RDF/A and ODF

[Bernd Schuster <schuster@mskgmbh.com>]

Hi all,
I followed the discussion of the last meeting with great interest and I aggree, that using RDF/A is a good approach to seamlessly integrate meta data into ODF documents.
What I'm concerned about is the "content" attribute in Ben's example concerning the cal:dstart, because it leads to a situation where the information which is human readable might be different from the information which is processed by a machine.
From 1dok.org's point of view this kind of redundancy must be avoided to make the document format robust, failsafe and reliable.

Especially in situations, where digital documents shall be used in mission critical applications such like medical expertises, governmental, legal, or business documents, the document format must guarantee that the information stored inside a document  is exactly the same on both the human readable and the machine side (presentation <-> semantic structure). In other words: The human being must be able to validate  the content of a document visually and must be shure, that he/she got what he/she sees, before he/she releases documents for further processing (e.g. payment of an invoice).

If not, we introduce a backdoor for intended or unintended inconsistencies which will reduce the acceptance of ODF in wide ranges of the community.

From my perspectice a solution to this issue might be to store the data information in machine readable format and delegate the presentation to the displaying application. The ODF  (I hope, my understandig of the ODF is right ;-) ) provides document fields in conjunction with special style elements for several data types which might be useful to solve this.

Regards

Bernd


-- 
mit freundlichem Gruß

Bernd Schuster

MSK Gesellschaft für Automatisierung mbH
Friedrich-Ebert-Allee 33
D-22869 Schenefeld

Tel.: +49 40 839286-10
Fax : +49 40 839286-29

eMail: schuster@mskgmbh.com