Comments

• Alex H.

  2008-05-09 08:06:13

  I was going to raise this issue the other day, but you made the same mistake again today.
  
  BSI haven't been sued. What happened is that UKUUG requested a judicial review: that means they've asked a court to review whether or not the process BSI followed was correct (_not_ whether or not their conclusion was correct). Unlike a lawsuit, if they are granted a review and then the review goes against the BSI, there is no penalty. The decision is set aside, and BSI reconsiders it: they can of course reach the same pro-OOXML conclusion again.
  
  Don't expect the BSI review to happen before the end of this month. The administrative court has a large backlog of cases and is actually doing extra shifts at the moment to get through it: a review (if granted) would be largely symbolic.

• ZiggyFish

  2008-05-09 11:35:58

  Athouvh it's not a lawsuit as such it's still a inquiry into the conduct of the BSI and at anytime can revoke the ISO standardization of OOXML

• Roy Schestowitz

  2008-05-09 11:44:33

  I'll correct the text in any event. Thanks, Alex.

• AlexH

  2008-05-09 13:38:44

  ZiggyFish - I actually don't think that the ISO status of OOXML is under threat from that enquiry.
  
  We have this "appeals" time limit. An "appeal" doesn't mean "I want to change my vote", it means "there was a problem with this process". For this review to result in BSI issuing a formal appeal to ISO, the court has to request a judicial review, the review has to happen, the BSI decision has to be set aside, the BSI has to re-consider it's decision, the BSI then needs to come to a different decision *and then* decide to lodge an appeal.
  
  I just don't see that happening in three weeks.
  
  Personally, I doubt the decision will be set aside - I'm not even sure a review will be granted. The premise of the request - that the technical committee at BSI voted against OOXML - seems to be wrong, so the issue of the process being flawed seems to be moot.
  
  That's not to say I'm defending what happened at ISO; I think both ODF and OOXML have caused trouble at ISO (ODF for being developed outside of ISO and the ISO standard now effectively being obsolete, OOXML for the obvious pressures Microsoft & those invested in the Office ecosystem put on the process). I don't think BSI did anything wrong though, and I think overall it's better for OOXML to be in ISO control than Microsoft control - people forget that Microsoft and ECMA no longer control the development of the format. If that had happened with ODF, I don't think OOXML would have been approved.

• Roy Bixler

  2008-05-09 14:00:54

  Maybe a better way to phrase it would be that the BSI vote was "formally challenged"? This may not result in BSI changing their vote, but it does shine a light on the whole process and show that there was strong opposition to the BSI's stance on OOXML. I can only hope that other countries will follow this example if, as it seems, some other national standards bodies like Norway, Poland, Germany, etc. remain defiant that "they did nothing wrong." If the latter is true, then I think their processes leave a *lot* to be desired. I even recall that someone pro-OOXML like Jesper Stoclund agreeing that the ISO standardisation process itself is weak and relies heavily on the integrity of the national standards bodies, which seem highly suspect in a number of cases.

• Roy Schestowitz

  2008-05-09 14:02:00

  Alex,
  
  Microsoft does not obey what's in ISO's hands, regardless. It never intended to. It said so explicitly over half a year ago, so that last argument of yours is moot, IMHO.
  
  Moreover, your description of what happened there in pursuit for a standard underplays what was a "brutal and corrupt process" (Tim Bray's description and one of the biggest scams in computing history (another person's take). It was sheer corruption, I assure you as one who has watched this since 2006.

• AlexH

  2008-05-09 15:27:26

  Roy S,
  
  The Microsoft/ECMA proposal for maintenance wasn't accepted; the ISO plan is the one laid out with a maintenance committee responsible for the standard and another for harmonisation with ODF. So, you shouldn't confuse the current situation with what was previously put forward: if Microsoft don't play ball with the ISO maintenance process, then they can't claim to output an ISO standard document format. That would then make the entire process of standardising their format pointless: and so far, they've been pretty good at keeping up with the changes in the format as it went through previous standardisation.
  
  As for sheer corruption - I think that's the in the eye of the beholder. Every standard which goes through has significant corporate supports who have a vested interest in seeing it accepted. Take for example MPEG: it's an ISO standard, but it's heavily patented and you have to pay a pretty penny to use it. The latest MPEG isn't like some technical work of art, either - they basically took Apple Quicktime and documented it. Adobe did the same thing with PDF. ISO standards are actually, more often than not, just a proprietary standard which was later documented.
  
  Whoever lost the argument over OOXML would have complained bitterly about the process being corrupt, undue influence, etc. The basic math was that Microsoft has more friends interested in working with Office than IBM has friends working with ODF.

• Roy Schestowitz

  2008-05-09 15:35:15

  ODF is not IBM. That's just the same old Microsoft FUD (trying to warp the problem, making it a seem like a giant vs. giant duel).
  
  Regarding the rest, Microsoft won't care about ISO compliance. It's all about pretense and perception. It's about marketing.
  
  On the corruption bit, no... it was by all means corrupt. Maybe you just haven't watched it closely enough to see the bullying, the bribery, the blackmail, etc. It was disgusting, it was scandalous.

• AlexH

  2008-05-09 16:00:22

  I wasn't saying ODF is IBM (it's not; it's Sun ;) I was saying that the primary protagonists anti-OOXML within the ISO system were IBM or IBM-supported. You can choose not to believe that, but I know what I've seen.
  
  If you don't think Microsoft care about ISO compliance, I honestly think that you're misguided. Standards compliance isn't important to Microsoft's home and business customers, it is important to their Government customers. Government purchasers tend to be less affected by marketing, and if Microsoft claim ISO compliance when it doesn't exist I suspect they would be relatively well aware of that.
  
  As for bullying/bribery/blackmail.. I think that's too easy a retort, to be honest. There are stories on both sides and if you only listen to one side's stories then of course you come away with that impression. I think it's very easy to make allegations about the process when you disagree with the result; let's see how many of those check out factually and turn out to be true. The BSI case will be a very good example.

• Roy Bixler

  2008-05-09 17:26:30

  Here's a pretty good example of one aspect of the argument, which is the question "was OOXML really suitable for ISO's fast track approval?":
  
  http://lehors.wordpress.com/2008/03/26/clarification-on-what-the-fast-track-is-really-about/
  
  It seems that it's an honest matter of interpretation but I do feel that, if Stoclund's interpretation is correct, then there is little that is respectable in a standard that's been "blessed" by ISO. It would mean essentailly "we've cajoled enough national standards bodies and ISO officials to see things our way." It does not mean originally what I thought an ISO standard means, which would be that there is broad consensus among technical people around the world that the proposed specification is a mature and tried-and-true one.

• AlexH

  2008-05-09 18:49:33

  Roy B - indeed.
  
  You don't need to look through many ISO standards to disabuse yourself of the notion that they are technically brilliant often.
  
  For example, ISO 9660 - the CD format we all use (.iso files, right?) Would anyone who cared about doing things technically correctly limit file names to 8.3 format? Who other than DOS/CPM couldn't support long file names, and was anyone using DOS really burning CDs? If you wanted to "do things right", you'd just remove that restriction and not have it available in the standard.
  
  But that's not how standards work.