Re: [xml-dev] What A Document Means

[Bullard, Claude L (Len]>
>
> That is the direction I thought they might be headed in with that,
> but without a reference, I wasn't sure.   So it is an infoset.  That
> is a bit vague and doesn't quite tell us why it is an expensive
> approach, or precisely how it determines a document's "meaning"
> except insofar as information items are grouped.   In other words,
> we can use the infoset to describe it's meaning as a well-formed
> document (as document is defined therein), but not the meaning
> of the content.    ...

Let's also remember that an http: url retrieves a "representation of a
resource".  What then is the "document"?  Is it the resource or the
particular instance of its representation?  I presume that it is the
instance, in whatever MIME (or infoset) from it may have, but it would be
well to have this distinction spelled out.

Cheers,

Tom P