DR 09-0295 ? WML: gridCols measurement units

[MURATA Makoto (FAMILY Given)]

First, let's see what was the decision in the BRM.  It was Resolution 5. 
The following text is extracted from SC 34 N 990.

Units of measurement 
FI-10 (R 705) 
Ecma: the goals of Edition 1 are satisfied by the granularity chosen; the improvement proposed 
should be worked out thoroughly during maintenance. 
 
	Resolution 5: The BRM accepts the editing instructions contained in 
	http://www.itscj.ipsj.or.jp/sc34/def/BRM/Response_FI-0010_measurements_v2.doc
	with “informal” being replaced by “informative”, 
	in replacement of R 364, R 19, R 705, R 362, R 620 and R 103?
	so resolved. 
	 
	Resolution 6: Additionally, the BRM accepts the editing
	instructions contained in  
	http://www.itscj.ipsj.or.jp/sc34/def/BRM/Response_FI-0010_percentages.doc
	but with the following changes: throughout the text, add the specific Unicode value U+0025 
	for the percentage sign; and 
	change all ranges to use the English words “from x up to [and
	including] infinity”, “inclu-
	sive[ly]”/”exclusive[ly]” etc.?so resolved. 

And what is http://www.itscj.ipsj.or.jp/sc34/def/BRM/Response_FI-0010_measurements_v2.doc? 
It is contained in 0989_reference_docs.zip, which is a part of SC34/N989.

> http://cid-c8ba0861dc5e4adc.office.live.com/view.aspx/Public%20Documents/2009/DR-09-0295.docx
> 
> This DR covers the fact that measurements units (a feature added to several elements at the BRM) 
>are optional on the elements to which they apply, and requests that
>they are made mandatory. 

I have thought that Resolution 5 mandates measurement units in S while
allowing raw numbers in T.  But I cannot find any suppporting evidence 
in Response_FI-0010_measurements_v2.doc.

>There's a core assumption that it was the intent of
>the BRM to make these attributes mandatory - I've done as much research
>as I could and as far as I can see the BRM voted on a proposal (written
>up by Tristan and Rex) which incorporated optional attributes but
>maintained the existing Ecma-376 measuring system as a default, if no
>units were specified. This is how the standard is written at the moment.

I believe that, at present,  measurement units are simply disallowed in 
many places.  Even if raw numbers are allowed, should we allow 
measurement units as well always?

> Aside from the BRM debate, I'm not in favour of taking the modifications 
>proposed this DR. There seems to be no inherent disadvantage to having
>default units for particular measurements, and I think it makes the
>migration path from transitional to strict implementations harder for no
>obvious technical benefit.

Rick said to me that we shouldn't force programmers  to remember
different measurement units for different places but allow 
them to explicitly specify their favorite measurement units especially 
when they are writing programs for emiting OOXML.

Cheers,
Makoto