W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > October to December 2000

RE: LC-16 ( LC-132 ): Allow arbitrary order with occurrence > 1

From: <Noah_Mendelsohn@lotus.com>
Date: Mon, 16 Oct 2000 15:20:44 -0400
To: <mtimmerm@opentext.com>
Cc: www-xml-schema-comments@w3.org
Message-ID: <OF18795A40.51C7F0A3-ON8525697A.0069E480@lotus.com>
Matt Timmermans writes:

>> I will assume that all groups with all maxOccurs <= 1 are 
>> useful in at least some situations, simply because we 
>> include them in Schema 

This model is useful for a broad range of cases involving data.  Many 
storage systems (our own Lotus Notes/Domino product comes to mind) have 
the characteristic that named fields can occur at most once in a document, 
with order being insignificant.  When you store an email in Notes, there 
is a "From" field, a "To" field, etc., but order is not significant (we 
have forms that work like stylesheets to lay out a convenient presentation 
for users, but that is a different issue.)  The columns in a relational 
database are similar:  the name is significant, the order is not, and 
there is at most one with a given name.

Indeed, one of the reasons that I was so anxious to get <all> into the 
language is that I saw many real-world data DTDs with content models like:

        (name | rank | serial-number)*

in which, at the application level, it was clear that either exactly one 
or at most one of each was expected.  The repetition was done merely as an 
approximation to order-independence.  The problem is that a design time 
tool reading the DTD cannot see this restriction, and cannot establish the 
obvious mapping to a database or programming language structure.  The DTD 
implies that multiplicity can occur, when in fact it cannot. 

Another way to look at it is that our current <all> rules happen to map 
exactly to the rules for attributes on an element:  at most once, order 
does not matter.  These rules have surely proven useful in that context. 
In data scenarious, many sets of elements are used to model 
characteristics that feel like (structured) attributes of the parent.

------------------------------------------------------------------------
Noah Mendelsohn                                    Voice: 1-617-693-4036
Lotus Development Corp.                            Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------------
Received on Monday, 16 October 2000 15:26:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:12:48 GMT