XSD dependency (urgent problem)

Although XSD 1.1 [1] went to CR some months before OWL and was scheduled
to end CR at the beginning of August like us, the preliminary
information I have so far indicates it will be some more months before
they are ready to go to PR.

As I understand it, we have three options:

    1.  Wait for them.  The one detail I heard was that they aim to get
        to PR "in the fall".  I interpret that as delaying us 2-4
        months, and possibly more.  We may be able to get better
        schedule data in a few days.  (It's been hard to make contact
        this month; I finally got a response yesterday.)  I think we
        could wait before going to PR, or before going to REC, but I
        think it's best to wait before going to PR so that we can still
        use Options 2 and 3, below, without doing a second PR.

    2.  Excerpt the relevant parts of their text.  I'm told this has
        been done before with XML specs at W3C.  The advice I got is to
        do it explicitely: to create a normative appendix that has all
        the parts of XSD 1.1 we need, and note why it's there,
        explaining that a future edition [2] of the OWL spec may remove
        it, after XSD 1.1 gets to REC.

    3.  Rework our spec to not depend on XSD 1.1.  This might include
        creating some parallel datatype spec, as in Option 2, but using
        a different namespace.  I'm not sure of the details, but I
        believe it would create a permanent partial-incompatibility
        between OWL 2 and XSD 1.1.  This option might require another
        Last Call and CR, depending on how deep the changes turn out to
        be.

I'm in favor of option 2, as I suspect are all of you.  The catch is
that we'll need approval from the Schema WG, and their approval will
depend on their confidence that the parts we're excerpting will not have
substantive changes before REC.

I think the way to proceed is to draft the appendix for option 2 and
show it to the Schema WG.  Then they can consider its stability, and
decide whether to support our use of it.  [If I were saying this in a
meeting, I expect Boris would complete the drafting before the ensuing
discussion ended.  :-)  ]

Procedurally, I see this as new information, leading us to revisit the
PR publication decision of August 5, allowing us to try Option 2, and
possibly even Option 3, so we can avoid Option 1.

     -- Sandro

[1] http://www.w3.org/TR/xmlschema11-2/

[2] "Editions" are minor revisions of recommendations, fixing errata,
    and do not go through the whole WD/LC/CR process.  The XML spec is on
    it's fifth edition: http://www.w3.org/TR/2008/REC-xml-20081126/

Received on Wednesday, 19 August 2009 13:46:55 UTC