IS29500 - Open XML Is An International Standard

ISO has distributed the results of the Open XML vote to national bodies (**UPDATED) resulting in leaked information - but now confirmed by ISO in a press release. DIS 29500 has become IS 29500. In regular person speak – Open XML was adopted as an international standard by ISO/IEC JTC 1.

This standard has received more technical and political attention than any specification in the history of the ICT industry. Genuine industry competition around office productivity applications fed a heated exchange of technical considerations and philosophical positions regarding standardization, intellectual property, and most certainly document format technology. In the end, all of this – from the most ardent critics to the dedicated supporters of the specification – contributed to the improvement of the Open XML specification which ultimately led to its adoption as an ISO standard.

DIS 29500 failed to pass the Sept. 2 ballot by missing the super majority requirements. The fact that the ballot resolution process – conducted intensively over a seven-month period - produced such strong technical work and quality engineering engagement resulted in many of the “no” votes changing to “yes” as contemplated by the JTC 1 Directives. This happened because these national bodies considered the substantive changes to the specification and felt that their concerns had been sufficiently addressed -- leading to the position that the specification should be adopted as an International Standard.

The spec is genuinely better today than when it started. The Ecma 376 version is already broadly implemented across multiple platforms. Now the work continues in SC34 on maintenance and with product teams throughout the industry as they make use of the new standard.


Given the focus on this topic throughout the industry, I’m going to take the time to put down some extended thoughts about the standard and where we are today. (Caution – long post) I think it is worth discussion some of the macro factors that surround Open XML.

1. It is imperative that you own your data – and control its future. Within this context, XML-based document formats are good for society at large. The promise of the W3C specification that gave us XML in the 1990’s is now being realized in a myriad of ways -- through new protocols, new document formats, and new capabilities of applications. The innovation made possible by this technology – XML – is changing the very face of the industry (of course coupled with the advancement of processing capacity, and core infrastructure such as the continued spread of network ubiquity etc.) and the extent to which users are gaining greater control of their data.

2. Those who create applications need to invest in building the bridges that enable the effective exchange of data. This is a significant statement, and one that I could write a many thousand word essay on to explain. Suffice it to say that this is a critical aspect of what Microsoft has been and will be doing on interoperability. Every vendor needs to think across the full spectrum of the products they build, the work they do with the broader community, the access they give to their IPR, and the work they do with standards to achieve this goal.

3. Hundreds of organizations were involved in the JTC 1 process. Like most software standards, Open XML was a contributed specification. While the original smaller specification came from a single organization (Microsoft in this case), the specification was improved through the input of more than 20 organizations in Ecma TC 45 originally (including key users and competitors), and then from hundreds of parties through the JTC 1 process. I am greatly encouraged by the participation by all parties (for and against) as the process resulted in a better specification. As I stated before, significant industry competition factors around applications created a highly contentious environment, but out of that came an improved specification that is meticulously documented and gives independent implementers what they need to be successful.

4. Governments requested that Open XML be standardized. More than 2 years ago governments requested that Microsoft standardize the default format of Microsoft Office. Since the late 1990s Microsoft had been working to move away from binary formats to an XML format. That engineering effort culminated with the move to Open XML as the default format for Office 2007 and the contribution of that specification to Ecma. That has all resulted by responding to this request from governments.

5. Open XML is creating new opportunities throughout the industry. Open XML has opened the door for partners of Microsoft, customers, and competitors alike to create innovative solutions and to tap into the marketplace opportunity of Microsoft Office customers. We are already seeing applications in the healthcare, mortgage lending, manufacturing, eGovernment, insurance, document management, mobile devices, and many more segments that are taking advantage of Open XML. While some of these implementations are in association with Microsoft Office, many more are completely independent of it. I hope to see an explosion of innovation surrounding this format as it will benefit all parties.

6. The future of the specification is in ISO/IEC’s hands. A contributed specification is just that – it is “contributed.” Microsoft is like everyone else in regards to the future direction of IS29500. Long-term maintenance of the specification will be under the direction of JTC 1 (via SC34) and in partnership (of some form – to be determined if I’m not mistaken) with Ecma TC 45. Microsoft has committed to implementing IS 29500 in its next version of MS Office, and we will document that implementation in accordance with the interoperability principles we announced earlier this year.

I could go on and on, but I want to shift focus to the question of lessons learned. Anyone who has worked on this standard will likely reflect on their experiences. The long-term effects of this effort will be felt throughout the international standardization community as well.

1. International standards are about the long-term, not the short-term. The ratification of Open XML is just the first step in a longer discussion about document formats at the international standards level. ODF 1.2 and 1.3 (currently only under review at OASIS) will come before JTC 1, as will other activities: Open XML maintenance; PDF 1.7 maintenance; China may decide to move UOF to JTC 1; the work that is happening in DIN (the German national standards body) on translation and harmonization. It is a testament to the importance of document formats that so much engineering capacity and business investment is focused on these issues.

2. Politics matter. It became apparent to me early on in the cycle of DIS 29500 that the technical merits of the specification were only partially influential in the discussions of Open XML. The detractors and supporters alike (myself included) took the debate to the court of public opinion and into the political arena. The long discussions held over ODF procurement mandates being placed into statutes, the attempt to exclude Open XML from state-level procurement through policy mandate, and the debates over economic opportunity and industry competition were significant discussions. Each of these exposed different aspects of the debate. And each, to some extent, touched on foundational issues regarding the role of ICT in society. “Lobbying” was thrown about in an accusatory manner by both sides, but the pejorative hides what are some truly critical issues. Frankly, I welcome the fact that OSS advocates, private industry, NGOs, academics, and even other government agencies reached out to each other to discuss these issues. The invective that often accompanied some of these activities was unfortunate – but the end result was an overall, relatively deep discussion.

3. Speed of standards development remains an open-ended debate. One of the most basic criticisms of Open XML was that it should not have been put through the “Fast Track” process. The fact is, Fast Track is a more rigorous process than PAS (by the fact that it has the contradiction phase tacked on to the front of the process) and both move straight to the JTC 1 plenary vote rather than the SC-level work. As governments continue to drive toward eGOV frameworks where they have a favoring of standards that carry the ISO imprimatur (I’m not necessarily saying this is good or bad), then the old issue of speed of development will come to the fore. The software industry has relied heavily on industry consortia over the years due to the desire to get specs into circulation faster than through the slower, and more formalized, international standardization process. Open XML has shown a very bright spotlight on this issue which should merit the healthy, and rational, discussion of how this issue is to be dealt with over time. PDF 1.7 just went through the Fast Track process (with nary a peep from the vast majority of people about it), and ODF 1.0 went through the PAS process. (These are just doc format examples – anyone with a DVD player is using ISO Fast Track standards without famine or draught being caused.) My personal opinion is that nimble performance is going to be a real test for ISO in the future.

Quite frankly there is much more, but this is enough to think about for one day. I’m really proud of the work done by so many organizations in reaching this goal. The FUD-throwers will paint a picture of Microsoft taking unilateral action and thus this result. Anyone saying that is either purposely obfuscating the truth or spectacularly ignorant. The deep technical work done by engineers from around the world representing a truly staggering number of interests is what brought us to this point. Congratulations to everyone involved in helping Open XML become an ISO/IEC International Standard.