i>
</p>
but
<body><p>z</p><p>z</p></body>
is the same as
<body>
<p>z</p>
<p>z</p>
</body>
There is no syntactical (lexical) difference in the way white space
content of p and body are marked up in html: the significance or
otherwise of the space is specified in the processing of xhtml not in
the way it is parsed.
David
