Good PR, poor arithmetic
Jason Matusow of Microsoft is echoing in his latest post some so called “facts” brought up by Jan van den Beld, the former Secretary General of Ecma, who claims in his blog posting that the volume of comments on OOXML compared to that of other specifications is “completely normal – one comment for every 4-6 pages”.
I’m sorry to burst their bubble (not!) but it is well known that the comments that were registered by National Bodies during the vote of September 2nd are based on partial reviews of the specification. This is simply because it was impossible for anybody to do a complete review of the 6000+ page document in the 5 months allocated by the fast track process (which was totally inappropriate in this case). Several NBs have said so and I don’t know of anyone who has had a chance and the courage to do a complete read through.
It doesn’t take much effort to then figure out that, given the general complaint for lack of time to review the specification thoroughly, not all possible comments have been raised and have been registered. It then naturally follows that if the number of comments registered against OOXML is the same as what is typically received for specifications that are much smaller and therefore much more thoroughly reviewed, it actually means the OOXML specification is the object of many more comments/problems than what is normal!!
So while these so called “facts” might constitute good PR for those that have for agenda to downplay the mediocrity of the OOXML specification it makes for very poor scientific work for sure.
It is really appalling to see what is being done here. First, Ecma and Microsoft limit the amount of comments that could be registered by NBs by abusing the fast track process. Then they use that fact as an alleged proof that their specification is of the same quality as other specifications.
3 Comments »
Leave a comment
About
Arnaud Le Hors 2024 – CC BY 4.0
After several years focusing on strategic and policy issues related to open source and standards, including in the emerging markets, I decided to get back to my roots and focus more on technical work. I am involved in several software standards and open source development efforts with IBM. I am currently the main representative for IBM at W3C and INCITS, and serve as the Vice Chair of the OpenSSF Technical Advisory Committee.
I am a Senior Technical Staff Member part of the Open Technologies group at IBM. I have been working on standards and open source for over 30 years, both as a staff member of the X Consortium and W3C, and as a representative for IBM. I have been involved in every aspect of the open technology development process: technical, strategic, political, and legal. I was editor of several key web specifications including HTML and DOM and was a pioneer of open source with the release of libXpm in 1990 (which is still in use today! 😊). I participated in several prominent open source projects including the X Window System and the Apache XML parser Xerces, and more recently the blockchain framework Hyperledger Fabric. I’ve chaired several technical groups at W3C and OASIS, such as the XML and Linked Data Working groups, as well as more recently the Hyperledger Technical Advisory Committee and the LF Edge Open Retail Reference Architecture group.
Arnaud's Blog 
Actually it is wel worth noting that most comments ens in by the national bodies were as a result from 3rd party findings (like IBM) and seemed to cover just about every part of the specification.
You suggest that their has been only a partial reveiw but actually several national bodies used aid of public review methods (unprecedented on this scale) which gathered a lot more reviewing power than any document standard was ever subjected to before.
Evidently a lot more scrutiny than the ODF specification for which a japanese translator even without checking the content has found 100 relativly minor editorial issues which is actually more issues than were send in by Japan on OOXML.
All national bodies aren’t alike. Each national body is free to decide how it comes to voting one way or another.
Some countries have well organized and documented structures with experts in different areas they can rely on. Other countries have much looser structure and little expertise. Some don’t even have a structure per se and the vote is simply made by some government official on an unknown basis.
The question of whether the comments cover every part of the specification is interesting but actually if, as you say, national bodies relied on the same comments generated by IBM and other third parties then it is actually more likely that they only cover parts of the spec. IBM for one did NOT review the whole specification in detail. There was just no time for that.
When/if somebody ever endeavors to translate the OOXML specification we’ll see how many editorial issues get raised and we’ll have something we can compare with. Translators obviously read documents in a much closer way than anybody else.
Interesting point about translations, which is true to my experience. However, it’s undeniable — sadly — that Microsoft has only used politics to try to gain ISO approval for MS-OOXML rather than technical merit.
Maybe someday, someone will write a book on how Microsoft made every wrong turn possible with what has become “the MS-OOXML debacle.”