RE: Summary of strings, markup, and language tagging in RDF (resend)

> Well, it doesn't convey whether an attribute was single-quoted or
> double-quoted, but then even the XML parser may not convey that
> information. It does eliminate, in certain well-defined ways,
> the difference between attributes and elements, because that is
> seen as a syntactic distinction. But I'm not aware of any
> actual information in attributes or elements that is just
> thrown out of the window.
>

Ordering, namespace prefixes, where the colon appears in many qnames, ...
there's lots and lots.

> I'm quite sure many readers and users will understand the treatment
> of XML literals as typed literals as just a technical mechanism,
> and therefore actually be surprised. That will be the case in particular
> because the users that use XML literals will be familiar with XML.
> After all, it's XML, not something else.

That's certainly true.


> >I believe we have already fulfilled our charter obligations to
> clarify the
> >I18N support already in M&S;
>
> Just for the record, I think that both the current removal of xml:lang
> from xml literals and the creation of a distinction between xml
> literals with plain text only and plain literals are in conflict with
> the M&S spec, and I think you have agreed with at least one of these
> points.
>

As our group has realised, you can prove anything from M&S, ex contradictio
quodlibet.

It is a question of degrees of conflict and the amount of weight one gives
to different sections. M&S is not an unambiguous document.

Jeremy

Received on Thursday, 3 July 2003 12:11:28 UTC