The Future of ODF and OOXML

Tomorrow is the last day that a National Body (NB) can change its vote on OOXML.  Only a few NBs have announced what they have decided, and of those, not enough have changed their votes to reverse the outcome of last summer, in which DIS 29500 (a/k/a OOXML) failed to gain approval.  It will not be until Monday that the final vote will be announced by ISO/IEC JTC1 (or become public through disclosure by an NB committee member, as the case may be).

Many journalists and others have asked me whether I have a prediction on what the outcome will be, and also what I think it will mean if OOXML is approved.  I don’t have an answer to the first question, as there are too many countries involved, and too much may change until the last minute.  But I do have an answer to the second question, and that answer is the same one that I have given every time that a new decision point has loomed in the ongoing quest for a useful format standard that can bring competition and innovation back to the desktop, as well as ensure that the history and creativity of today will remain accessible far into the future.

That answer is this:  if anyone had asked me to predict in August of 2005 (the date of the initial Massachusetts decision that set the ODF ball rolling) how far ODF might go and what impact it might have, I would never have guessed that it would have gone so far, and had such impact, in so short a period of time.  I think it’s safe to say that whatever happens with the OOXML vote is likely to have little true impact at all on the future success of ODF compliant products. 

Here are ten reasons why I believe this prediction will be borne out.

1.  The quality of OOXML is not there yet.  Each of the NBs that voted on OOXML would agree, regardless of their vote, with the following statement:  DIS29500 is still a standard with many issues that need to be resolved before it is as good as an ISO/IEC JTC1 standard should be.  The great majority of the hundreds of substantive proposed revisions were adopted in the form proposed, while virtually all of those that were discussed and approved required modification before they were adopted.  The choice with respect to these changes was “leave broken,” or “accept this fix, whether you think it’s good enough or not.”  Many identified issues have been deferred to the “maintenance phase,” and therefore won’t be initially resolved at all.  In short, it will be some years before OOXML is of the same quality that an ISO/IEC JTC1 standard is expected to represent by the time it is adopted.

This is hardly the outcome you would expect from a committee co-chartered by ISO, the global organization that gave the world the widely adopted 9000 series of standards for quality management systems.  ISO 9001 requires that a compliant organization must institute the following requirements, among others:

·    a set of procedures which cover all key processes in the business;
·    monitoring processes to ensure they are effective;
·    keeping adequate records;
·    checking output for defects, with appropriate corrective action where necessary;
·    regularly reviewing individual processes and the quality system itself for effectiveness; and
·    facilitating continual improvement

Would the Ballot Resolution Meeting in Geneva meet this test, or indeed, the entire Fast Track process, as applied to OOXML?  The bottom line is that “much improved” does not equate to “good enough,” and never has for ISO/IEC JTC1 in the past.  If that changes tomorrow, JTC1 will lose more credibility than OOXML will gain.

2.  There will be no OOXML-compliant products for some time.  If the value of approval by the formal standards process is that some governments will only procure products that comply with ISO/IEC adopted standards, then for some yet to be determined time period, only ODF-compliant products will meet that test.  Office 2007 does not now conform to DIS 295000, because it cannot – the formalized version of OOXML has not yet been publicly released, if indeed it yet exists.  Only when it has been published, obtained, and studied by a vendor can the actual coding, debugging and distribution of a compliant product update.  Not surprisingly, Microsoft has not yet announced the date by which a fully DIS 29500 compliant version of Office will ship. 

3.  The “billions and billions of documents” that already exist were not created in OOXML.  The primary reason given for OOXML’s existence at all has always been to maintain the accessibility of the "billions and billions" of documents that had already been created using prior versions of Office.  However, looking backwards, those prior versions of Office were not based upon OOXML, which had not yet been created.  And looking forwards, the hundreds of millions of pre-existing copies of Office still in use cannot open documents created by OOXML-compliant software (or Office 2007) without a plug in.  In contrast, products such as OpenOffice natively support opening those billions and billions of documents already in existence with more than adequate fidelity to permit (for example) citizens to round trip the type of documents they need to exchange with their governments.

4.  Many governments do not wish to support a monopoly.  Microsoft remains unpopular in many countries around the world.  In Europe, the European Commission continues to investigate its behavior – including the manner in which OOXML was promoted through the Fast Track process.  In many emerging countries, governments would prefer to help local industries create jobs in the information technology industry, rather than continue to support a US-based multi-national.  Indeed, the popularity of the United States itself is at a low ebb after seven years under the current administration.  And finally, a government can switch to another ODF compliant suite in the future, if another ODF product develops more quickly.

5.  Many Governments promote competition.  Independent of any bias against Microsoft and/or the United States, some governments would like to promote innovation and competition, rather than help support the continuing existence of an ecosystem centered on the products of a single dominant vendor.  By making their procurement budgets publicly available for ODF-compliant products, these governments can provide market incentives to bring competition back to the desktop, where it has been largely extinguished for almost two decades.

6.  Governments may not wish to tell their citizens what software they must buy and use.  While various plug-ins and converters exist, their performance thus far has been sub-par.  While other types of products besides office suites will implement OOXML, the core implementation that most citizens will use when they interact with government will remain a product that creates text-based documents.  Already a variety of such products exist that are based upon ODF, and several of these suites are available for free. 

7.  “Open” means more now than it used to.  For an increasing number of governments, open means not only that a standard can be implemented by any vendor, but also that it can, and has been, implemented in open source software.  While Microsoft continues to assert that it has (unnamed) patents that would be infringed by office suite software, such as OpenOffice, and also that it reserves the right to charge royalties for access to those patents, there will be a cloud over the ability to implement OOXML in an open source software package licensed under the General Public License (GPL).

8.  The contest between ODF and OOXML has raised public awareness over the importance of document standards.  There are now substantial numbers of citizens in many countries that have become active in support of open document formats, and who are not happy with the progress or outcome of the OOXML Fast Track process.  The same events have made the substantial and growing open source community aware that the development and adoption of truly open standards can be as important to securing the freedoms they care about as can open source licensing agreements.  These citizens will remain active, and will pay close attention to the procurement policies that their governments adopt. 

9.  Recognition of the existence of “Civil ICT Rights” secured by “Civil IT Standards.”  The incredible success of the Internet and the Web has transformed virtually every aspect of modern existence.  I recently articulated a case for the existence of a subset of standards that I termed “Civil Information and Communications Technology (ICT) Standards” to protect what I called Civil ICT Rights. I explained this concept in part as follows:

We are entering an era in which IT technology is to society as earlier very different modalities were to human rights, from freedom of expression and free access to information (the unfettered use of the printing press), to civil rights (the abolition of separate schools and separate seats on buses for people of color in the US), to freedom of religion (the ability to openly practice one’s religion in houses of worship).

In this new interconnected world, virtually every civic, commercial, and expressive human activity will be fully or partially exercisable only via the Internet, the Web and the applications that are resident on, or interface with, them.  And in the third world, the ability to accelerate one’s progress to true equality of opportunity will be mightily dependent on whether one has the financial and other means to lay hold of this great equalizer….

[A]s the world becomes more interconnected, more virtual, and more dependent on ICT, public policy relating to  ICT will become as important, if not more, than existing policies that relate to freedom of travel (often now being replaced by virtual experiences), freedom of speech (increasingly expressed on line), freedom of access (affordable broadband or otherwise), and freedom to create (open versus closed systems, the ability to create mashups under Creative Commons licenses, and so on).

I believe that governments will increasingly recognize the existence and importance of Civil ICT rights, and will use their procurement power to promote and protect them.

10.  Cost.  The total cost of ownership of ODF-compliant products can be lower.  Open source products such as OpenOffice are free to download, and StarOffice, with full support, is significantly cheaper than Office.  Moreover, converting from Office 2003 to OpenOffice may require less employee training than upgrading from Office 2003 to Office 2007, due to the number of changes made in the new version.  This is especially relevant in the emerging nations that will represent the highest growth markets in the years to come.  Many purchasers will be first-time buyers, and of those many will be more interested in cost than edge functionality.

With all that said, will I be happy if OOXML is not approved?  The answer to that is certainly yes, for a variety of reasons, including the fact that it will act as a first step towards reestablishing the integrity of the formal standard setting process.  But do I think that an approval of OOXML will reverse the forward momentum of ODF?  I firmly believe that the answer to that question is not at all. 

History has demonstrated time and time again that once people realize that important rights are at stake, securing those rights cannot be denied, by anyone.  The importance of open document formats has now achieved mass recognition, and compromise in the protection of the Civil ICT Rights that rely upon Civil ICT Standards is no longer an option.


For further blog entries on ODF and OOXML, click here

sign up for a free subscription to Standards Today today!

Comments (7)

  1. The impact of approving OOXML in its current state will be primarily felt by ISO – its reputation as a technical standards authority has been badly tarnished by the progress of OOXML through a "fast track" process that has been set up to process existing standards (not standards that have never been subject to a critical review). To pass a standard in the state OOXML is reported to be in is ridiculous and, given the debate around it,  also fails to address the need for "consensus" is ISO standards.

    The victims of the resulting loss of reputation will be the other standards ISO produces in this domain and the people and institutions that have invested time and effort to produce standards of an acceptable quality (my own direct interest is in C++, but I’m sure that other working groups have similar concerns).

  2. This blog entry is totally biased in my opinion. Most of the bullet points look irrelevant to me since they keep recycling the concept that Microsoft is having a monopoly which is a bad thing. This very statement is debatable but overall I may agree with it, however I cannot figure out how the approval of OOXML as a standard will reinforce their monopoly. On the contrary, having a standard format – whatever it is – makes it much less dependable on a single software provider, don’t you think?

    Also, re: your point about the billions of billions of existing documents: as far as I know, most if not all these existing documents are using Office formats (doc, xls, etc). And as you admitted yourself, they can be opened and exchanged by many different applications including open source suites "a la Office". So, what is the point with them in relation to OOXML? They are some sort of "de facto" standard already. And the fact that Open Office or Star Office are able to read them does not constitute an obstacle to the adoption of another standard, nor does it represent an overwhelming criteria in favor of the adoption of ODF just because these applications support this format as well.

    Finally, I find the point about the adoption and training costs of Open Office and Star Office compared to Office 2007 very specious. I won’t argue on financial figures, but again I don’t really see how it makes the debate on standard formats move forward. If you want to demonstrate that a specific application is better or cheaper than another, then you totally miss the whole point of why we need standard formats, that is to facilitate the exchange of information across as many applications as customers are willing to buy and support for whatever reason.


  3. 1.  The quality of OOXML is not there yet.
    Yup. Well-played. By the time OOXML will be cleaned up the real harm will be far, far away. MS Office 2007 will be history and competitors will be forced to support MS Office 14 or MS Office 15.
    2.  There will be no OOXML-compliant products for some time.
    What are you talking about? The whole hoopla was to make it irrelevant if there will be OOXML-compliant products or not. The fact that there are no OOXML-compliant products will put MS Office in the position of "golden standard": "well, you know – no products are 100% compliant yet, but MS Office 2007 is the most common one so please make sure you are supporting it".
    3.  The “billions and billions of documents” that already exist were not created in OOXML.
    Wonderfull, doesn’t it? Now competitors will be forced to support yet-another-standard in addition to .doc/.xls/.ppt.
    4.  Many governments do not wish to support a monopoly.
    But many clerks in these some governments do. By one reason or other. Now the have carte-blanche.
    5.  Many Governments promote competition.
    Yup – but now it’ll be harder to do.
    6.  Governments may not wish to tell their citizens what software they must buy and use.
    Yup. But then some key officials will get "free" copy of MS Office 2007 and will demand to use "standard" OOXML format since it’s what this thing uses by default.
    7.  “Open” means more now than it used to.
    Yup. That’s the only good thing. But of course Microsoft will claim that since OOXML is an open standard and free software supports it so bad it just means you can not trust FOSS at all.
    8.  The contest between ODF and OOXML has raised public awareness over the importance of document standards.
    That it did, but the end result is: Microsoft can continue to use what it likes and can crush alternatives by weight of it’s monopoly.
    9.  Recognition of the existence of “Civil ICT Rights” secured by “Civil IT Standards.”
    We’ll see. It’s indeed a problem for Microsoft. Will need a lot more "standards" like OOXML. Remember: XPS and others?
    10.  Cost.
    You’ll have harder time with compatibility-wise: most of your respondents are using MS Office, not! Is it enough to justify the cost? We’ll see.

    Remember POSIX support in NT 3.1 (back in 1994)? It was strict requirements for the OS to by buyable by government. Microsoft implemented some kind of POSIX subsystem – enough to pass the tests, not enough to really use. Here we have such a situation right now: OOXML is incompatible with MS Office, but it’s close enough to claim that MS Office 2007 is using international standard for it’s documents. Enough to sell, not enough to harm, because true interoperability will be a myth. Well-played, Microsoft.

    Of course Microsoft will be forced to do some motions which are looking like a fullfilment of promises: provide some mapping (probably disfunctional), participate in OOXML “maintenance phase” (the slower the better), etc. But Microsoft is great dissembler – I’m pretty sure it’ll be able to it.

  4. Someone said (Weir?) that the worst nightmare for Microsoft would be if OOXML was approved now, and someone wrote a comprehensive, detailed, conformance test that gained widespread acceptance. Of course, given all the defects in OOXML and the imperfections in the future MSOffice support, Microsoft would have a hard time convincing governments that Office even conformed to the standard.


    • Someone said (Weir?) that the worst nightmare for Microsoft would be if OOXML was approved now, and someone wrote a comprehensive, detailed, conformance test that gained widespread acceptance. Of course, given all the defects in OOXML and the imperfections in the future MSOffice support, Microsoft would have a hard time convincing governments that Office even conformed to the standard.

      (Raises hand)

      Yup — however, let’s not get too excited.  The conformance clauses to DIS-29500 are such that it’s very difficult to not comply on paper.  You only have to be able to read a conformant file without crashing and write it back out again, so even cp is technically compliant.  That is, after all, the objective: to have a euphemism for "Microsoft Office" without flagrantly violating the government purchasing requirements.

      • <i>You only have to be able to read a conformant file without crashing</i>

        What are the odds a Microsoft product will be able to do that reliably?


    • Err, but there are published schemas which state which elements and attributes are allowed, in which positions and usages, and what the values of properties are supposed to be.  (And the use of BNF for formulaes, etc. means that it is not difficult to validate those things as well.)

      So for document conformance, the XML documents can be validated *now* with W3C XSD, and there will be ISO RELAX NG schemas as well.  Test suites only need to be made for the things that the schemas do not express.

      If there are more tricky constraints, the newer generation of more powerful rule-based schema languages such as ISO Schematron can be used. SC34 has been developing these languages in part to reduce the number of constraints that must be expressed in plain text in a standard without an automated way to verify them.

      For application conformance, that is a much more difficult issue, and it is something that all these standards have a problem with. (Which is why SC34 prefers to concentrate on document conformance rather than application conformance.)

      Traditionally, test suite creation and testing is performed by different bodies than ISO/IEC. E.g. NIST in the US.  They have skills and budgets in this area.

      Rick Jelliffe

Comments are closed.