Hmmm... from Section 4.1.1.... "Its element children consist of metadata elements followed by zero or more atom:entry child elements."
This is pretty clear. The entries are expected to come last. On Thu, May 20, 2010 at 11:53 PM, eric scheid <[email protected]> wrote: > > On 21/5/10 10:06 AM, "Pablo Castro" <[email protected]> wrote: > >> The Atom spec requires all the atom:link elements in a feed to come before >> the >> actual entries (section 4.1.1 of the RFC). > > Really? > > The only time the spec references the question of significance of order of > child elements (eg. atom:entry in atom:feed, child elements of atom:entry, > child elements in a Person construct, etc) it is quite clear that no > significance is assigned. > > In context of RFCs, the "exception proves the rule" doctrine does not apply. > > Note too that the RELAX NG expressions in the spec are _informative_ only, > and not _normative_. > > "Some sections of this specification are illustrated with fragments of > a non-normative RELAX NG Compact schema. However, the text of this > specification provides the definition of conformance." > > Since the _text_ spells out no MAY SHOULD MUST specification for the order > of child elements of atom:feed at all, this means there is no such > requirement in effect. > > Feel free to to put your atom:link elements following your atom:entry > elements. > > e. > > -- - James Snell http://www.snellspace.com [email protected]
