[tei-council] Re: [rng-users] time to really allow instance to name schema?
sebastian.rahtz at oucs.ox.ac.uk
Mon Feb 19 17:54:42 EST 2007
Jirka Kosek wrote:
> I think that we are not agreeing on what is "information content" of
> document. I include elements/attributes/text nodes under this term,
> while I take PIs and comments as something which doesn't convey primary
> information of document, just some stuff which can be used by some
> users/tools as a suplemental information when document is processed in a
> particular way.
I probably do agree with you. But I am still prepared to use
elements for metadata; after all, you and I both work on a
W3C scheme (ITS) which is happy to store meta information
about processing information in elements the header....
> Nope, PIs can be before root element, for example
> <?schema href="sdocbook.rng"?>
> <article ...>
ah, OK. true. But you still get at that PI with an XML parser,
don't you? or do you just read the first few hundred bytes
and hope you find it?
> OK, so then you will not use <?schema at all or you will just ignore
> this PI. But there are still people who need this functionality.
If OASIS or W3C want to start a workgroup to standardize
a PI, that's fine, of course. I can't believe the W3C will,
though, because they only recognize the existence of
XSD, which already has its truly horrible attribute.
So how does one start an OASIS activity to look at this?
Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
OSS Watch: JISC Open Source Advisory Service
More information about the tei-council