ISO Miscounted Cuban OOXML Vote

An anonymous reader notes Groklaw's coverage of the apparent mix-up ISO made with Cuba's vote in the matter of recommending OOXML as a standard. Cuba apparently voted against OOXML in September, but ISO recorded their vote as a "yes" — which is odd on its face, as Microsoft is forbidden to sell any products in Cuba. The Cuban NB head has apparently now officially responded to the BRM, but Groklaw's PJ notes that verification remains problematical, and "...the bottom line to me is that a process that worked perfectly well when folks all trusted each other falls into chaos when there are allegations of dirty tricks or undue pressure."

Comments

obligatory

[shentino]

"the bottom line to me is that a process that worked perfectly well when folks all trusted each other falls into chaos when there are allegations of dirty tricks or undue pressure" ...well DUH...

No, NOT "Duh."

[spun]

Dealing with unfairness and undue pressure is a central theme in building societies and groups that work. If everyone were good and played fair, any system would work. We need social systems precisely because some people do not play fair. Thus, we have checks and balances in our American political system. Where are the checks and balances here?

The author is basically saying, the system is flawed because it does not take into account certain facts about human nature, and fails at one of the most basic tasks any socio-political system should strive to accomplish, namely limiting the ability of participants to put undue pressure on each other and use dirty tricks.

Re:No, NOT "Duh."

[Chris Burke]

If everyone were good and played fair, any system would work.

With the corollary that if nobody had any reason to cheat, then everyone would play fair, and the system would work.

I think this is part of why "communism" only seems to work on the scale of the "commune", where ultimately even the most corrupt person could, what, lord over the persons and crops of fifty people? Have the most sweet potato of any villager? Scale that up to the level of a nation-state, and suddenly taking control and abusing the system provides a lot more gains in wealth and power.

Similarly with ISO, in the past the system worked because, by and large, nobody had any significant reason to game the system entirely. Sure different companies had their reasons to promote their standard, but ultimately it was still about cooperation and interoperability. While I may be missing some cases, I feel confident stating that this is the first time a standard presented to ISO has the potential to make or break a multi-billion dollar monopoly.

So of course when that kind of cash is on the line, a system that before survived because there wasn't much incentive to abuse it is found to be completely vulnerable. Kind of like if that commune suddenly found itself sitting on a gold mine.

Re:Y'know...

[cloakable]

And those standards went through, I'd guess, on the non-fasttrack route? Like the SQL standard? That took years to go through.

The abuse here is trying to push OOXML through on fast track, when it's obvious to anyone following the process that this should take the same route as SQL, for example. But that wouldn't be quick enough for Microsoft to stem the organisations mandidating open standards to look at their options, and choose OpenDocument over OOXML.

Re:Y'know...

[belmolis]

Your characterization of the Groklaw crowd is quite inaccurate. Plenty of people there are familiar with standards processes. And yes, they know that there has been controversy before, but rarely has there been controversy of this magnitude, to my knowledge, NEVER in the fast-track process. The purpose of the fast-track process is to expedite the formalization of what are already de facto standards, which means standards that are well thought out and carefully written, that exist in multiple implementations, and on which there is substantial consensus. Microsoft's attempt to use the fast-track process for OOXML is outrageous given that OOXML is a bloated mess, has yet to be implemented by anyone, not even Microsoft, and is a single-company effort on which there is no consensus.

Rules always rely on trust

[microbox]

The author is basically saying, the system is flawed because it does not take into account certain facts about human nature,

I'd argue that it's impossible to build a system that will work when people don't respect it. For example - the third-party payola loophole farce. Another example - democracy might work when people respect it, but the rules mean nothing to "I have a PhD in violence" Mugabe.

In essenence, the rules themselves are only useful if they are followed in spirit. When they are not followed in spirit, then we need more clarifying rules until we come down to some basic rules that are followed in spirit. That's why are law books are so large - and it's still not large enough for people like Darl McBride... proof that the more we disrespect each other, the bigger the rule book needs to become.

The traditional solution is to turn your back to people fail to follow the spirit of the rules. You just tell them that they can go bother someone else. You can't force other people to learn ethics, and there'll always be that fuzzy area where the amorale can do horrible but legal things like deliberately spread disinformation about global warming. These people should be charged with treason, because they are subverting the public good.

When an untrustworthy entity enters a situation where a certain level of trust is already assumed (M$ and ISO), then the rulebook needs to catch up *a lot*.

Standards were ALWAYS cutthroat politcs

[Ungrounded Lightning]

PJ says: "...the bottom line to me is that a process that worked perfectly well when folks all trusted each other falls into chaos when there are allegations of dirty tricks or undue pressure."

But standards operations have ALWAYS been about cutthroat politics and dirty tricks to gain competitive advantage. (For instance: There's stuff in an international protocol standard from the '70s or so that was transparently-crufty weirdness a US delegation proposed to get the French to back down from something they didn't like - but the French instead embraced the cruft wholeheartedly and the US negotiators couldn't admit it was just a bluff...)

The ideal is to standardize exactly what you're already marketing (or are about to release), so you continue to sell it and become (or become more) the dominant and entrenched market player while everybody else is delayed while they make changes - and become incompatible with their previous prototypes or products. This is a massive advantage even if you DO have to give up your patent locks on the technology to make it into a standard.

What's different about this is just the scale and the ability of the multibillion-dollar gorilla to afford tactics that weren't cost-effective enough to be common.