Open XML Will Have A Ballot Resolution Meeting


As I said in my last post, Open XML did not get approved. Frankly I was going to leave the topic alone for a while given that the process now moves to the ballot resolution process, but someone pointed me to a posting by Rob Weir of IBM that raises questions about the BRM. It is important that we clarify this point as it is really important.

To make things really complicated, some may be confused as to which rules govern the process being discussed. There are ISO/IEC rules for Technical Committees that fall under ISO or IEC control…and then there are ISO/IEC Joint Technical Committee 1 (JTC 1) rules for technical areas which are jointly under ISO and IEC control. To be very clear – Open XML (DIS 29500) falls under the jurisdiction of the JTC 1 Fast-Track rules, which differ from the ISO/IEC Fast-Track rules in significant ways.

Rather than quoting a whole bunch of legalese at you – I’ll be brief. During the balloting phase of the JTC 1 Fast-Track process, the JTC 1 Subcommittee that has responsibility for the draft international standard (DIS) establishes the time/location of the ballot resolution meeting. This is done before the close of voting and the rules clearly state that the final approval status of the DIS is not determined until after the close of the ballot resolution process.

Rob rightly points out that there are instances where a vote took place and no BRM was held. That is a decision taken by JTC 1 and the contributing body. Like, for example, ODF 1.0 – it received a unanimous vote in favor of the specification and thus the decision was made to not hold a BRM. It is important to recognize that this did not affect the final outcome. I have been told by more than a few folks that this annoyed several national bodies as it meant that their comments (submitted with their “Approve” votes) were not ever directly addressed. This fact contributed to the issues around “Yes with comments” votes, thus my blog posting earlier. Another example was C++/CLI – where the vote was negative and considering that the comments were not going to be addressed by the submitter, the decision was made to cancel the BRM – not affecting the final outcome

Open XML is different in that the BRM may well affect the final outcome, and thus is very important to hold. ODF 1.0 went through JTC 1 easily not because it was a superior specification (here are some comments I made on this point), but arguably because there was very little interest in the spec at the time. Open XML is clearly a different story. There is HUGE interest in this specification and the implications of the spec on the industry as a whole. OK, it is being held to a higher bar (I have no problem with that provided the rules remain consistent), but that also means it would be out of the question to cancel the BRM.

There will be a ballot resolution meeting at the end of February and this has already been announced on the JTC 1 SC34 web site. Between now and then ALL of the comments will be taken into consideration. I totally agree with the comments others have been making on my previous posting that transparency into this aspect of the Open XML journey is important.

Oh, one more thing from elsewhere in Rob’s blog. If I were the head of a National Standards body and being told that my choice to participate in this vote as a P-Member was to be scorned – I would be insulted. While I was not surprised by Rob’s insinuation about the Transparency Index due to the unfortunate situation in Sweden, I am saddened to see that turn into a means for him to insult the integrity of those national bodies. People seem fond of talking about respect in the process, but that only works if you show some yourself. I have a great deal of respect for the work Rob has put into something about which he clearly feels very strongly. I don’t have respect for the line of logic that says the mistake made in Sweden shows that the MS teams and National Body members are acting unethically elsewhere categorically. (Caution, snarky comment – but by Rob’s logic, IBM’s activities in Argentina in the mid 1990’s would suggest a pattern of behavior unacceptable to most. I don’t see it that way, individuals who do improper things do not necessarily represent the company as a whole.) I think we could all wish for this entire process to have been less messy.

At this point, I’d like to see things move towards more constructive dialogue. Open XML is being widely adopted, it has significant independent implementations, and now stands to benefit from the hard work put into the comments submitted through the JTC 1 process. 

Comments (25)

  1. Thanks for the post.  Could you clarify what is allowed to happen with the comments?  Can they be posted somewhere public, either before or after somebody culls out the duplicates, but before the BRM?  Are they considered confidential at all?  As you have said, this is a learning experience for all of us in terms of standards, and I would certainly like to be able to understand more.  While I am sure Rob Weir and others will address this as well, I’d like to understand your views.

  2. Alex Brown says:

    Jason hi,

    On a point of process detail, the JTC 1 Directives state that the final decision on the DIS is the result of the BRM’s deliberations, and that "if it is impossible to agree to a text  […] the proposal has failed and the procedure is terminated." If the BRM agrees to a set of revisions, the proposer (Ecma in this case) has one month to prepare the final text which is then automatically published as a full International Standard.

    – Alex.

  3. Simon Phipps says:

    Jason:  Good to hear you’re so committed to the BRM. As you imply, now is a great point to clear up issues left over so…

    At the London NCC/BSI meeting, Stephen McGibbon and Stijn Hendrikse promised that Microsoft would be asking JTC1 to manage the maintenance of DIS 29500, rather than ECMA handling it. Is this in fact Microsoft’s intent?

  4. Andy says:

    //Oh, one more thing from elsewhere in Rob’s blog. If I were the head of a National Standards body and being told that my choice to participate in this vote as a P-Member was to be scorned – I would be insulted. While I was not surprised by Rob’s insinuation about the Transparency Index due to the unfortunate situation in Sweden, I am saddened to see that turn into a means for him to insult the integrity of those national bodies.//

    It is nothing wrong about questioning the design problem of the UN style ISO process. It is madness that banana republics get as much rights as industrialized nations.

    It is also a matter of principle to be open. When I am in certain African states I know that governments are sort of corrupt over there. To a certain degree I have to participate, e.g. bribe the airport guys. In ideal terms the ISO process rules ensure that these nations cannot just block a standard with bad intentions. Nationality does not really matter.

    On slashdot I found the http://openiso.org proposal. I think Microsoft should participate in these efforts!

    // If I were the head of a National Standards body and being told that my choice to participate… //

    It is nothing wrong to call a whore a whore. It creates a bad impression among technologists to spin around. The issue is not moral, the issue at hand is how to get reasonable procedures for standardization. ECMA made it too far simple for you, so get the problems in fast-track. ECMA failed because it had no proper review process. ECMA is no real standard body.

    Now, I can get x in a compromised nation, but in my own nation I prefer a non-corrupt government with sound procedures,. Rob Weir’s comments were bullet-proof. You can’t say that any nation can vote approval without comments when the review process workd. That is not honourable. so even if it benefits you, we know there is a problem. Or how funny the Kenya thing…

    We should work to get a reform for all of us that makes abuse less likely. Bias away, think benevolent. An open ISO is what we need. ECMA didn’t do its job. It is like a teacher that let’s you graduate although you don’t deserve it. So you get the trouble in a later stages. You can’t always get away with it.

  5. jasonmatusow says:

    Ben – If I am not mistaken, your questions are good and reasonable, but not answerable at this time. I think that the Project Editor and crew are working through this now. We’ll all find out soon (I hope) as to what the next steps really look like.

    Alex – good to hear from you. Fair enough on the process point. The first question as to if there will be a BRM was the point of the blog posting. What/how in the work leading up to it, and how the BRM itself is going to run is still being discussed as far as I understand it. If you have insight into this, please share.

    Simon – I have said this in a post already. There is a proposal for a joint maitenance agreement already submitted to SC 34. My understanding is that will be a point of order during the SC 34 plenary this fall. Hope that helps.

    Jason

  6. jasonmatusow says:

    Simon – one more quick thought. (just a friendly jibe really) Since maintenance is really an issue for an approved standard, can I take this to mean that you are eager to see the DIS move to IS after the BRM? That way the maintenance agreement can be put into use.  😉

    Jason

  7. Simon Phipps says:

    Only if all the comments are adequately addressed and there’s no more gerrymandering, Jason… 🙂

  8. RichL says:

    The key line from Rob’s blog for me was //We need to remember that a participation rate of 50% of JTC1 P-members is required to transact most JTC1 business. So this "rejuvenation" may very well paralyze JTC1 entirely unless the new members are earnest and participate in ballots beyond OOXML.//  I don’t see that level of participation by many of the late comers as credible.  The last minute signups by so many NB’s for this particular ballot, whether heavily influenced by Microsoft lobbying or not, could have disastrous unintended consequences for other standardization efforts.  Again, it seems like the single-minded pursuit for a specific standard does long term damage to the ISO and JTC organizations.

  9. jasonmatusow says:

    Simon, there is deep irony here that the term "gerrymandering" comes from a Massachusetts politician. (His last name was Gerry and he created an electoral district that looked like a salamander on the map.)

    Rich – thx for the comment, point taken. I’m not sure of the implications myself at this point, but it is worth some deep consideration.

    Jason

  10. Anonymous Coward says:

    Hi Jason,

    you wrote:"Open XML is being widely adopted, it has significant independent implementations, and now stands to benefit from the hard work put into the comments submitted through the JTC 1 process."

    Would you mind putting up few URLs to those "significant independent implementations" of Open XML for us who would want to check them out?

    Thanks!

  11. jasonmatusow says:

    Anon – you bet. That would be a good top-level post. I will put in on my to do list.

    Jason

  12. carlos says:

    what is Open XML? do you mean Office Open XML ( OOXML ) ?

    ( please don’t rename the specification… there is too much confussion in all this rushed standardization to add one more )

    If you choosed the name Office Open XML ( confusing with Open Office product ) , stick to it and face the bad decission.

  13. Len Bullard says:

    There has been heavy lobbying and last minute sign ups on all sides.   The ISO process must continue unimpeded or there is no use in a process at all.    The general atmosphere of acrimony is far more poisonous to the pursuit of nation-consensus standards than the acts  of any single entity.

    In English:  it is time to put the attack dogs back in the cages.  As said on XML-Dev, attack dogs tend to circle back on the hands that held the leash.  Caveat vendor.   MS, IBM, Sun, Red Hat and the others, pay attention.   Your reps as contributing members are on the line here and the web is not a forgetful technology.  These incidents resonate for years to come.

    Two things are very important:

    1.  ISO process must continue in accordance with ISO policy without extra burdens created to make difficult work impossible.

    2.  Two standards will not do any damage to the status quo but creating an atmosphere where NO standard will be accorded legitimacy given the events will.

    Be professional and clear even if it hurts.   All organizations involved must discipline their contributors to act accordingly.  The stakes are very high and the negative results will break in all directions equally.

    Breathe and let the people with the difficult job get on with it.

  14. Lars Gaarden says:

    Jason, I see you have neglected to respond to Andy’s comment.

    Of the 12 members that joined in 2007, 10 voted yes. Add the influx of new participants in the NBs whose only participation in the standard process was to vote yes without comments.

    How is a neutral observer supposed to come to any other conclusion than this being a serious attempt at ballot stuffing going on? You can spin it as an attempt to "insult the integrity of those national bodies", but not calling what is obviously a spade a spade seems intellectually dishonest.

  15. Lars Gaarden says:

    "OOXML [..] now stands to benefit from the hard work put into the comments submitted through the JTC 1 process.".

    Unfortunately, many of the comments expose a lack of polish. For an example, forgetting to specify whether trigonometry functions use degrees or radians should have been identified and fixed a long time ago. These kinds of errors – at least in the volume identified by the NBs – raise serious questions as to how much review has actually gone into the proposed standard. Submitting it to Fast Track seems premature.

    Documenting the MS Office file formats is a huge undertaking, and submitting it as a standard doubly so. MS should be commended for doing this work, since a proper OOXML standard would be a huge benefit. But please MS, do a proper job of it instead of ballot stuffing it through the Fast Track.

  16. Anonymous says:

    Lars, go ahead and call spade a spade. Which countries were "bribed"? Which countries "stuffed the ballot"? USA? Germany? Brazil? China? India?

  17. Anthony says:

    This is an interesting article, though I have to say I came here from groklaw.net, so you’ll probably be able to guess what my interpretations of events are, yet it’s interesting to read yours.

    One thing that has me puzzled, however, is your interpretation that ODF passed JTC1 because there was little interest in the spec, while there’s significant interest in OOXML, causing it to be held to a higher bar.  Yet, the difference in attention appears to be the addition of a lot of countries that had no problem rubber stamping OOXML.  The same countries that had no issues with ODF are the ones now raising issues for OOXML.  I hardly see that as a lack of focus on the standard when ODF passed the process.

    Anyhow, I still fail to see why Microsoft continues to push for a standard that is, on the one hand, completely different from it’s past formats, and yet insists on doing stuff like incorporating a date bug, to retain compatibility.  It’s pretty hard to see why using a non-broken date format for the new proposed standard would prevent microsoft from being able to save it’s formats to a backward compatible format. Maybe you could explain that in another posting.

    Just to finish up on that point, your comment about OOXML being held to a higher bar also really leaves me irked.  If the bar was consistent, I’d expect issues like the buggy date format to have been wrapped up long before a fast track proposal was ever pushed forward, at which point the BRM would clearly not have been needed – clearly that leaves observers wondering why Microsoft is in such a rush, if they’re serious about the standards process.  If they are (as you claim they are), should we expect to see a vastly improved (and ISO compliant) standard at the BRM?

    In any case, thanks for the blog – it always gives me something to think about.

  18. SM says:

    (quote)

    Lars, go ahead and call spade a spade. Which countries were "bribed"? Which countries "stuffed the ballot"? USA? Germany? Brazil? China? India?

    (unquote)

    US, Sweden, Germany, Portugal, Switzerland, Malaysia, Singapore, Denmark, Hungary, and most of the new P members all had irregulatities pointing to corruption and ballot rigghing on behalf of Microsoft corporation.

  19. Tim Wesson says:

    I’m sorry, but with regard to the transparency index, a correlation would most likely imply corruption, but this does not imply that every case with low transparency would be corrupt:  that is bad statistics.

    Further, not to call a spade a spade through "respect" is to invite corruption.  There is a reason why there is a strong correlation between transparency and corruption generally, and that is that "respecting" a system invites corruption.  Those systems that are the most self-critical might get PR problems in the short term, but they grow to be the systems with the most integrity in the longer term.

    Political correctness here, "respecting" the participants will doom the ISO;  no institution should be "respected" in this way.

  20. jasonmatusow says:

    Lots of comments here, I certainly appreciate the activity – sorry for my being slow in responding.

    Len – I really appreciate your comment. You are spot-on about the implications of the overt commercial competition bubbling up into the international standards arena. The pressure from both sides of this issue have been tremendous and have certainly tested the limits of the whole process. ISO, and the other international standards orgs, perform a critical function for global commerce and it is important to keep that in mind. Very astute comment  – thx again.

    Lars – you are right, I did not yet respond to Andy – I will. As for your other comments – I have addressed this numerous times in earlier postings. The difference with Open XML has been the elevated interest in the standard (both for and against). I have laid out arguments about ODF and could pick other examples as well where similar paths, maturity curves, etc. for specs have come through ISO, been approved and moved on. In this case, it has turned into a much broader set of issues beyond the spec – and that is what makes it different.

    SM, Tim, and others – I have been part of the team at MS involved with the submissions and subsequent work with Ecma and now the national bodies. We have been very clear with our teams about these engagements and hold all employees to a clear code of conduct. While easy to take shots at us on this – it is not reflective of what has been going on. Sweden happened – it was unfortunate – it was addressed in a professional manner – it does not define all other activites. I pointed this out in a posting before, the champions of the other side (IBM) had a massive bribe scandle that resulted in employees going to jail – does that mean that their behavior in all other endeavors is corrupt?

    As for the smaller countries moving to P- membership. And for the comments about what countries should, or should not, be doing. The whole point of the "big I" standards bodies is to engage in standards work within the context of national soveriegnty. It is their choice as to how they want to participate. It is allowed (in most countries) for private interests to express their opininion on a given matter – but it is ultimately the government’s decision as to what action to take. The IBM GM of South Africa sent letters to all national bodies in Africa advocating a no vote. Considering the size of IBM’s contracts in those countries, their influence globally – is that improper behavior? In my book it is not.

    Tim – my use of the work "respect" was not about turning a blind eye as I think you are suggesting. I do mean though that they are soveriegn entities and there is something to having a regard for the desire to particiapte in international commerce. Hope that clarification helps.

    Jason

  21. jasonmatusow says:

    Andy –

    I disagree with your comments on a number of levels. On the personal level – I have been to more than 40 countries over the past number of years and I have yet to bribe a single airport employee.

    I do agree that ISO rules provide a context that creates a level playing field for all nations, but then you have to be ready for all players to play on the level field. ISO participation is not based correlated to population or GNP = # of votes. It is one country, one vote. JTC-1 has rules, and participation functions within those rules. Then, they have purposely allowed for each country to determine their own structure/rules for national bodies. This means that you need to value diversity even within the highly structured interactions of JTC-1.

    Jason

  22. Andrew Sayers says:

    Anthony,

    Brian Jones talks about technical issues in his blog (http://blogs.msdn.com/brian_jones/) more than Jason does, so you might find a stroll through his archives informative.  I don’t particularly like Microsoft’s solution to this problem, but it’s a genuine problem, and dismissing it as a legacy issue isn’t a solution at all.

    The example I like to use to show the problems Office Open XML is trying to solve is to ask what strategy you’d use to import a formula like this from Excel into ODF:

    =IMREAL(IMEXP(COMPLEX(0, "03/01/1900 03:23:53.606"))) + WEEKDAY("4 March 1900")=0

    This exploits several of Excel’s bugs and eccentricities, and it’s worth spending a while with it to get a feel for what "100% fidelity" really means.

    – Andrew

  23. jasonmatusow says:

    Anon – you asked me to put up some links to independent implementations of Open XML. I promised a top-level post, and just completed it – hope it is both satisfying and delicious. 🙂

    Thx

    Jason

  24. Romeo Pruno says:

    In questi giorni la rete freme di opinioni, molto interessanti, circa le attauali questioni suoi nuovi

  25. In questi giorni la rete freme di opinioni, molto interessanti, circa le attauali questioni suoi nuovi