[tei-council] MD chapter revised: namespace rules

[Daniel O'Donnell]

On Wed, 2007-04-11 at 11:37 +0100, James Cummings wrote:
> Lou Burnard wrote:
> > Wittern Christian wrote:
> >>> Like everybody else, I'm having trouble with simple project-based
> >>> renamings, since they can remain in the TEI namespace under principle 3.
> > I think there's a misunderstanding here. Renaming an element is *not* a
> > clean modification, since the set of documents now regarded as valid is
> > not a pure subset of the documents regarded as valid before the
> > modification.
> 
> 
> I think this misunderstanding arises from my proposal...

Oh good--it's not just me.

> 
> I was suggesting that TEI Conformant documents could have renamings in the
> TEI namespace, as long as they could be reverted according to the
> information in the referenced ODD.  And then TEI Interchange Format would
> insist that any such renamings be reverted before interchange.

This is what I understood as well.

> 
> However, it seems like a majority of people now do not see renamings as a
> clean modification, so renamings must then appear in a new namespace.

The only issue I have with that is that it seems to vitiate any reason
for following canonical methods of altering or renaming elements. It
seems to me that your original explanation in the paragraph above
represents a case where people are working "within" the TEI. If the ODD
references the changes and they are reverted before interchange, then it
seems to me that non-conflicting renamings are suitable for remaining in
the namespace of the version or translation to which they belong.

-dan

> 
> So:
> 1) Any new element must be in user-defined namespace (UDN)
> 2) Any new attribute must be in UDN
> 3) If you make a dirty (i.e. non-subsetting) change to an element it must
> move to a UDN (and possibly be renamed)
> 4) If you make a dirty change to an attribute it must move to a UDN (and
> possibly be renamed)
> 5) The only changes to elements or attributes which do not result in them
> being moved to a UDN are subsetting changes.  That is, changes with further
> restrict content models, attribute value lists, or datatypes, in such a
> manner that a document which validates against this schema also continues
> to validate against tei_all.
> 
> Is that a fair summary?
> 
> -James
> 
-- 
Daniel Paul O'Donnell, PhD
Department Chair and Associate Professor of English
Director, Digital Medievalist Project http://www.digitalmedievalist.org/
Chair, Text Encoding Initiative http://www.tei-c.org/

Department of English
University of Lethbridge
Lethbridge AB T1K 3M4
Vox +1 403 329-2377
Fax +1 403 382-7191
Email: daniel.odonnell at uleth.ca
WWW: http://people.uleth.ca/~daniel.odonnell/