Re: [docbook-apps] XInclude: multiple <preface>s

[Paul Grosso <pgrosso at arbortext dot com>]

At 16:58 2003 04 01 +0100, Dave Pawson wrote:
>At 16:04 31/03/2003 -0500, Daniel Veillard wrote:
>
>>  The XIncluded parts must be well formed XML documents. That's
>>completely indepedant of the DocBook syntax, you don't have the
>>right to put multiple roots in an XML document, be it XIncluded or not.
>
><grin/> Does that come next year Daniel?
>Multiple roots I mean.
>xml-dev keeps hinting it could be useful?


I don't think it's fair for xml-dev to put pressure on an implementor 
to implement something that is against the standard.

If xml-dev (or whoever) wants to be able to xinclude things with
multiple roots, then it should press for a change to the XInclude
specification (which, I'll note, is not even yet a Recommendation,
so Daniel is to be commended for his work to date on it).

For what it's worth, XInclude is defined in terms of merging 
infosets, so the XInclude spec, in fact, can't do anything about 
the problem directly.  Rather, the XML Infoset would need to be 
augmented to define an infoset for things with multiple roots--that 
is, an infoset for things that are NOT XML documents.  

It's not like others haven't thought about this before.  This is 
trickier than it looks which is why it hasn't yet been done, but 
it may yet get done, especially if the user community puts the 
pressure in the right spot.

But don't push implementors to implement non-compliant code.  
That does no one a service.

paul

p.s.  I know Dave is quite sensitive to the issues of the standards
process, so I hope he will understand that I'm not speaking directly
to him, but rather using his relatively innocent comment to make a
point that I think needs to be made both on this list and on xml-dev.




---------------------------------------------------------------------
To unsubscribe, e-mail: docbook-apps-unsubscribe at lists dot oasis-open dot org
For additional commands, e-mail: docbook-apps-help at lists dot oasis-open dot org