W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > July to September 2002

Re: xsi:schemaVersion

From: Dan Chang <dtchang@us.ibm.com>
Date: Fri, 13 Sep 2002 14:14:58 -0700
To: "Noah Mendelsohn <noah_mendelsohn" <noah_mendelsohn@us.ibm.com>
Cc: www-xml-schema-comments@w3.org
Message-ID: <OFEC91E3B6.E59A49EC-ON88256C33.0072AB6C@boulder.ibm.com>


Thank you for pointing me to the TAG discussions. I see that the
xsi:schemaVersion suggestion has also been made by Tony Coates and there
are some support for it. Regardless of this, I think, quoting you, "this is
the time to choose a group to think seriously though at least some of the
versioning issues" and hopefully come up with 80-20 solutions. The reason
is that being able to have versioning without changing the namespace is of
importance to XML Schema repositories and XQuery processors.

Regards,  Dan

IBM Almaden Research Center/Silicon Valley Lab
Notes:     Dan Chang/Santa Teresa/IBM@IBMUS
Internet:  dtchang@us.ibm.com
Phone:    (408)-927-1714/(408)-463-2319
Office:     ARC(G1-018)/SVL(D168)

                      Noah Mendelsohn                                                                                                               
                      <noah_mendelsohn@        To:      "Dan Chang" <dtchang@us.ibm.com>                                                            
                      us.ibm.com>              cc:      www-xml-schema-comments@w3.org                                                              
                                               Subject: Re: xsi:schemaVersion(Document link: Dan Chang)                                             
                      09/13/2002 07:27                                                                                                              


There's coincidently a discussion of XML versioning springing up on the
Tag mailing list (TAG is essentially W3C's architectural steering
committee).  See [1].  For my own opinions, see a posting I made in that
thread last night [2].

You specifically suggest:

>> For this to be useful, I would like to
>> suggest that an xsi:schemaVersion be
>> defined for use in an XML document to
>> provide hints as to the version(s) of a
>> schema that may be used for
>> assessment.

I think this implies some assumptions about how schemas for different
versions are somehow registered so a processor could "find them".  The
usual way for finding things on the Web is URIs, and we'd have to be
careful that a schemaVersion wasn't effectively a non-URI based mechanism
for naming or searching for schema documents.

As it stands, the schema recommendation gives nearly total flexibility to
processors to find schemas in any manner desired, and new attributes are
allowed on schema document constructions, so it seems that certain
solutions could be implemented without changing the schema recommendation
itself.  They could be published as additional conventions to be agreed
upon and followed by processors.  Like Tim Bray (see thread referenced
below), I think this is a very hard problem and to be approached with some
caution.  I do agree it is important.  For the moment,further discussion
should probably be on the TAG list (note, however, that the TAG strongly
discourages rambling or repetitive discussions:  the idea there is not to
drill this into the ground, but just to help the TAG decide whether to
take any initiatives in this area.)  Thanks.


[1] http://lists.w3.org/Archives/Public/www-tag/2002Sep/0082.html
[2] http://lists.w3.org/Archives/Public/www-tag/2002Sep/0106.html

Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
Received on Friday, 13 September 2002 17:15:37 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:50:00 UTC