Proposed Response to FPDAM Part 1 BR-0001, et al

MURATA Makoto (FAMILY Given) eb2m-mrt at
Tue Jan 26 16:31:09 CET 2010

As convenor, I think that we should not spent too much time on this
since all of us anyway agree not to introduce any version attribute.
But I also would like to make the record straight.

I can now explain why I do not quite like the current wording.

Member bodies correctly pointed out that there are no easy and reliable
mechanisms for distinguishing Ecma-376:2006 documents and 
transitional ISO/IEC OOXML documents.  Implementations have to rely 
on heuristics:  software information recorded as part of OOXML documents.
Since such heuristsics are not very appropriate as part of ISO/IEC standards, 
we cannot say that no more mechanisms are needed. 

But we also believe that a versioning attribute would not allow easy and
reliable detection as long as we want to make most of the existing documents
conformant to both  Ecma-376:2006 and ISO/IEC 29500.  So, we are struck.
We gave up.

Concerns about possible conflicts with other mechanisms such as
namespaces and MCE are valid.  But I think that there concerns are 
not the biggest reason for not incorporating a version attribute as part 
of this FPDAM.  Conflicts can probably be minimized or avoided if the
version attribute is kept extremely simple (e.g., @ecma376_2006 = "false")

The biggest reason is that version attributes would not allow easy and
reliable detection.  For example, although presense of @ecma376_2006= "false"
would imply ISO/IEC 29500, absense would provide no information.


More information about the sc34wg4 mailing list