Or at least that’s what ISO’s Secretary General says. [I had hoped to stop writing about this subject, sigh]. There are multiple appeals against OOXML; let’s try to read the tea-leaves without too many guttural snickers.

The Story Thus Far · When Microsoft decided to ram OOXML through the ISO “fast-track” process, a number of voices spoke up saying that would be inappropriate given the scale of the spec, and that doing this would be bad for the ISO and for the industry.

ISO thought it over (well, one assumes they must have) and decided that no, fast-track would be OK and they’d proceed, making up ad-hoc process as necessary. Predictably, and as predicted, the process was rife with corruption and bullying; at the end of the day, the pretence of carefully evaluating standards-ware had mostly worn away and it became a straightforwardly-nasty political dogfight between Microsoft and its allies on one side, and anti-Microsoft partisans on the other.

After Microsoft won, there were appeals from four countries saying, mostly, that the process was inappropriate given the scale of the spec, and had damaged ISO’s image.

Next? · ISO will examine the appeals. Let’s bear in mind that a favorable outcome for the appeals would mean, in effect, ISO acknowledging that they’d made a big high-level mistake. And then let’s not hold our breath waiting for transparency or neutrality.

Especially when ISO’s head honcho Alan Bryden goes on the record to say (quoting from Reuters) that “criticisms that a fast-track process was abused to rush through the Microsoft standard were unfounded” (note Reuters doesn’t use quotation marks, so presumably they’re summarizing Bryden).

Um, excuse me, doesn’t it seem wildly inappropriate for the chief executive of an allegedly-neutral international agency to comment dismissively on an in-progress appeal? If I were on ISO’s Board of Directors or equivalent, I’d be hauling Mr. Bryden in right now for a short unpleasant interview.

Whatever; the damage is done. I really hope my personal impression, based on the OOXML experience, that ISO alternates between bumbling and whorishness, is wrong. The world needs a reasonably competent and transparent standards organization whose integrity is not a standing joke.


Comment feed for ongoing:Comments feed

From: John Cowan (Jun 30 2008, at 13:34)

You have to judge ISO WGs separately. JTC1/SC2/WG2 (the ISO side of the Unicode or UTC process) is consistently sensible, professional, and responsive. Other groups dealing with related subjects (not close by on the ISO tree, though) are complete cock-ups. I hear only good things of the programming language standardizers, too.


From: PJ (Jun 30 2008, at 13:55)

The IETF is real and respected, but not really appropriate. You need to start one, call it the the 'Whole World Task Force' .. hrm, I suspect that WWTF is not an acronym an organization that wants to be taken seriously should use.

More seriously, I've always liked the IETF's pragmatic "rough consensus and running code" rules, and wish more standards organizations would place a similar emphasis on actually getting things done.


From: Craig A. Eddy (Jun 30 2008, at 14:59)

I'm sorry, but I don't hold any hope for ISO actually admitting they have made a mistake. They have managed to do the impossible: they've made a sow's ear out of a silk purse. Unless and until they manage to straighten out the mess they've achieved, I can no longer trust anything that they propose as a standard without outside confirmation from a neutral body. As there appears to be no neutral body, in this case, that makes things a bit more difficult.


From: FRank Daley (Jun 30 2008, at 15:37)

Thank you for highlighting the ongoing and highly questionable methods used by Microsoft to garner support for its OOXML format.


From: Tony Fisk (Jun 30 2008, at 15:40)

I'm afraid my response to the whole sorry saga is a 'guttural' snipe:

Two months on, and It Smells Ook alright!


From: Andre (Jul 01 2008, at 00:34)

"The world needs a reasonably competent and transparent standards organization whose integrity is not a standing joke."

That is very blunt or as Bryden himself admits:

"Irrespective of the outcome of the current appeals, we are confident that the robustness of the system will again lead to the answer the market place wishes to see and, in fact, **reinforce** ISO's credibility,"

Now, I personally believe that Bryden should not take the criticism against the standard process as directed towards ISO. ISO is not the bully here. While ISO did show a lack of leadership you can well make the point that OOXML as a standard is good enough. But saying multiple standards are fine makes him lose face.

Diplomats like Bryden put a kafkaesque nightmare on critics of the process. If you complain diplomatically about abuse of process a diplomatic ISO's governance does not show the leadership to preserve raison d'etat. If you complain vocally ISO governance regards the messenger as the problem but at least you make your point. Either way we have a leadership crisis of ISO because of its "Chinese" habbits. If ISO staff gets under public criticism then because they made their choice not to walk their talk.

Just look how an established conservative party in standardisation as IBM lost trust in ISO and went more and more offensive.

But you can also argue that ISO needs criticism: "ISO Secretary-General Alan Bryden told Reuters the uproar that erupted over the voting process was largely due to a poor understanding of the way that ISO works". This is the sort of arogance we know from other technocrats such as those that ran the Soviet Union. On the contrary the process demonstrated the wisdom of crowds; when the internet public observes the bully it cannot get away with it. It is public scrutiny to be proud of and to be embraced. Civil society and public sphere at work; Entering an institution that is usually so boring and technocratic that it mostly gets out of touch with that democratic element.

"Some of the negative publicity is quite extreme," writes Bryden

but how the whole ISO process went turned people into making more and more extreme comments. This is a breath of fresh air for ISO. Revolution does only happen where an institution is unwilling or unable to perform reforms. And I trust ISO to perform the necessary steps despite Bryden's comments.


From: Alex Brown (Jul 01 2008, at 11:59)

@John Cowan

One of the troubles with the OOXML project has been that it never went through a WG (working group) or even an SC (sub-committee). In that respect, Fast Tracked standards are completely unlike "normal" ISO/IEC standards which spend a lot of time getting worked over by ISO/IEC technical experts.

- Alex.


From: Rick Jelliffe (Jul 03 2008, at 00:03)

JTC1 already looked at various issues raised by October last year, at the time of the Brisbane/Gold Coast plenary, including the various claims of impropriety at that stage, and found no actionable substance to them.

And there had been JTC1/ITTF determinations already made on various issues brought up during the contradiction period. For example, the issue that the size of the standard could be used retrospectively to override the clear statement in the JTC1 directives that there criteria for fast-tracking were the proposer's business was rejected.

So, without new information, merely restating objections already made during the contradiction period, during the initial ballot, at the JTC1 meeting, or at the final NB vote, will go no-where. How could it?

I think it is quite unfair to say that ISO made "a big mistake" because firstly, it confuses which parties are involved. JTC1 is the body which makes up the rules. ITTF is the body which administers the technical challenges. The ISO secretariat administers the process. (The SCs are pretty well out of the picture.)

ITTF has to interpret the JTC1 Directives, and some of them are deliberately vague to allow discretion and prevent procedural mummification (e.g. the conduct of meetings), others are vague due to bad drafting or new situtations, others are vague because it seemed obvious to the drafters what the "ISO" or "JTC1" context was, but there are many other places where the Directives are quite clear. In general, the high level is clear, the low level is played by ear, just like any other body.

But while ITTF may certainly interpret and fill in gaps and finesse and develop interpretation-precedents for the JTC1 Directives consonant with the organization's goals, they cannot just make up new Directives or ignore clear ones.

Where there is a roadblock, the joint ISO/IEC Secretaries-General do have some extra discretion in the Directives (e.g. to allow extra deadlines).

The mistake was JTC1 for allowing any fast-track in the first place. It may be OK for a NB to vote no for a standard because it doesn't agree with the JTC1 Directives, but it cannot challenge the standard on those grounds. It has to challenge the Directives at JTC1.

What Tim sees as "ad hoc" I see as "discretionary". A voluntary standard for describing an existing technology is simply a different beast than a standard for making a perfect or best-practice technology. I think in Tim's mind is some idea that you cannot have a worthwhile standard for a bad technology, whereas sometimes it is exactly the bad technologies which need all their uglinesses and oddnesses exposed, documented and sifted though: you want to try to get towards an excellent standard even for crappy technology.

The mouse in room that no-one mentions is that ISO/OASIS ODF 1.0 lowered the bar for what is acceptable in a fast-tracked standard so significantly, allowing only the most general descriptions and gaps. If not-enough information is OK, why is too-much information not OK? I am not ODF bashing in saying that, nor promoting OOXML.

But I think there is a large aspect of the DIS29500 process and ballot being ultimately one about procedural fairness: should we apply one set of easy rules to an external technology from people we like while another, harsher set to an external technology from people we don't like? The "two wrongs don't make a right" argument didn't win, instead the consistency and equity argument won, IMHO.


From: MURATA Makoto (Jul 06 2008, at 08:08)

OOXML reminds me of Unicode. It is not perfect, but it is here to stay. I am thus happy to see that OOXML has been standardized and placed under the control of SC34.

There have been no practical alternatives to Unicode. Although ODF exists, I am not impressed by the technical and editorial quality of the ODF standard, and I am unhappy with the way ODF has been maintained (or not maintained). Note: I have spent a *lot* of time on ODF, since I am the convenor of the JIS ODF committee.

I have been continuously astonished by the lack of careful reviews of the ODF standard. People appear to praise ODF blindly only because they hate Microsoft. In the long run, the lack of constructive criticism will cause significant damage to ODF.


author · Dad · software · colophon · rights
picture of the day
June 30, 2008
· Technology (85 fragments)
· · Standards (19 more)
· · XML (135 more)
· Business (112 fragments)
· · Microsoft (19 more)


I am an employee of Amazon.com, but the opinions expressed here are my own, and no other party necessarily agrees with them.

A full disclosure of my professional interests is on the author page.