[tei-council] xml-colon-thing

Sebastian Rahtz sebastian.rahtz at computing-services.oxford.ac.uk
Mon Nov 22 06:31:19 EST 2004


Syd Bauman wrote:

>I am sure I understand the current problem: editing a single file of
>a multi-file document (like the Guidelines) is a big pain with XML
>IDs that point to another file.
>
>  
>
they point to other places within the same document; that
document _happens_ to be instantiated as several files on disk
at present, but that is by no means set in stone. We could switch to
a single XML file tomorrow if we felt like it.

>I agree with SOW09's point (David Durand's idea, I think, but might
>have been someone else's) that if one used a relative URI reference
>this would not be a problem.
>  
>
but you can't use a relative URI. a URI points to another document. the
other chapters are not other documents.

>But I'm a bit confused ... since XML editing software doesn't
>typically validate bare name references, wouldn't using just a bare
>name avoid most the problems as well as using a relative URI
>reference would? E.g. <ptr target="#COXR"/>
>  
>
yes, thats what I assumed we would do

>
>Seems a little counter-intuitive to me -- why wouldn't the wave /
>particle dilemma apply?
>
'taint a dilemma for me :-}

> I think of the Guidelines as a single
>document, as I think most people do, but chapter 2 is even produced
>as a separate document.
>  
>
no. a separate document can be produced which uses
the same content as chapter 2. not the same thing....

>I'm not sure what you mean by "internal hackery", but if it isn't
>tricks like using a relative URI reference ("./name#here") instead of
>a bare name ("#here"), what is it?
>  
>
when you process chapter X which references #foo in chapter Y,
you consult the eXist database to see which chapter #foo is in,
and say "somewhere in chapter Y".

sebastian



More information about the tei-council mailing list