Updated (twice): 9:30 AM and 2:30 PM
Last night, I projected that the OOXML vote in ISO/IEC JTC1 would fail (the New York Times predicted the opposite). I have now seen the official vote tally, and confirmed that the vote failed both tests for approval (details are included at the end of this blog entry). The official ISO announcement is here.
Microsoft issued a press release early this morning, seeking to put the best face on the OOXML in advance of the official announcement by ISO. The release is titled Strong Global Support for Open XML as It Enters Final Phase of ISO Standards Process, and subtitled, "Significant participation by National Bodies in ISO/IEC ratification process for Open XML; final decision expected in March 2008 at close of ballot resolution period."
The release focuses on the degree of participation (51 National Bodies), and level of "support" (74% of all qualified votes, without differentiating between P and O countries). It also refers to this level of support at "this preliminary stage of the process," and compares it "favorably" to the number of countries participating in the votes to consider ODF and PDF, but without mentioning percentage levels of support: the OpenDocument Format received a total of 31 votes - all to approve. Moreover, there were so few comments offered along with those votes that no Ballot Resolution Meeting was required.
The Microsoft press release tries to draw good news from the level of participation in the OOXML vote, stating perhaps more accurately than intended (out of 41 P members, 24 voted either against approval, or abstained) as follows:
This widespread participation and support is consistent with the rapid adoption of the Ecma Office Open XML file formats across multiple platforms and products from a wide range of IT vendors (including Apple, Novell, Corel, Sun, Microsoft, Java developers and Linux distributors), creating real value for IT users around the globe.
More tellingly, the cornerstone quote from Microsoft's Tom Robertson reads in part as follows,surely placing it in the running for winner of the "Pony in the Pile" award for today:
This preliminary vote is a milestone for the widespread adoption of the Open XML formats around the world for the benefit of millions of customers. Given how encouraging today’s results were, we believe that the final tally in early 2008 will result in the ratification of Open XML as an ISO standard.
The actual numbers for the final tally are now what you could rightly call "encouraging." All 41 P members voted, with the following breakdown: 17 yes, 15 no, and 9 abstain. Or, as the ISO press release more neutrally described the result:
Approval requires at least 2/3 (i.e. 66.66 %) of the votes cast by national bodies participating in ISO/IEC JTC 1 to be positive; and no more than 1/4 (i.e. 25 %) of the total number of national body votes cast negative. Neither of these criteria were achieved, with 53 % of votes cast by national bodies participating in ISO/IEC JTC 1 being positive and 26 % of national votes cast being negative.
Rather incredibly, some early articles in the press are buying the Microsoft press release at face value. Here is an article at PCPro, by Simon Aughton, which reads in part as follows:
The ISO has given preliminary backing to Office Open XML. A large majority of the international standards body voted to support ratification of Microsoft’s file format. Although the "yes" vote was short of the required majority, many of the members who voted no, including the UK, indicated they would approve OOXML as a standard once some technical issues have been addressed. Microsoft was naturally delighted with the result and welcomed the "invaluable technical comments designed to improve the specification"….The NoOOXML campaign, which had predicted a "no" vote, has yet to respond.
Although I have described the next steps in the past, it is interesting to see them described from an authoritative source, in this case, the ISO press release:
Comments that accompanied the votes will be discussed at a ballot resolution meeting (BRM) to be organized by the relevant subcommittee of ISO/IEC JTC 1 (SC 34, Document description and processing languages) in February 2008 in Geneva, Switzerland.
The objective of the meeting will be to review and seek consensus on possible modifications to the document in light of the comments received along with the votes. If the proposed modifications are such that national bodies then wish to withdraw their negative votes, and the above acceptance criteria are then met, the standard may proceed to publication.
Otherwise, the proposal will have failed and this fast-track procedure will be terminated. This would not preclude subsequent re-submission under the normal ISO/IEC standards development rules.
Finally, recall that I have consistently predicted that the vast majority of the O members that upgraded to P membership would vote in favor of approval, presumably inspired to do so at the urging of Microsoft. That expectation has been confirmed, although the motivation must continue to be inferred from the results, which are as follows: of the 11 upgrades, 9 voted to approve, or approve with comments, and only one voted against approval (Ecuador; Trinidad and Tobago abstained).
For further blog entries on , click here
subscribe to the free Consortium Standards Bulletin
* * * * * * * * * * * * * * * * *
Strong Global Support for Open XML as It Enters Final Phase of ISO Standards Process
Significant participation by National Bodies in ISO/IEC ratification process for Open XML; final decision expected in March 2008 at close of ballot resolution period.
September 04, 2007: 06:31 AM EST
REDMOND, Wash., Sept. 4 /PRNewswire-FirstCall/ — Today the International Organization for Standardization (ISO) released the results of the preliminary ballot to participating National Body members for the ISO/IEC DIS 29500 (Ecma 376 Office Open XML file formats) ratification process. The results show that 51 ISO members, representing 74 percent of all qualified votes, stated their support for ratification of Open XML. Along with their votes, the National Bodies also provided invaluable technical comments designed to improve the specification. Many of the remaining ISO members stated that they will support Open XML after their comments are addressed during the final phase of the process, which is expected to close in March 2008.
With at least 87 countries taking part in some way, the Open XML review represents an unprecedented level of participation in the standardization of a document format. Fifty-one ISO members voicing support at this preliminary stage of the process compares favorably with the 32 ISO members supporting Open Document Format (ODF) 1.0 at the end of its process and the 15 ISO members supporting PDF/A-1 at the end of its process. This widespread participation and support is consistent with the rapid adoption of the Ecma Office Open XML file formats across multiple platforms and products from a wide range of IT vendors (including Apple, Novell, Corel, Sun, Microsoft, Java developers and Linux distributors), creating real value for IT users around the globe.
"We are extremely delighted to see that 51 ISO members, representing 74 percent of the qualified votes, have already voiced their support for ISO ratification of Open XML, and that many others have indicated they will support ratification once their comments are resolved in the next phase of the ISO process," said Tom Robertson, general manager for Interoperability and Standards at Microsoft Corp. "This preliminary vote is a milestone for the widespread adoption of the Open XML formats around the world for the benefit of millions of customers. Given how encouraging today’s results were, we believe that the final tally in early 2008 will result in the ratification of Open XML as an ISO standard."
Technical Input Will Enhance the Standard
Today’s results represent the beginning of the third phase of the ISO/IEC process, called "ballot resolution," during which time Ecma International will respond to all comments that have been submitted by ISO National Bodies and provide them with a final opportunity to voice their support.
"Technical experts around the world have provided invaluable feedback and technical recommendations for evolving the format," Robertson said. "The high quality of the Open XML format will be improved as a result of this process, and we take seriously our role in working within the Ecma technical committee to address the comments received. We believe that the ISO National Bodies will be pleased with the results."
The ISO/IEC process is also designed to create maximum opportunities for ISO members to move from "no" to "yes" in this final phase, and we fully expect the total number of supporting votes to grow. "The objective of the standardization process, whether with Ecma International, ISO/IEC or another standards body, is ultimately to refine a specification and achieve a positive consensus around its formal adoption for the benefit of the entire industry," said Hugo Lueders, group director of EU Public Policy for the Computer Industry Technology Association. "Given the remarkable level of participation from the global standards community, the results from this preliminary ballot are very encouraging."
Although no date has been formally set, the final tally is likely to take place in March 2008. ISO/IEC requires that at least 75 percent of all "yes" or "no" votes (qualified votes) and at least two-thirds of "P" members that vote "yes" or "no" support ratification of a format in the Fast Track process. More information about the ISO/IEC JTC-1 process and its participants is available at http://www.iso.org/iso/standards_development/processes_and_procedures/iso_iec_ directives_and_iso_supplement.htm.
Widespread Support for Open XML
The Ecma Office Open XML file formats are being rapidly adopted across multiple platforms and products from a wide range of IT vendors, creating real value for IT users around the globe. Thousands of companies from 67 countries on six continents have raised their voices in support of Open XML and its ratification by ISO/IEC at http://www.openxmlcommunity.org. As well, the open standard has also been gaining broad adoption across the software industry for use on a variety of platforms – including Linux, Windows(R), Mac OS and the Palm OS. Independent software vendors (ISVs) and platform providers around the world – such as Apple, Corel, Sun, Microsoft and Novell – are developing solutions using Open XML. Many developers working with the formats are active contributors to http://www.openxmldeveloper.org.
Those working with Open XML can attest to the benefits of this open file format in the areas of file and data management, data recovery, interoperability with line-of-business systems, and the long-term preservation of documents. Open XML is optimized for the level of precision and detail that facilitates carrying forward billions of existing files. As well, Open XML file formats are uniquely capable of integrating other types of systems and data with Open XML documents, while maintaining a clean, simple separation of presentation (Open XML markup) and data (custom schemas and instances thereof). This means that organizations can use Open XML formats to report information from other applications and systems without having to translate it first, which is a key innovation for developers seeking to incorporate real- time business information into their documents, or those who seek to "tag" documents with their own categorization system to improve their understanding of its contents.
More information about Open XML, as well as available solutions using the open standard specification, is available at http://www.openxmlcommunity.org/momentum.aspx#technology.
Commitment to Interoperability
The Open XML Translator (http://www.sourceforge.net/projects/odf- converter) is one among many interoperability projects Microsoft has undertaken. We continue to work with customers and others in the industry to address the critical interoperability issues of our customers through the delivery of products that are interoperable by design, collaboration on interoperability projects, providing access to interoperability technologies, and standardization. Other evidence of collaboration includes our technical collaborations with AOL LLC and Yahoo! Inc. for instant messaging interoperability, the broad collaboration with Novell on virtualization, document formats and intellectual property, and the creation of the Interoperability Vendor Alliance. More information is available at http://www.microsoft.com/interop.
Founded in 1975, Microsoft is the worldwide leader in software, services and solutions that help people and businesses realize their full potential.
it would appear that the Microsoft minister for information has his sums wrong. They are claiming 51 yes votes at 74%. 51*(100/74)=69. They think that of 69 countries 51 voted yes, so that leaves 18 to abstain or vote no. I think we know about 24 no or abstain so far in the P list. There could be more in the O list. Maybe they are counting abstain as a vote in support?
Or maybe they uses Excel to calculate it.
Art
I have been over the maths many times. The only way I can get to Microsoft’s 74% in the press release and repeated in blogs is to add up the approvals, approvals with comments and abstentions to get 51 and then divide by the total of P and O countries to get 69. 51/69 is 74%.
Microsoft are claiming the abstentions they caused to be votes in support of their flawed standard.
Andy,
Combining the upcoming BRM and the voting results, how many "no’s" would microsoft have to flip in order to gain approval? Do the 2/3 and no more than 25% rule still apply? Thanks!
From my reading of the ISO regulations section 2.7.3, the final ballot would have exactly the same rules for majority as this one.
My WAG is still that they spend the next 10 years trying to address all the comments.
That, or sign up more "P" members. There are a lot of countries yet to be heard from, after all.
MS even don’t need to convince any NB to change their vote!
They just have to
– buy a few more votes (there are not a lot left, but surely enough to go from 74% (51 out of 69) to 75% of the valid votes – possible candidates are Algeria, Bahrein, Botswana, Estonia, Ethiopia, Libya, Mongolia, Oman, Saint Lucia, Sudan…) and
– upgrade a bunch of NB to P status, to reach 2/3 of the P votes.
Done. Game over. Everybody (except MS) loose.
Wait… This may not be correct. IIRC, only the NBs present during the BRM may vote in March ’08. So, MS will indeed have to
1. ensure that everybody participates to the BRM, and
2. convert a couple of No or Abstain votes to Yes.
The tricky part will probably be Step 1 : how to convince so many NBs to pay the travel and attend a one week technical discussion meeting, while some of them probably haven’t any expert that opened the OOXML file ?
Upgrade to P status is possible until the last minute, so this part should be easy.
Unless ISO/IEC rapidly cleanup their JTC1 membership rules… or the BRM is cancelled because to much change is needed in the 6000 pages in a too short timeframe.
Luc Bollen
This is not how I understand it. Votes that are casted remain until changed. If a NB doesn’t show up at the BRM, their vote remain. They only lose their ability to change it. I understand that typically NBs that vote yes don’t show up and those the vote no come to see if their comments are addressed adequately. But now we have NBs that voted "Yes with comments". Will they show up and change their votes if the comments are not addressed? This BRM may end up being quite spectacular.
Indeed, your explanation make sense…
But now we have NBs that voted "Yes with comments". Will they show up and change their votes if the comments are not addressed?
I think some will – particularly those that were suckered by Microsoft’s mis-information that a ‘yes with comments’ means that comments will be/must be addressed. Come to think of it, they may change their votes anyway – just because they were suckered…
Atleast the NYT corrected itself: http://www.iht.com/articles/2007/09/04/technology/msft.php
ISO News Release out now:
"A ballot on whether to publish the draft standard ISO/IEC DIS 29500, […] as an International Standard by ISO (International Organization for Standardization) and IEC (International Electrotechnical Commission) has not achieved the required number of votes for approval."
As Microsoft has been spinning that their stuffing is just the same that IBM is doing, I wanted to do this little experiment.
The only place I’ve found the votes of all the countries has been [1], so my numbers are based on it (feel free to correct me). It’s somewhat limited, as it doesn’t discriminate "Yes"/"Yes with comments" or "No"/"No with comments" [2], but I guess the outcome wouldn’t be much different.
My burning question is: how would the vote had been without all those late-P countries on board? You know, the ones that supposedly Microsoft and IBM bribed on equals merits? These are my sad findings [3]:
Do you need me to plot it??
So, an open quistion to Microsoft: who bribed them? did they just came on their own? (would you still argue *that* even if many of them have never shown interest in the standards-settings process?).
To anyone who knows something about the story in Ecuador (an opportunity to uncover evil evil IBM!), please share your knowledge.
[1] http://www.theopensourcerer.com/2007/09/03/ecma-376-dis-29500-ooxml-the-voting-so-far/
[2] so, "Yes" groups all "Yes" and "Yes with comments" votes (which is accurate), and "No" groups all "No" and "No with comments" (which is a little misleading as a "No with comments" means "Conditional Approval", meaning that the casting party made a voluntary non-binding commitment to approve in case its comments get addressed).
[3] The first group sums 100,1% because of rounding errors.
nachokb
Some interesting figures :
– amongst the 30 "original" JTC1 P-Members : 8 Yes, 14 No, 8 Abstain
– amongst the 11 "late-comers" JTC1 P-Members : 9 Yes, 1 No, 1 Abstain
– amongst the 15 "late-comers" SC34 P-Members : 12 Yes, 0 No, 3 Abstain
I think that these figures speak for themselves.
Another perspective :
– Original JTC1 P-Members: 30 ( 8 Yes, 14 No, 8 Abstain)
– Other ISO Member Bodies: 57 (43 Yes, 4 No, 10 Abstain)
I call "Original JTC1 P-Members" those registered as Participant Members at the end of the contradiction period. These NB normally knows what they are talking about. By definition, the 57 other voters were not participating to the process until very recently… Again, we clearly see from where the "overwhelming support" to OOXML is coming !
Note also that many of the 8 Yes and 8 Abstain votes in the original 30 P-Members group came after strange things happened as part of the voting process (e.g. Finland, Germany, Italy, Malaysia. Netherlands, Spain, Switzerland, USA).
Luc Bollen
"Some interesting figures :
– amongst the 30 "original" JTC1 P-Members : 8 Yes, 14 No, 8 Abstain
– amongst the 11 "late-comers" JTC1 P-Members : 9 Yes, 1 No, 1 Abstain
– amongst the 15 "late-comers" SC34 P-Members : 12 Yes, 0 No, 3 Abstain
I think that these figures speak for themselves."
This document shows another view of the results…:
http://docs.google.com/Doc?id=ddzgv9cv_1gkb7rm
Dario
There’s two questions I haven’t read about yet:
1. Can a "yes"-vote turn to a "no" vote again in the PRM-phase, or is this impossible?
2. Can new members still be added to the voting process from now on until February?
Can a yes vote turn into a no vote? The answer is yes
Can a new JTC1 member participate, or an O member upgrade to a P member, and in each case, have their vote count as such? I believe that the answer to this is yes, but that’s second hand. I haven’t confirmed it from the rules.
– Andy
Thanks Andy. This is intesting for Switzerland, where the yes result was reached by just one more vote (just one vote less, and it’d have been a no). For the second question, I should add this question:
2 b) On national levels, can new members (businesses, organizations etc.) join the NBs for the February voting? In other words, can not only new NBs join, but can new members be voters at NBs? If yes, then we might see all the lobbying again.
"Can new members still be added to the voting process from now on until February"
My understanding is that no new ISO member can now participate : the final result will be based on the existing 87 votes list. All these votes will remain valid and taken into account. The NBs that participate at the BRM will have the opportunity to change their vote during one month after the end of the BRM (from any of Yes, No, Abstain to any of Yes, No, Abstain). I’m not sure about the possibility to change the vote if a NB do not participate at the BRM.
However, it seems that the upgrade to P-Member can take place at any time, and that the status taken into account will be the status at the end of the ballot (planned in Mar-08). So, new P-Members can be "added" to the voting process : to balance the current 15 No voters, 13 Yes voters could be upgraded to P status (to reach 30 positive P voters). This is a "by the (letter of the) rules" approach that MS could try to take… but it would have a clearly negative impact on OOXML credibility.
Luc Bollen
The tricky part will probably be Step 1 : how to convince so many NBs to pay the travel and attend a one week technical discussion meeting, while some of them probably haven’t any expert that opened the OOXML file ?
That’s easy. I’m sure that some generous corporation or foundation can be found to sponsor those travel expenses.
The BRM head has a big role to play now. Does he have a say as to whether each comment has been properly addressed or does he simply put each comment & reply to vote and simply collate the results ?
If this is an ECMA standard then why is Microsoft announcing this result, isn’t that ECMA’s job? At least give the illusion of ECMA having some role in this, I would suggest that Microsoft write the PR and allow ECMA to release it. Microsoft has after all been very insistent that OOXML is an ECMA standard not a Microsoft spec.
I think its clear that Microsoft has so much riding on this that they have to spin really bad news into good news.
The interplay between Ecma and ISO plays out as follows:
1. Ecma adopts standard first; submits it to ISO/IEC JTC1 in its special role as a "Fast Track" partner
2. ISO runs, and owns, the process for global adoption. Hence, the right to issue the definitive press release properly "belongs" to ISO (the Microsoft press release, of course, has nothing to do with any official position)
3. Ecma, as the initiatior of the OOXML and is the home of the committee that packaged OOXML for submitssion, has the right to respond to the comments and recommend changes.
4. ISO SC 34 then has a whack at it, and finally the memberrs do as well.
5. If OOXML is finally adopted, a committee will be needed to maintain it on an ongoing basis. That committee can, but need not be, the original organization (in this case, Ecma) that created the specification in the first place. I believe that Ecma has submitted a proposal for filling that role if OOXML is finally approved.
– Andy
1. Ecma adopts standard first; submits it to ISO/IEC JTC1 in its special role as a "Fast Track" partner
Always bearing in mind that ECMA has placed itself subordinate to Microsoft WRT ECMA-376
2. ISO runs, and owns, the process for global adoption. Hence, the right to issue the definitive press release properly "belongs" to ISO (the Microsoft press release, of course, has nothing to do with any official position)
3. Ecma, as the initiatior of the OOXML and is the home of the committee that packaged OOXML for submitssion, has the right to respond to the comments and recommend changes.
Always bearing in mind that ECMA cannot make substantive changes without Microsoft’s approval.
4. ISO SC 34 then has a whack at it, and finally the memberrs do as well.
Which raises a question: from the wording of Microsoft’s "patent pledge," they can withdraw it if they don’t like the resulting specification.
You’re the lawyer here, Andy, and with the understanding that you’re not my lawyer, still …
Doesn’t this amount to Microsoft having a de facto veto power over any outcome of the BRM? What if a Microsoft employee stands up as a point is brought to the floor, and announces that Microsoft would not support a spec with that change and would, in fact, not extend its patent pledge to that spec? (This is not entirely hypothetical: I’ve seen it happen in working groups.) Have they left themselves enough freedom to do that?
5. If OOXML is finally adopted, a committee will be needed to maintain it on an ongoing basis. That committee can, but need not be, the original organization (in this case, Ecma) that created the specification in the first place. I believe that Ecma has submitted a proposal for filling that role if OOXML is finally approved.
With the usual observation on what that amounts to.
Without having gone back to refresh my recollection on the patent pledge, I’d say you’re pretty well on target for your supposition on point 4. I think that after Microsoft has talked so long about it being "no longer in control now that OOXML is an open standard" that they would have some limits on what they can do, especially with the EU looking over their shoulder. That said, if things got down and dirty, all bets could be off.
There’s a particularly interesting response that Brian Jones recently had at his blog to a question, in which he made one of the points that you raised – they have no obligation to make their patents available for free with respect to additions to the spec that didn’t come from them.
– Andy
The linked PCPro article no longer exists, having been replaced by Microsoft OOXML proposal falls short, also by Simon Aughton, which provides the correct facts and ends with an apology. I, for one, am glad to see a journalist willing to face up and do the right thing about a mistake.
I don’t know if anyone will klnow the answers to these questions but I think they are worth raising anyway?
1) How many other JTC1 votes are likely to take place between now and March that all of the new P members are now expected to participate in?
2) If the new P members fail to live up to their responsibilities is there a mechanism in place to downgrade them or encourage them to drop their P status and if so could that happen before the BRM?
"Finally, recall that I have consistently predicted that the vast majority of the O members that upgraded to P membership would vote in favor of approval, presumably inspired to do so at the urging of Microsoft. That expectation has been confirmed, although the motivation must continue to be inferred from the results, which are as follows: of the 11 upgrades, 9 voted to approve, or approve with comments, and only one voted against approval (Ecuador; Trinidad and Tobago abstained)."
5 of this 9 NBs that made a rushed upgrade to P, voted "approve without comments", i.e: "inconditional approval" : Cote-d’Ivore, Cyprus, Jamaica, Lebanon and Pakistan
bad smell…
Carlos