Re: CR-18 Response: Validating XPointer IDREFs?

At 10:19 AM +0100 2/15/01, Eric van der Vlist wrote:
>> It would be helpful to us to know whether you are satisfied with the
>> decision taken by the WG on this issue, or wish your dissent from the
>> WG's decision to be recorded for consideration by the Director of
>> the W3C.
>
>Not really.
>
>I am still thinking that it would be much more flexible and powerful to
>remove the restriction that a field must identity a node:
>
><quote>
>This must identify a single node (element or attribute, not necessarily
>within the
>selected element) whose content or value, which must be of a simple
>type, is used in the constraint.
></quote>
>
>If a field was the result of a XPath expression without further
>restriction, then the issue I have reported would be solved and it would
>open many possibilities.

I don't know which part of the 'field' you are talking about (key-part or
keyref-part?) and how it works, since I don't know how dropping the
restriction will automatically solve your issue.  The current key/keyref
mechanism allows you to define a set of elementary fields (element or
attribute contents) as a key and another set of elemenary fields as a
key-reference. For a key-keyref relation, the values of these fields must
exactly match each other.  Dropping the restriction would bind more than
one physical fields to a single logical field. But how can this be used to
check the integrity of XPointer's IDREFS?

Could you give me some concrete examples and your proposed solution?

>I don't see how the spec would become more complex by removing this
>sentence and I don't think the implementations would be more difficult
>either as one can expect that most of them will rely on existing XPath
>implementations.

Whether the full XPath needs to be supported or only a subset was raised as
an issue (CR-49) and the Working Group decided to use a subset. This will
be in the spec and it will be reviewed by a joint task force of the XSL and
XML Schema Working Groups.

>BTW, another point with which I am dissatisfied, although not directly
>related to the spec, is the fact that despite many requests to many
>people my request to be subscribed to this list hasn't been taken into
>account, making it very difficult for me to track the progress of this
>spec.

I don't now about subscription to the www-xml-schema-comments list, but
this list is publicly archived at W3C and you could view the messages.  But
for tracking the progress, the w3c-xml-schema-ig list is more appropriate
one because technical discussion takes place in that list.

Thank you.

Best regards,
Aki Yoshida

Received on Friday, 16 February 2001 08:10:57 UTC