There is a great conversation happening in the comments section of my last post about independent implementations of Open XML. Andy Updegrove asked a few questions (and later made some observations) along with a group of folks who are actively building implementations. I'd encourage you to read that in conjunction with this.

Topics to tackle today (in as few words as possible) based on the earlier thread:

  1. What Is the utility of the Open XML specification in terms of implementation? Is that utility to be measured solely on the result of a competing office suite to Microsoft Office? Or are there other possible benefits?
  2. Are implementations of Open XML that were expressly done to interact better with MS Office somehow "less" than implementations of Open XML in a form completely unrelated to Microsoft Office?
  3. If there are indeed applications stemming from the availability of the Open XML specification that do not have anything to do with MS Office - is it important that they be backwards compatible with MS Office old binary formats?

The Utility of Open XML

It is important to remember that the European Commission (yes, I am quite aware of the CFI decision, and very aware that we have some more work ahead of us to meet expectations on interoperability) expressly recommended that Microsoft standardize the MS Office document formats. This happened in conjunction with our preexisting work on XML formats, customers requesting more openness in formats, the Massachusetts ETRM policy, ISVs looking for more ways to work with XML and our formats, etc., etc. My point here is that many, many people felt there would be great benefit to the standardization of Open XML.

The utility of a fully documented Open XML specification (commented issues and all) is manifest in how the individuals and organizations are making use of it. To me, this is a critical part of the discussion. The desire to have greater openness (I have written about this before) has led directly to increased choice, and a wider range of available solutions. The specification has created new economic opportunity which will ultimately lead to customer benefit. 

These implementations are directly competing with Office while others are utilizing it to work more effectively with MS Office, and there are those that have absolutely nothing to do with MS Office. This is exactly what was being looked for in requesting the format be standardized in the first place.

Working With MS Office - Good or Bad?

There is no reason to disparage an implementation of Open XML for working with MS Office or not. The value of those implementations will play out in the quality of the overall solution in which they are part. This is a big part of what standardization is all about. Does the specification result in market-viable implementations. OSI vs. TCP/IP ring any bells? Standards bodies are constantly on the lookout for work items that are relevant to the marketplace, and offer high value to implementers.

If working better with MS Office were the only way that implementations of Open XML worked, it would be hard to argue that there was no utility in that, nor that it did not create massive economic opportunity for the implementers. Better yet, this is not the case. The opportunity is considerable for non-Office-related work as well as for those with a direct relationship.

Backwards Compatibility

Backwards compatibility with old MS Office binary formats remains a core tenet of the Open XML work at Ecma. Vast amounts of data is stored in the old binary formats and TC45 is clear in the importance of maintaining the bridge back to that data.

But, is it absolutely critical for any app that uses Open XML as its data format to have high-fidelity backwards compatibility? Especially if it has zero relationship to the MS Office suite, or even to any recognizable office automation function? Probably not, but who am I to judge? It may be that they want to take streams of old Excel data into some new-fangled solution - ok. In my book, the option needs to be there, and certainly in the case of the Microsoft Office 2007 implementation it is something that our customers have told us is critical. Outside of MS, I can imagine a document management company building with backwards compatibility in mind, while someone doing a vertical market data analysis solution with a stand-alone reporting engine that pumps out Open XML dynamically to not care a whole bunch about it.

Implementation is often used as a measure of the viability of a standard. For those National Body representatives looking for clear evidence that Open XML is already a successful standard, is already providing the exact opportunity and choice it was supposed to - just observe the comments made to my last posting. The specification most certainly will be improved by the work done considering all of the comments. Maintenance will continue to improve it as well - and that will be done (hopefully) under the proposed joint maintenance agreement with SC34.

Just today I heard of 2-more completely independent implementations in France, plus an additional 8 or so doing improved integration with MS Office.

The fact that there are so many credible, and high quality implementations merits some deep consideration in the ballot resolution process this coming February.