ODF / Open XML – Technical Specifications Mature Over Time

Clearly things are heating up to a fever pitch this week. September 2 is just around the corner and the media knows a good story line when they see it. Great drama for coverage as country votes become public. The U.S. looks like it is headed for a Yes with comments and Germany has cast its Yes with comments vote. Brazil, No with comments. Countries all over the world will be registering their votes this week and some will publicly post their votes while others we'll have to wait until after the 2nd to know how they voted.

One argument that keeps getting raised by individuals representing companies with deep commercial interests in ODF is that Open XML is not technically ready to be approved. This seems inconsistent to me as those same folks are clearly ok with the fact that ODF was clearly not mature at the time of its JTC1 adoption.

Before I go any further, this is not about criticizing ODF. It is meant to be illustrative of the weakness of the arguments against Open XML.

Some facts to consider:

8 of the 32 P- member countries voted Yes with comments for ODF. Some of the issues raised were items such as:

  • The British Standards Institute pointed out non-conformance issues to mandatory ISO PAS procedures and requested the be addressed, normative references to specs that were different from similar existing ISO standards, incomplete specification of compression formats, and constraints due to implementation-specific references (meaning features in Star/OpenOffice represented in the spec). There was no BRM to address these comments.
  • Egypt noted that at the time of submission for international standardization, there was no support in the specification for Arabic languages.
  • Japan noted that it would be nice if the spec would address the technical capabilities to support accessibility.
  • China noted in their first comment that ODF should be integrated with China national standard final draft (UOF), and that it lacked support for user defined schema.

ODF is not a static specification - ISO/IEC 26300 (ODF) is no longer the most current version.

  • The OASIS Open Document Technical Committee continues to expand the capabilities of ODF. The ODF 1.1 has been released and 1.2 is in draft form already. These represent significant ongoing work based on elements either flawed or missing in the 1.0 specification.
  • The PAS management guide (JTC1N5746) states a desire to avoid divergence of the JTC1 spec from any future versions. Therefore the management guide requests that the originator of the specification work closely with JTC1 on revising or amending a PAS submission. Given that ODF 1.1 was not submitted to JTC1, should we interpret this to mean that it is not mature enough and that they are waiting for 1.2? What should be inferred about the maturity of the 1.0 specification?
  • The technical committee is not only progressing on the core spec, there are three subgroups working on accessibility, formulas, and metadata. This would seem to suggest that a significant amount of work was required to bring the spec forward from its IS 26300 form.

My point IS NOT to say that ODF should not have been standardized. In fact, Microsoft did not work to block this contributed specification from our competitors - we voted in favor of its adoption by ANSI of ISO/IEC ODF as a recognized National Standard. It is remarkable that so much effort is going into the blocking of an ISO/IEC specification. In other words, when ODF was submitted the same sense of civic duty was not as strongly felt by those opposing Open XML today.

My point IS that people with glass houses should not be throwing stones. Specifications mature over time. A real litmus test for the viability of the ISO/IEC DIS (draft international standard) 29500 (Open XML) is whether or not there are independent implementations. The answer to this question for Open XML is an unequivocal yes. There are independent Open XML implementations based on the existing specification in applications that run on Linux, Mac, Palm OS, iPhone, and Windows. My colleagues in Germany informed me recently that there are more than 70 in that country alone. I think the point is made that the spec is mature enough to drive independent implementations. Could it be better? Sure, and the comments coming in will certainly help achieve that goal.

Technical comments are valuable, and there is a commitment on the part of Ecma TC45 and the DIS editor to consider and respond to all comments. Furthermore, a proposal has been made by Ecma TC45 to establish a joint maintenance agreement with SC34. This is a level of working relationship the ODF TC at OASIS has not established with JTC1 (it is not mandatory as far as I understand it). The commitment is there to take the comments and resolution process very seriously.