[tei-council] MD chapter revised: namespace rules

[James Cummings]

Sebastian Rahtz wrote:
> James Cummings wrote:
>>
>> I recognise I'm side stepping some of the issues there.
>>   
> When you contact Johannes and say "can you send me your
> document for my corpus", what does he do?

He says "My document is freely available at the following URI:
http://www.projektmann.de/meinDokument.xml
It's teiHeader includes a pointer to the licence it is released under, and
also a pointer to the ODD, just where the Guidelines say I should include
such things."

> I am rapidly coming to the conclusion that Roma
> should seriously downplay the use of <altIdent>,
> and use it mainly for the full-scale language
> translations, where the whole doc changes
> namespace. Using <altIdent> to rename
> div1 to section, and then forcing JP to
> use <my:section> seems unfriendly and
> unhelpful. If he really wants to use alternate
> notation for the TEI abstract model, better if
> he switches namespace for the whole document.

So you are suggesting that if I want to rename 'div1' to 'section' then the
entire document is no longer TEI, doesn't use TEI elements and so shouldn't
be in the TEI namespace??!  That seems much more extreme than anything
we've been suggesting, doesn't it?

> I shudder to think about the work needed on
> Roma to support the last 2 weeks discussion
> properly.

Yes, that is my worry as well.

> I am now assuming that, inter alia,
> Roma should support a "canonical mode", in which
> <altIdent> and <equiv> follow a different
> route, in which the schema created is canonical,
> but as a side effect an ad hoc canonicaliser
> is created for you, to apply to documents
> which are valid against a non-canonical
> schema.

I'm not sure I buy this two types of conformant documents, a local
non-namespaced one and a canonical one.  While command-line Roma maybe
should have this distinction, I'm convinced that we should strive to have
webRoma only produce ODDs (to generate schemas) to validate Conformant
documents.

-james

-- 
Dr James Cummings, Oxford Text Archive, University of Oxford
James dot Cummings at oucs dot ox dot ac dot uk