Re: [oiic-formation-discuss] Profiling for Interoperability

["Dave Pawson" <dave.pawson@gmail.com>]

2008/8/10 Dennis E. Hamilton <dennis.hamilton@acm.org>:
> I see from Dave's comments that I need to clarify a couple of items.
>
> 1. INTERCHANGE-AGREEMENT PROFILES (or whatever better name works - "document application profile" has been used for similar purposes in the distant past, but I don't want to prejudice thinking with that particular approach).

exchange profile has been used here, but I get the meaning.



> 1.2 Because of the necessarily-intimate connection of this kind of feature with the definition of the document format itself, I would expect that providing a standard interoperability-agreement profile mechanism would have to be incorporated in the document-format specification, at least as a supplementary capability.

+1

>  That is what had me think this is more for the document-format technical committee rather than the OIC, although OIC might
> provide important foundation as part of its more-generic conformance and interoperability exploration.

The harder part would be gaining agreement on what is needed (say a
schema modification) by sufficient
number of users to define such an interchange profile.  Quite possibly
different for any pair
of orgs exchanging data.


>
> 1.3 I'm not fussy about this.  I am not disturbed that it be out of scope for OIC.  I simply want to raise a cautionary consideration against cutting oneself off from the OIC work being evolvable toward such a capability by someone.

+1 The iic could contribute. It could only provide a recommendation though.


>
> 1.4 Because this is about implementation of a kind, I disagree with Dave.  I think OIC might inform such a development but that its provision is out of scope and its institutionalization would be a matter for the ODF TC to the extent that the document format must enable it.

I'm OK with that.


 > I think I failed to indicate how much I envisioned that processor
implementation would be impacted by normative
>  requirements and supporting markup.

I don't understand that.


>
> 2.  PROCESSORS AS VALIDATORS/DEMONSTRATORS OF CONFORMANCE AND INTEROPERABILITY

> 2.3 It struck me since that an interesting reference implementation would be one that was incorporated in an automated conformance and interoperability confirmation fixture and that demonstrated how a product might support its own validation in this way, too.

Hell of a lot of work Dennis? I'm sure you'd get NIH responses from
current vendors.
It's been suggested, but I'm not sure how to take this forward unless
you chose vendor X,
in which case vendor Y would scream.


regards




-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk