Re: Blueberry is not "closed" (was: Closing Blueberry)

[John Cowan <cowan@mercury.ccil.org>]

Elliotte Rusty Harold scripsit:

> Even without this, I think Blueberry documents require a Blueberry
> DOM. They'll need to use a different group of legal character tables
> for Blueberry and non-Blueberry documents. (though I admit one thing
> that really hasn't been addressed at all yet is the changes Blueberry
> will require in other specs like XPath, DOM, Schemas, etc.)

I believe that no other spec is sensitive to either "What characters are
allowed in names?" or "What characters are whitespace/line terminators?",
the only two Blueberry questions.  (If anyone knows of any, please
speak up.)

The changes, therefore, would be mostly a matter of reference:
the specs would have to be updated to point to Blueberry instead of 1.0.

IMHO there is no need to discriminate at the DOM level.  A compliant
DOM need not reflect whether a document is Blueberry or not; the
generation of XML text from a DOM (or from XSLT, or whatever) can
generate the Blueberry mark only as needed.  I agree that a DOM
implementation that always generates the mark is a Bad Thing,
but that does not require a separate DOM.

-- 
John Cowan                                   cowan@ccil.org
One art/there is/no less/no more/All things/to do/with sparks/galore
	--Douglas Hofstadter