OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-comment] ODF schema is faulty (ODF all versions)


Michael,

> Regarding default values: We have dropped them already. I unfortunately
> missed to remove them from the current schema so far, but the TC already
> agreed to do so.

I'm happy to hear that default values will be dropped.

> Regarding replacing xsd:ID and xsd:IDREF with xsd:NCName: As James Clark
> pointed out, we then would not follow the recommendation in the xml:id
> specification.

Appendix D of the xml:id recommendation is non-normative.  I think
that it should be ignored here.

> Right now we are normatively referencing the xml:id specification and are
> following its recommendations for RNG. We are not (normatively or
> informative) referencing the RNG Compatibility specification.

If you use xsd:ID as part of RELAX NG schemas and expect checking
uniqueness of identifiers, you are using the "RELAX NG DTD
Compatibility" specification anyway.  The RELAX NG specification
never checks uniqueness of identifiers; it simply validates the lexical
aspects of xsd:ID.  This is clearly described in Clause 4 of the
"Guidelines for using W3C XML Schema Datatypes with RELAX NG".

http://www.oasis-open.org/committees/relax-ng/xsd.html#IDAOGYR

> Is what you request that we anyway break the recommendations of the xml-id
> specification, and instead make the schema compliant with the RNG
> Compatibility specification, which we are not referencing, and for which
> James Clark says that it might also be a good idea to revise it to be xml:id
> aware?

Because D.3 ("With RELAX NG Validation") in the xml:id specification
is non-normative.  It can be safely ignored even when you use xml:id.

John Cowan, in his comments on James' article, already pointed out
that xml:id processors do type assignment, not validation.  So, I
think that ODF application programs should be required to check for
uniqueness of ids and check for ID/IDREF consistency.

Incidentally, W3C XML Shema schemas in OOXML Part 1 never
uses xsd:ID or xsd:IDREF although identifiers are certainly used.
In my understanding, checking uniqueness of ids and checking for
ID/IDREF consistency are tasks of OOXML application programs.

Cheers,

Makoto


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]