Re: [xml-dev] XML 1.1 grinds to halt?

[Michael Kay]

> Since the bottom line in the Namespaces spec is that parsers are
> supposed to accept any character string in the role of a namespace name,
> the obvious solution is to remove all the fine words about IRIs which
> don't actually affect conformance anyway.
>

If an IRI is supposed to be able to use any local encoding, how is the
network supposed to know which encoding is being used for any particular
IRI?

Cheers,

Tom P