Re: 1343 messages later

-----Original Message-----
From: keshlam@us.ibm.com <keshlam@us.ibm.com>
To: xml-uri@w3.org <xml-uri@w3.org>
Date: Wednesday, June 14, 2000 3:19 PM
Subject: Re: 1343 messages later

John C:

>>It lets everyone *except me* go back to work, you mean.  As editor of
>>Infoset, I can't accept creative ambiguity, or echo it.  I have to
>>have a definite resolution: what does the Infoset expose?
Joe K:
>Which in turn affects what the DOM exposes (or, at the very least, how it's
>exposed). Remember, the DOM's primarily an API for the Infoset, plus some
>convenience features.


If it were just what the DOM exposes, then it would not be so bad.
But the DOM has to prevent multiple occurrences of the same attribute,
so it has to do comparison.  The base URI of a document has to be
made read-only.

This is IMHO a good thing as there may be relative URIs in other contexts.
To make a new document  with a different URI requires carefully thinking
about any occurrences a relative URIs (or even absolute URIs of
resources close in URI space) to decide whether they should be made
to a copy of that resource ort the original.  This is of course a question
of what the user wanted.  (which is why when you try to do that in an editor
you often get various options on the SaveAs box.)

The absolute URI of the namespace is of course just a function of the
URI-reference and the base URI - does the infoset define many properties
which are just functions?  It is however more fundamental than the
URI-reference.

Received on Monday, 19 June 2000 08:45:03 UTC