2. OOXML Ballot Resolution Meeting: Why You Should Worry About All Comments

In English language literature, the usual example of a very long book is Leo Tolstoy’s War and Peace which, at least in the hardcover version to which I’ve linked, comes in at 1296 pages. Suppose you were an editor for this book, one of many, and you came up with 25 serious issues. However, when you went to the editorial meeting, you discovered that there were, in fact, hundreds of serious problems uncovered by the other editors. Should you worry only about your list?

The answer is no, and I think it is obvious why. If the author makes changes to the book based on the other editors’ comments, then the revised book is not at all the one you read. Not only might the changes affect the issues you had, but they might introduce many new problems, at least in your mind. It is the entire book with all the proposed changes that should concern you as a conscientious, thorough, and responsible editor.

There are, in fact, other longer books, but this one is still pretty big. It is dwarfed, however, by Microsoft’s OOXML specification which is four times longer at over 6000 pages. National bodies submitted comments during the ISO/IEC JTC1 balloting period that ended on September 2. That ballot failed on both measures by which it needed to succeed, significantly discrediting the specification, in my opinion.

Nevertheless, we are moving on to a ballot resolution meeting in Geneva at the end of February. Representatives from many of the countries that submitted comments will attend, though it is not clear yet how many and from which countries will make it there. In the last few days I’ve also heard that there is some question about whether one country can attend and have a proxy vote for another.

Just as in my War and Peace example above, a given national body must concern itself with all the comments submitted on OOXML, not just its own. The new document that will be produced after the comment resolution meeting, if it is successful, will be significantly different from what was reviewed before September 2. Changes can be made that affect your own comments as well as modify the document in serious ways in other areas. If you do not review all the comments, in my opinion, you are not being thorough and you should not possibly vote in favor of the specification.

Now I’ve been told that some countries are being urged to only look at their comments by the supporter(s) of OOXML. Of course they would suggest this, because their only motivation is to try to get OOXML passed as a JTC1 standard. Resist this.

A reasonable excuse might seem to be “but we don’t have time to review all the comments.” Well, reasonable perhaps but not acceptable. This specification was too big to be analyzed fully for the Contradiction Period last January, but nevertheless it was pushed through into the Ballot Period. At over 6000 pages, it was still too big for many people to fully analyze it in those 5 months.

At this point there are thousands of comments and just one week in Geneva to resolve them as a group, though Microsoft and ECMA’s “solutions” will be delivered in the middle of January – still too tight a schedule for something of this sheer magnitude.

If this specification is too big to analyze and too big to process all the comments, it should not be successful at any point now in this too short Fast Track process. It should never have been placed in this process in the first place, but we’re here.

If you cannot process all the comments deeply and diligently, then reject OOXML. If you are still not satisfied for any reason that it makes sense for OOXML to be an ISO standard, regardless of Microsoft’s market share, then reject OOXML. If you fundamentally believe that OOXML is a bad idea built for the wrong reasons, reject OOXML.

Do not satisfy yourself that looking at some small subset of the overall comment resolutions will make this whole immense specification somehow perfect and deserving of having the ISO mark. That mark should mean quality. As part of the process, you need to either be sure that OOXML fully deserves it, or else is rejected. This might be your last chance.

Don’t set a bad precedent when it comes to ensuring that ISO standards must be of the highest quality possible, particularly if the length of the specification or the number of comments is large.

Also See: 1. OOXML Ballot Resolution Meeting: Questions and Answers