I don't think this is really about conformance. The point is that in P4,
the @value attribute was defined as being a normalised version of the
date, normalised according to some undefined (user-chosen) scheme. We've
now decided to go with one particular normalisation scheme and defined
the attribute accordingly. If you want to normalize according to another
scheme, then you need to define another attribute. I don't think you're
at liberty to use this one with such a radically different meaning, any
more than I think it would be a good idea to take the TEI table element
and use it to describe objects for putting cups on.
Sebastian Rahtz wrote:
> does this take us back to conformance?
> TEI P5 out of the box expects you to use W3C datatypes for
> machine-readable dates, for good or bad. This means that others
> can parse your XML with a well-understood set of expectations. If you
> need to change the datatype for date/@value
> to use some other scheme, for your project, you could make an ODD file
> (or whatever), where you expressed
> your degree of divergence from the out-of-the-box, and your intent.
> That said, I bet software for doing every conceivable date
> transformation is widely available....