Re: [Fwd: Re: Ameliorating no change on XML Literal design]

At 10:45 03/07/18 +0100, Graham Klyne wrote:

>Martin,
>
>what you're saying here about XML doesn't seem to fit with what you were 
>saying earlier about text with markup.  Particularly:
>
>[[
>I.e. an XML literal denotes an XML fragment the same way an
>integer denotes an integer.
>]]
>
>Here, you seem to say that an XML literal is *not* text with markup, but 
>your original comments seemed to be focused on the disparities between 
>treatment of XML literals and plain text.  If we're to have a constructive 
>discussion, I think we need to remain clear about the requirements we're 
>trying to address.

Ok. I think the core requirements are:

- Clear distinction between markup and text that looks like markup.
- Easy transition (as easy as possible) from text without markup to
   text with markup.

I think that a lot of current, and potential future, use of XML
literals in RDF can be explained most easily as 'text with markup'.
This is in particular true for various i18n-related usages.
And this way of looking at things helps understand the transition
issues. However, I think it would be a bit stretchy to say that
all XML literals in RDF are text with markup. Some of them may
only be markup, in other cases, there may be element content that
if of course a sequence of characters but barely will qualify as
text (in a similar way that it may be difficult to call some of
the character sequence in plain literals 'text', depending on the
definition of 'text' one uses).

I don't think there is any contradiction in saying:
    "An XML literal denotes an XML fragment the same way an
     integer denotes an integer."
and
    "An XML literal/fragment that consists only of element content
     (i.e. no markup) is equivalent to the plain literal with the
     same sequence of characters (and the same language information)."

Hope this helps,       Martin.

Received on Friday, 18 July 2003 16:58:31 UTC