[Fwd: Re: [tei-council] Conformance draft: namespace purity]

Syd Bauman Syd_Bauman at Brown.edu
Tue Apr 3 14:00:21 EDT 2007


> I will make the field have a default of "-" (say), and make this
> illegal. They have to put in some text other than - or something
> which does not start with "http://www.tei-c.org/", or they can
> consciously empty it.
> 
> people who really want to force it to add elemnets in the TEI
> namespace have to edit the ODD by hand.

I think that (given the non-TEI namespace for new element
requirement, which I think is a bad idea) this all makes sense,
except that I still lean towards making the default be a valid (if
useless) namespace URI. Makes it a lot easier to use Roma for quick
testing, a use to which webRoma is often put, I think.

I agree that a namespace URI in www.tei-c.org, even if it has
"user-extension/" carries too much of an implication; I agree that
putting an e-mail address inside the URI by default is a bad idea.
But note that the "tag" scheme that Conal pointed us to does not
require an e-mail address. It only requires a DNSname, and if I read
the BNF right, a dotted-quad IP address qualifies as a DNSname. Thus

  tag:128.148.123.321,2007-04-03:P5customization/$name

(where $name is whichever one of prefix, filename, schemaSpec/@ident
that Sebastian finds easier to use) seems reasonable to me.




More information about the tei-council mailing list