The Standards Blog

OOXML Payback Time as Global Standards Work in SC 34 "Grinds to a Halt"

OpenDocument and OOXML

One of the more egregious behaviors observed in the recent vote on OOXML was the sudden and last minute surge to join not only various National Bodies just before they voted on OOXML, but also the relevant committee of ISO/IEC for the same purpose. At the latter level, not one but two unusual membership changes occurred. During the voting period, more and more countries joined SC 34, the committee within ISO/IEC's Joint Technical Committee 1 (JTC1) that addresses document formats, at the Observer (O) level. Then, in the final weeks and days before the voting closed, many of these new members as well as many longer term members suddenly upgraded their status to Principal ((P) membership, thereby gaining greater influence in the final vote under the complex rules under which the committee operates (those rules are described in detail here).

 

 

SC 34 is one of the more important and active committees in JTC1, and has a constant stream of standards under active consideration and balloting.  In anticipation of the OOXML vote, its membership surged – with 23 new National Body members, and the number of P members spiking by 11.   When almost all of the new members voted for adoption (most of those countries that were long term members voted against adoption, with comments), many felt that the standard setting process had been abused.

 

 

But unfortunately, the damage has not stopped there: since the OOXML ballot closed on September 2, not a single ballot has received enough votes to count in this important committee. Why? Because the last minute arrivals to SC 34 are not bothering to vote.

 

The resulting gridlock of this committee was as predictable as it is unfortunate. The extraordinarily large number of upgrades in the final months, and particularly in the final days, therefore seemed attributable not to an abiding investment and interest in the work of SC 34, but in the outcome of a single standards vote. That conclusion is now certain, given the voting performance of the upgraded members since they cast their votes on OOXML.

The specific problem arises from the rules under which JTC1 committees operate, which are intended to ensure that specifications do not become official standards unless there is sufficient interest in them, as well as adequate review, to merit issuance as global standards. One of those rules is that at every balloting stage, at least 50% of the P members eligible to vote must in fact return a ballot. Even this requirement, however, does not set a high bar, because a member is permitted to return a ballot of "Abstain," and inadequate review to form an opinion is accepted as a valid reason to abstain. As a result, returning a vote of "abstain" constitutes at best only the most minimal level of good citizenship.  

The result is that a very important committee has, in the words of its Secretariat Manager in frequent pleas to the non-responsive members, "ground to a halt." The impact is significant, since this is the committee that controls standards such as RELAX NG (ISO/IEC 19757 Part 2), Schematron (ISO/IEC 19757 Part 3) and Topic Maps (ISO/IEC 13250) – not to mention ODF and PDF (if will be interesting to see if participation increases when Microsoft's PDF-competing XML Paper Specification advances to SC 34 from Ecma, where it is currently in preparation).  It is also a committee that attracts top talent on behalf of its traditional members, such as Ken Holman, Jon Bosak, Murata Makoto, Steve Pepper, Patrick Durusau, Francis Cave, Martin Bryan, and Rick Jelliffe, to name only a few. [Update/Correction 10-22-07:  James King, of Adobe, tells me that the latest PDF submission is tracking through a different committee:  ISO TC171.

It's a sad story, and one that for now does not have a happy ending in sight. For now, those that want to advance the cause can only grow increasingly frustrated.  If you want to sample the depth of that frustration, read the excerpts reproduced below from Secretariat Manager Ken Holman's plaintive weekly memos. And if you know anyone involved in standard setting in Malta, Cyprus, Lebanon, Cote d'Ivoire (and so on), do everyone a favor and ask them to consider either voting, or dropping back to Observer membership.  

Sadly, even that level of citizenship has been lacking in the newly upgraded members, whose numbers have dramatically raised the number or P members required to vote in order to advance a standard towards final approval. While I'm told that 90% of committee votes have achieved the necessary 50% return in the past, the current numbers tell a far different story: the three most recent (SC 34 N 870, SC 34 872 and SC 34 N 874) have all failed because of P member apathy.   As I read the tallies at those links, only one recent P member responded to a single ballot, even after some ballots had been reissued for a second or even a third time. Had it not been necessary to include the new P members in the calculations, the second two votes would have passed (the first related to establishing a liaison relationship with another organization, and not a standard).  

 

 

* * * * * * * * * * * * * * * * * *  

8/26/2007

 

 

.

 

.

Speaking of contributions, please remember that participating members have an obligation to participate in *all* ballots for the subcommittee and not just in some of the ballots.  There was a disappointing response to the last ballot, and so far a disappointing response to the ballot that is due in the week ahead. This week I am excited to report an additional five (count 'em!) participating memberships in our committee, adding Malta, Venezuela, Pakistan, Poland and Egypt to our ranks.  We also have a new observer membership from Indonesia.  We welcome all of these new members to our committee and look forward to their contributions

 

  9/7/2007

 

  This week we welcome six new P-members to JTC 1/SC 34, three of them new: Lebanon, New Zealand and South Africa, and three of them as transfers from O-membership: Romania, Sri Lanka and Chile.  We look forward to their contribution to our committee.

 

  Regarding contributions, last week I tried to remind all participating members that you have an obligation to participate in *all* ballots for the subcommittee and not just in some of the ballots you might be interested in.  My comments went unheeded and as a result the outstanding letter ballot 870 regarding Category A Liaison membership for the XML Guild failed due to a lack of response.

 

  Directives 9.1.10 explicitly indicates "if more than 50% of the P-members have not voted, the vote will have failed."  Until we get 50% of *all* P-members responding to ballots, the work of SC 34 will grind to a halt. Please consider your obligations of P-membership.

 

  In the next few days I will re-issue the liaison ballot, this time for 1 month.  This is the third time this question has been put forward to the membership of SC 34

 

  9/18/2007

 

  In the past week we have had two new ballots issued from WG 3, and a (second!!) reissue of the liaison request from the XML Guild.  Until we get sufficient response on committee ballots, the work of SC 34 will grind to a halt and all insufficiently-responded ballots will have to be reissued. This is a critically important issue to our committee and I commend all P-members to assess their responsibility to respond, and to consider changing their membership to Observer status if they do not plan on participating in the work of our committee. Please remember that abstention is a valid vote and will help the committee continue its progress.

 

  9/30/2007

 

  You will see at that link that (as of Sunday evening) only 7 member bodies of our 38 participating members have actually submitted a ballot response….Since the recent influx of new P-members to SC 34, not a single ballot has been able to be processed…

 

  It is critically important that P-members remember their obligations:  if we do not get 20 responses per ballot, the work of SC 34 will grind to a halt….If you do not plan to participate in the work of SC 34, please consider changing your membership to Observer status.  For those national bodies that joined in the interests of DIS 29500 Ecma 376 OOXML, remember that P-member/O-member status in SC 34 has no effect on attendance and voting at the Ballot Resolution Meeting being held in February.  If this is your only interest, it would serve SC 34 well to change your membership status to O-member.

 

  10/7/007

 

 

Regarding the two ballots that closed in the last week, even with all of the reminders I sent out we still fell short of the required number of ballot responses for either ballot to have their results considered. …Since the recent influx of new P-members to our committee, not a single ballot has had a sufficient number of responses to be considered.

 

 

 

[On a different topic, and if you still have a few more minutes:  as Rick Jelliffe noted when he linked to my last blog entry, there are some things that are much more important than whether your file format says <tomato> while mine says <tomato>.]

 

For further blog entries on ODF and OOXML, click here

 

subscribe to the free Consortium Standards Bulletin

 

 

 

Comments

Permalink
I am in general opposed to mitigating a bad situation when a root-cause correction is available.

(Andy may appreciate this in light of recent USSC decisions which openly concede that the law is broken but also that the solution must lie with Congress.)

In that light, any approach which presupposes good faith but has no response to bad faith is begging for abuse as soon as the stakes get high enough.

Where do you get the idea that this is caused by bad faith?  I know conspiracy theories are attractive, but the more reasonable explanation is just that this is just the normal mix of newbie and old-timer SNAFUs: some National Body organizations not prepping committee organizers enough, some committee organizers not realizing they had to vote, some local committee equivalents not meeting, the votes being at the bottom of the list of activity, some committees not having the expertise to follow up,  and some people tasked to review not reviewing. 

If we have no evidence of bad faith in others, all we have is prejudice which is hardly good faith!

The trouble is that people think ISO is somehow like a law-maker or a policeman, and they get upset that it is doesn't act to defeat monopoly. But ISO is more like a library that accepts books that are properly catalogued from anyone, or a postman that delivers mail to anyone where the mail is properly addressed, or a notary who signs for anyone provided the proper formalities are attended to. You don't want the postman to refuse to deliver mail to Microsoft because it has a monopoly in the OS and desktop applications market and to IBM because of its mainframe monopoly and to Apple because of its music monopoly (if it is at that stage). 

In order to keep up with a committee like SC34 well, NBs need nearly monthly meetings: this is because there are 3-monthly general meetings around the world and agenda items have to be submitted (6?) weeks in advance. So NBs need one meeting to discuss the outcomes of the general meeting and propose agenda items for the next one before that (6 week?) deadline; then another meeting when agenda is published to come to NB positions. Then to arrange delegates or proxies to the meeting.  During the meeting, general votes are not taken on the same day they are proposed, in order to allow consideration by NBs in different time zones; so it is good to have a local NB meeting at the same time as the 3 monthly meeting. That is the kind of commitment that a P-country should be considering, but only the most interested NBs achieve this. Some NBs are responders not initiators, and they only really need to meet once every three months or so in order to attend to the ballot comments of their experts. And if they are not interested, they just need to register abstain votes. 

Rick Jelliffe

"The trouble is that people think ISO is somehow like a law-maker or a policeman, and they get upset that it is doesn't act to defeat monopoly. But ISO is more like a library that accepts books that are properly catalogued from anyone, or a postman that delivers mail to anyone where the mail is properly addressed, or a notary who signs for anyone provided the proper formalities are attended to. You don't want the postman to refuse to deliver mail to Microsoft because it has a monopoly in the OS and desktop applications market and to IBM because of its mainframe monopoly and to Apple because of its music monopoly (if it is at that stage). "

They say: Never attribute to malice that which can be adequately explained by stupidity. Never attribute to stupidity that which can be attributed to ignorance.

So I will assume you just didn't know. For your information, ISO indeed has "law-making power". Under WTO rules, governments are legally required to insist on ISO standards in procurement whereever they exist. ISO is not a library which is required to store MS books whenever MS brings them. It is a standards organization that sets legally enforcable standards.

It also seems that you have internalized the MS political line that portraits OOXML vs ODF as a dog-fight between MS and Sun or IBM (whichever is convenient). ODF has a broad and very legitimate userbase which transcends simple MS/IBM fights. But that is a point you would never ever suspect when listening to anyone from MS. You seem to follow this line. People complain here for very good reasons about MS' actions compromizing the ISO. Your reaction is along the line of "there is no criminal evidence, it is just a question of competing companies, and IBM did too". All three are just untrue and irrelevant.

However you slice it, the ISO rules are there to ensure consensus standards. I have still to see anyone tabling any evidence that there was dissent against ODF when it passed ISO. There has been, and still is, massive opposition to MS OOXML. But instead of trying to work out a consensus, as is ISO's aim in life, MS is trying to force OOXML upon ISO. In the course of this, the operation of ISO has been compromized, as has been exemplified here.

This is not a court of law where you are innocent until proven guilty. In the light of quite some documented evidence, we are allowed to assume bad faith from a company convicted for this type of illegal actions many times over. I think anyone who has followed the OOXML saga can be excused from thinking they have been resorting to immoral practises again. Nothing you wrote here or elsewhere has even come close to disproving these suspicions.

So if you are convinced MS has done everything to ensure ISO is able to really evaluate ECMA/MS OOXML on its merrits, and all stakeholders could table their views in an orderly manner, please give us arguments. All this "it's a game and we were not caught breaking the rules" and "<$ENTETY> did too" are not valid arguments for the concerns aired in this, and many other blogs.

Winter

Please read http://www.iso.org/iso/standards_for_technical_regulations.pdf which is ISO's own statement on the matter of the legal status of standards. It is a recent document and tackles recent misconceptions. Note in particular section 7.1 <i>Ensuring no delegation of legislative responsibility</i>

<blockquote>Using ISO and IEC standards for technical regulation does not imply that regulators have reduced power or that they delegate responsibility to other parties. Regulators still have the power to change or update their legislation at any time, or to delete a reference if the standard loses its validity for the relevant legislation. Referencing ISO and IEC standards in technical regulation simply means that regulators make use of the existing consensus at international level.</blockquote>

ISO does not make laws or regulations for IT. It makes standards. Nations may regulate to use them, but that is their business, not necessitated from the ISO side or the fact of any ballot result. Indeed, from the ISO side comes the following warning that the ISO system runs with the expectation that nations will not blindly adopting standards: they need to have procedures in place to ensure the fitness for regulatory purpose:

<blockquote>Regulatory procedures are therefore required when approving references to standards regardless of which method of referencing is used.</blockquote>

(And IIRC, the WTO treaty obligations concern local versus international standards, especially local standards that block off trade. So I don't see how they apply to OOXML/ODF.)

Rick Jelliffe

Sorry Rick, but I read these treaties completely differently. They are binding, and at least in the EU, an ISO standard must be used if available.

WTO agreements:
http://www.wto.org/ english/docs_e/legal_e/final_e.htm

ANNEX 4(b) AGREEMENT ON GOVERNMENT PROCUREMENT

Article VI
Technical Specifications

1. Technical specifications laying down the characteristics of the products or services to be procured, such as quality, performance, safety and dimensions, symbols, terminology, packaging, marking and labelling, or the processes and methods for their production and requirements relating to conformity assessment procedures prescribed by procuring entities, shall not be prepared, adopted or applied with a view to, or with the effect of, creating unnecessary obstacles to international trade.

2. Technical specifications prescribed by procuring entities shall, where appropriate:
(a) be in terms of performance rather than design or descriptive characteristics; and
(b) be based on international standards, where such exist; otherwise, on national technical regulations

3, recognized national standards

4, or building codes.

(my emphasis, sorry for the mangled layout)

From:
http://www.wto.org/e nglish/docs_e/legal_e/gpr-94.pdf

EU procurement rules have a likewise rule on non-discrimination in technical specification (by way of the UK site): http://www.ogc.gov.uk/documents/NonDiscriminationInTechnicalSpecification s .pdf

Sorry, can't copy a quote, but it is section 2, page 2.

Winter

Note the "where appropriate". It is not the case either that regulators must adopt an ISO standard merely because it exists, nor that they must adopt all ISO standards in the same area or treat them equally, under that WTO.

In the EU document you kindly reference, it only speaks of national standards, as far as I can see. (An ISO standard is not automatically a national standard.)

Rick

"Note the "where appropriate". It is not the case either that regulators must adopt an ISO standard merely because it exists, nor that they must adopt all ISO standards in the same area or treat them equally, under that WTO."

The WTO treaties are not toothless. They can be, and have been, enforced in (international) court.

Likewise, "where appropriate" has legal power. If there is an international standard covering the procurement, it must be applied unless there are legally valid reasons NOT to apply. So any government procuring document processing software is legally required to use the ISO ODF standard. Or the relevant ISO OOXML if it gets accepted. Not to do so is in breech of the treaty. Suppliers can go to court, as far as I know.

"In the EU document you kindly reference, it only speaks of national standards, as far as I can see. (An ISO standard is not automatically a national standard.)"

In the EU (IANAL etc.), international treaties overrule national law. So any international standard overlapping a national standard will overrule the national standard by force of the WTO treaties. There is not that much choice legally.

But this is moot anyway. As if there are many countries with their own, national document standards. Except for China, that is.

Winter

Rick and Winter,

First, thanks for engaging in a fascinating debate here, which I've enjoyed reading and hesitate to interrupt.  That said, here are a few observations:

1.  Rick, I appreciate your first-hand perspective from SC 34 and elsewhere.  Despite having knocked around the standards world for 20 years, all of my knowledge about what actually goes on in ISO/iEC committees is second hand.  So thanks for the insight.

2.  Both:  Regarding the WTO:  What the WTO requires, and what interested parties actually do, are two different things, for better or worse.  Unlike the ITU, which deals in frequencies that wil reliably propagate, ISO/IEC standards are voluntary.  The WTO has no prosecutors - it's only a structure within which affected parties (nations, and industries that lobby their governments to speak on their behalf) can bring greivances and, after a long and drawn out process, eventually achieve some sort of  result - often with few teeth.  Recall the US tariffs on steel in the early Bush years.  Other countries brought (justified) complaints, and the US fought it up to the point where sanctions could have applied, and then backed down.  There's no way that (for example) someone in Sweden could bring a WTO complaint against their own government about (for example) OOXML - because they don't have any standing (this would be an intranational, not an international complaint, and the WTO is exclusively international).  Moreover, as Rick points out, neither specification is national.

3.  Regardless of what the WTO says, or what "when appropriate" legally might mean, it's what countries do that matters.  In this case, as noted in the previous point, what Europe chooses to do isn't a WTO issue.  It may be an issue under EU law - as interpreted by the EU.

4.  No government is going to adopt standards that nobody uses, even if ISO/IEC has adopted them (there are examles), whatever the rules say.

Finally, I should say that I don't know as much about what goes on in EU purchasing in fact, as compared to what (for example) the IDABC may say.

Net net, what I'd suggest is that you look to the past history of how the EU has actually exercised its purchasing power vis-a-vis ISO/IEC adopted standards.  That's probably the most significant data - and what the vendors involved will certainly be looking to, rather than the WTO.

All that said, I look forward to your next posts.

  -  Andy

I don't know much about EU law, but as a British citizen I can safely say that any law which the press could spin as "Eurocrats ban Word" would be completely politically untenable in the UK, and would be lobbied pretty hard against by our government.

I realise this is bad in many ways, including the way it confuses the file format with the application, but tabloid newspapers aren't renowned for their fairness or attention to technical details when a juicy headline is available.

- Andrew

"Net net, what I'd suggest is that you look to the past history of how the EU has actually exercised its purchasing power vis-a-vis ISO/IEC adopted standards.  That's probably the most significant data - and what the vendors involved will certainly be looking to, rather than the WTO."

I think the point is that I am from a small EU country that found out the hard way that it doesn't pay to game the EU competition commisioner, even when she is Neelie Kroes (whom we know from the past). Our administration lost hundreds of millions (if not even more) for flouting the EU rules on tenders.

While the US can afford to tumb its nose to the WTO, my country cannot do that. The retaliation of the bigger countries would be devastating. That does give us a different perspective about the strength of the WTO rules and the way ISO standards work out in procurement.

But I leave these matters to people who know more about that.

Winter

"Which country? Which industry? Which fines? Which year? Which rules? Which ISO standards were involved?  A URL would be very welcome, it is interesting to me."

The EU is rather open about their procurement policies.

Please read here about the infringement procedures:

http://ec.europa.eu/internal_market/publicprocurement/infringements_en.htm

This one is about computers:
http://europa.eu/rapid/pressReleasesAction.do?reference=IP/04/1210&format=HTML&aged=0&language=EN&guiLanguage=en
With a ISO standard for processors, there would have been no problem (actually, in all these cases they wanted to procure MS Windows compatible hardware, but they couldn't say that, could they).

As usual, standards are only a (small) part of the infringments. But refering to an ISO standard will get you off the hook in that part. Why? Because you can refer to the WTO treaties.

Winter

"Sorry, I'm confused again. I don't see anything about WTO treaties there, just about the EU Directives."

A few more of these posts, and I am have become an expert on international standards ;-)

Firt, the EU directives implement international treaties, they don't cite them.

So go read:
Proposal for a Directive of the European Parliament and of the Council on the coordination of procedures for the award of public supply contracts, public service contracts and public works contracts

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52000PC0275:EN:NOT

Article 24, Technical specifications contains the following section:


3. Technical specifications shall be formulated by referring to national standards implementing European standards, European technical approvals, common technical specifications, international standards or when these do not exist national standards or national technical approvals, or any other technical reference produced by European standardisation bodies as defined in Annex VI, provided that the reference is accompanied by the words "or equivalent".
(emphasis mine) Note the careful formulation to exclude national standards that do NOT implement international standards.

If you read this carefully, you can conclude that any procurement for Office Software MUST refer to ODF, or an EU equivalent. It cannot refer to OOXML unless OOXML becomes an international standard.

Or did I understand that wrong? Please educate me if I do.

Winter

For a start, in the same document it makes a distinction between technical specifications and standards. You must realize that governments typically treat ISO standards for health and safety very differently from those for IT.  Here is the definition in that document for standard:
-------------------------------------------

2. "standard" means a technical specification approved by a recognised standardising body for repeated or continuous application, compliance with which is not compulsory and which falls into one of the following categories:

- international standard: a standard adopted by an international standards organisation and made available to the general public,

- European standard: a standard adopted by a European standards organisation and made available to the general public,

- national standard: a standard adopted by a national standards organisation and made available to the general public;

-----------------------------------------------
Note the "not compulsory". Milk standards would be compulsory technical specifications., I suppose.

Now you have explored a few different positions in this thread: first that ISO has "law-making powers". As a voluntary standards body it does not, as their own recent material makes clear.  Next that WTO required adoption of ISO standards: it does not, all voluntary standards only need to be adopted as appropriate, which is not to say that some lazy countries might just do it that way. Next  that EU regulations require that ISO standards be mandated:

I don't pass myself off as any kind of EU expert, but in that document I see
----------------------------------------------
5.1. The current provisions on technical specifications [20] are designed to require public purchasers to define technical specifications by reference to an exhaustively listed set of instruments so as to avoid conferring any advantage on a given economic operator or giving preference to national production. These instruments are not only well known, transparent and publicly available but also represent, as far as possible, harmonisation of specifications at European or international level. The most important of these instruments is the standard - preferably European, international or, failing that, national. Other instruments which are more sector-specific (European Technical Approval for building products, as provided for in Directive 89/106/EEC) have also been retained as possible references.
-----------------------------------------------
clearly interposes a layer between standards and adoption ("the instruments") which, I suppose, is where the EU can select which standards are useful for what purpose and for public policy.

Obviously ISO standards are in a prime seat for adoption, and being left out from being an ISO standard may put you at a disadvantage, especially if procurement people fail in their due dilegence to identify and select the correct standard(s) for the particular job. But that is a long way from standards being laws or automatically mandatory or even required, yes?

Cheers
Rick

"Obviously ISO standards are in a prime seat for adoption, and being left out from being an ISO standard may put you at a disadvantage, especially if procurement people fail in their due dilegence to identify and select the correct standard(s) for the particular job. But that is a long way from standards being laws or automatically mandatory or even required, yes?"
I don't see your point about the "optional" nature of ISO standards.

The EU provisions state that a technical specification should, if appropriate, refer only to international standards (EU or other). As mentioned before, appropriate does not mean if you like, but if covered. As far as I know, this means that if there are no technical reasons why a standard should not apply, it must apply.

This very simply means that if a public body writes a tender for a document processing suit that can process MS OOXML, it is in breach of the EU rules.

Why, because this specification would not "avoid conferring any advantage on a given economic operator or giving preference to national production." However, after OOXML becomes an ISO standard, suddenly, the "international standard" rule applies, and it can be required. If Intel would have succeeded to get a performance indicator for its processors recognized as an ISO standard the countries requiring them would be mostly out of problems (http://europa.eu/rapid/pressReleasesAction.do?reference=IP/04/1210&format=HTML&aged=0&language=EN&guiLanguage=en). This is what I meant with the combination of ISO and WTO rules forcing lawmakers.

But this is not about government choice, but consumer choice.

In many countries, public bodies cannot legally force consumers to buy specific brands. So there are movements to force tax and other authorities to accept input form "other" platforms. However, in all these countries public bodies can force consumers to buy standard compliant gear. So, in the end, ISO standardization can force consumers towards specific brands. And if a county does this, higher levels (up to the EU) cannot interfere easily, because the rules clearly specify that technical specifications (Standards) based on international standards are legal.

Wrt your insistence that nations are not force to adopt ISO standards. That is bogus. Few nations are capable of drawing up their own document standard as little as they can come up with a new WiFi standard. They will have to relie on ISO for that. So, in practise, just a an ISO standard for WiFi forces me to use WiFi or nothing, an ISO standard for documents forces me to use software that processes that specific standard.

Within the trade rules of the EU, this means that MS has only to convince a single EU nation (Malta?) to adopt an ISO ratified OOXML, and all other nations must accept OOXML bids. And yes, a company can go to court to get a tender to accept "equivalent" standards.

As the original (Dutch) article wrote, ISO standards have become de facto law. In my opinion that is the reason ISO denied it.

Winter