Re: [Simple] Presence Relax NG schemas

[Aki Niemi <aki.niemi at nokia.com>]

Perhaps this is exactly the type of activity that should take place in 
the future SIMPLEt events. As I understand the issue, creating an über 
schema from the various PIDF + extension schemas isn't exactly straight 
forward (because of the UPA issue). So I would consider this as a helper 
in doing document validation for implementors.

As an aside, I use the xml2rfc tool as probably most of you do, and I 
use Relax NG schema (that I converted from Marshall's DTD) to validate 
my drafts. And it seems to work fine.

I think the Relax NG schema should simply be something that implementors 
can take, validate their presence documents and (after the dust settles) 
see what happened vs. the XML schema validation. If there are 
differencies, then that info is probably very valuable.

There is a bigger issue here: do we mandate that the presence documents 
are always validated using XML schema only, or is e.g. Relax NG also 
allowed?

Cheers,
Aki

ext Jonathan Rosenberg wrote:
> I don't have any particular views on the pros/cons of XML schema vs. 
> relax NG. However, my concern about this is whether there is value in 
> having TWO definitions of the grammar for these documents. It seems that 
> this is a recipe for interoperability problems. An XML-schema-using 
> implementation might define a document that is valid according to the 
> schema, but invalid according to the RELAX-NG definition. That would be 
> a problem. If that never happens, and neither can the converse (a 
> RELAX-NG compliant document that is not schema compliant), then the 
> RELAX-NG description is isomorphic to the XML schema, and so what is the 
> value exactly?
>
> -Jonathan R.
>
> Jari Urpalainen wrote:
>
> > Hi all!
> >
> > I have submitted a draft 
> > <http://www.ietf.org/internet-drafts/draft-urpalainen-simple-presence-relaxng-00.txt> 
> >
> > (targeted as informational). This memo describes PIDF + extension
> > schemas written with the Relax NG schema language. These were
> > produced because with the W3C schemas it is not possible to write
> > reasonable combination presence schemas doing on-the-fly validation
> > (because of the paranoid UPA constraint specified in the appendix H
> > of W3C structures spec). So these are meant for publishers of
> > presence information and the rules given in these schemas are thus
> > stricter than the W3C schemas. Therefore, instance documents that
> > validate with these shall also validate with the W3C schemas. The set
> > incluces schemas for PIDF, datamodel, RPID, CIPID and CAPS. There are
> > still some minor differencies between these and PIDF extension
> > schemas (some "last minute" changes in I-D's). So these are not meant
> > to replace W3C schemas, they are just produced to decrease
> > interoperability problems with "real" implementations. W3C Schema 1.1
> > spec may eventually have a proper wildcard definition etc., but
> > before it happens these schemas can be utilized.
> >
> > Any comments ? Jari
> >
> >
> >
> >
> >
> > _______________________________________________ Simple mailing list 
> > Simple at ietf.org https://www1.ietf.org/mailman/listinfo/simple

_______________________________________________
Simple mailing list
Simple at ietf.org
https://www1.ietf.org/mailman/listinfo/simple