DR 09-0157 - WML: restriction on ordering of run properties

[rjelliffe at allette.com.au]

I am strongly in favour of a design rule that says that

1) When an element with element contents has child elements with no mutual
ordering semantic, they should be declared to be position-independent.

2) Child elements which are names or titles should go first (Head-body
pattern) for readability. However this can be a convention (or checked by
a subordinate Schematron schema).

Establishing the principle can help improve Open XML.

There is a school of thought I have heard that it is better to force an
order even for unordered elements. (IIRC correctly this is why in XSD
attribute declarations have to go after element declarations, not before
or after.) The only justification I heard was that this might save a bit
each time when compressed, which seems very 1970s.

(With xsd:all, remember that if you use it in a group, you cannot use that
group in elements that use xsd:choice or xsd:sequence too.  All or
nothing!)

Cheers
Rick Jelliffe

> Hi Jirka et al - is there room for an option D? I spoke to Zeyad (our
> WordprocessingML guy) and he proposed grouping the elements and using
> xsd:all with minOccurs=0. This ensures that the zero or more of the child
> elements can exist, in any order.
>
>        <xsd:group name="EG_RPrBase">
>              <xsd:all minOccurs="0">
>                     <xsd:element name="rStyle" type="CT_String" />
>                     <xsd:element name="rFonts" type="CT_Fonts" />
>                     <xsd:element name="b" type="CT_OnOff" />
>                     <!-- ... -->
>              </xsd:all>
>        </xsd:group>
>
> This change is specific to the schema and will avoid putting schema-style
> constraints into the prose, and making normative references to the Relax
> NG schemas.
>
> Chris
>
> -----Original Message-----
> From: Jirka Kosek [mailto:jirka at kosek.cz]
> Sent: 07 April 2010 08:48
> To: Shawn Villaron
> Cc: 'e-SC34-WG4 at ecma-international.org'
> Subject: Re: DR 09-0157 - WML: restriction on ordering of run properties
>
> Shawn Villaron wrote:
>
>> I was wondering if you thought you’d be able to provide updated text
>> for this DR response in preparation for the Thursday, April 8^th WG4
>> teleconference?
>
> The change is only in schemas. You can easily see diffed schemas at
>
> https://www.assembla.com/code/IS29500/subversion/changesets/89
>
> and use it in draft COR2 text.
>
> In my opinion there is no need for change in the text of the standard as
> there is nothing said about ordering of those properties and there is just
> reference to schemas.
>
> If we need some accompanying rationale for DCOR2, you can use something
> like:
>
> "This change allows run properties to appear in any order which is much
> more convenient for automatic OOXML content producers."
>
> Hear you tomorrow,
>
> 					Jirka
>
>
>
> --
> ------------------------------------------------------------------
>   Jirka Kosek      e-mail: jirka at kosek.cz      http://xmlguru.cz
> ------------------------------------------------------------------
>        Professional XML consulting and training services
>   DocBook customization, custom XSLT/XSL-FO document processing
> ------------------------------------------------------------------
>  OASIS DocBook TC member, W3C Invited Expert, ISO JTC1/SC34 member
> ------------------------------------------------------------------
>
>