The OOXML BRM: Secrets and statistics

I’ve provided a few links in another entry to some of the negative views from the ISO/IEC Ballot Resolution Meeting for Microsoft’s OOXML specification (via ECMA). For the most part, I’m letting others comment, especially those who were in the meeting. I expect we’ll be hearing a lot more in the next few days from those who were in the room.

I was not in the room, though at times I was a few hundred feet away. I’m not quite sure, since I never got up to the third floor of the convention center in Geneva were the meeting was held. Like others who were not part of the meeting itself I heard my news second or third hand, but the statistics that are starting to become public about who voted and how, and just how few issues were considered are just stunning.

Of all the “condensed” resolutions from ECMA, approximately 82% were not discussed at all, including counter proposals for these same issues. That leaves about 18% that were either discussed and voted on or else voted on early in batch. (I’m fine with batch votes for minor typographical fixes.)

What an utter and predictable embarrassment.

Did anyone actually think that so many comments could be processed in only one week? Is this an example of the “process working”?

I’ve been asking around to get some opinions about how much money it cost people to fly to Geneva and then stay and eat there for the BRM.  Geneva is a very expensive city to visit and the US dollar, for example, is very weak compared to the Swiss Franc.

Many people arrived on Sunday and left on Saturday, though some Europeans could escape on Friday night. There were many people from Asia and South America there. Conservatively, the general agreement is that it probably cost about $3000 to attend the BRM for the week, on average. I understand that there were 107 delegates from 32 countries at the meeting. Let’s lower that to 100 and say it’s not unreasonable that delegates spent at least $300,000 in Geneva to work on the OOXML BRM. That’s not even counting the cost of most of them not doing their “day jobs” for the week.

You’re welcome to your own estimates and, as I said, some colleagues of mine think I’m way too low on that $3000/delegate.

What a waste of money for such a poor percentage of the work getting done to the extent it deserved.

What a waste of money to individuals and standards organizations who could have spent their resources far more productively on other work. Maybe people should have just stayed home, everything balloted en masse, and all of that $300,000 saved.

“Well, sir, we spent a week at great expense and only resolved 18% of the problems. But it looks good because we approved most of the rest of the proposals without discussing them in detail, including how they might impact the spec, contradict each other, or even be implementable.”

If this were a plane, would it even leave the ground?

If this were a medical device, would a patient survive its use?

If this were a car, would you drive it at highway speed?

If this was your information, would you bet that your citizens’ and your government’s data would be preserved for years to come?

Do you want to bet your job on that?

Do you want to bet your professional reputation on that?

Do you know enough about the state of OOXML today to place either of those bets wisely?

Note that there is a lot you’ll never really know about what happened at the BRM. While I believe there was an audio tape, that’s just not sufficient because 1) it’s not public and 2) it’s hard to fully and correctly identify who was speaking.

This should have been digitally recorded and streamed live to the open web. There was too much controversy with OOXML to hide behind closed doors. People deserved to understand who said and did what. People deserved to see the process in action. Instead, interested people around the world were cheated of their opportunity for knowledge because transparency, and hence openness, was not part of the plans for the OOXML BRM.

I said to some journalist last week that I believe that OOXML has provoked a crisis in the standards system. There is a lot to fix but we must not accept any of the “just approve OOXML and we agree to fix the problems later” line of argument. In my opinion, OOXML should not become an ISO standard as part of this Fast Track process, and we should immediately embark on a serious reform of the standards system that got us to this point today.

Just how much evidence do you need?