Preliminary Google reply to DIS 29500 : the consideration of ECMA-376 OOXML for ISO standardization.
where Jeremy Allison (@Google) complains about OOXML mainly because of 2 technical reasons:
Google wrote:No useful guide to specification. Although it has made the specification available, Microsoft has made
it impossible to find the relevant information swiftly as the specification is over 6,000 pages long.
Google wrote:Continued use of binary code. In addition, the specification continues to use binary code although the
significant advantage of XML is that it is generally text based and therefore humanly readable.
In addition, many of these binary sections refer back to the Windows architecture which is unknown to
software engineers outside Microsoft, ...
As an example Jeremy Allison cites a binary font declaration and mentions
• Paragraph conditional formatting
• Table cell conditional formatting
• Table row conditional formatting
• Table style conditional formatting settings exception
• Pane format filter.
IMHO, the main conclusion on the technical side of the problem is this:
Although OOXML may formally comply with Ecma, it was clearly not designed with an “open” spirit.
Comparing the current with the future situation, interoperability is likely to become more difficult
instead of easier. The implementation of a fully compatible ODF importer (the current efforts regarding
.doc and .xls) is not an easy task, but it is dwarfed by the implementation of a fully compatible
OOXML importer, which we estimate to take something between 50 – 500 person years, or even
longer. Therefore, although it is theoretically possible to generate an OOXML document, this document
will probably only use a very small subset of the standard.
Then he states the judicial implications of what MS refers to as “Open Specification Promise”.
Hundreds of man years for an acceptable import filter
(leaving aside any export filter) and no guarantee (just a promise) that you won't be sued by MS or even third parties!