When Standards Collide: Microsoft Office XML and ODF

Lots has been written so far about Microsoft's recent announcement to take its Office 12 XML schema to Ecma International to standardize, and then use Ecma as an avenue to ISO standardization.  Microsoft then offered a "covenant not to sue" implementers of their future standard and this too has been well analyzed.  The best writing on the subject is covered here:

• Andy Updegrove on the covenant compared to Sun's on ODF, and compared to Microsoft's original Office 2003 XML patent licenses . Andy's commentary in this space has been balanced and deep since the discussions began and I would encourage you to track his blog.
• IBM's Bob Sutor has posted several excellent items covering the initial announcement and standards "factoring", as well as regular updates tracking the news and various letters to the Commonwealth of Massachusetts.  (Bob has also begun podcasting.)
  

The press should not be ignored on this issue and I've put a short summary of links at the end of this item.

I'll take a different focus here.  Let's demonstrate how Microsoft is either naive of the purpose of standards or arrogant in its attitude towards the very process with which they're claiming to engage.   

The interesting thing about this announcement is that clearly Microsoft doesn't understand standards.  Let's go back to first principles on this one. The economic purpose of a standard is to enable multiple implementations. Standards enable product substitutions to be made. As Tim Bray points out, Office represents a huge amount of the profitability of the corporation.  Honestly encouraging multiple implementations is the wrong thing to do from a business perspective.  The Board of Directors should be loudly questioning this strategy if that is Microsoft's intent. 

The reality is probably closer to an arrogance towards standards in general from the perspective of setting them and of their customers' desires to see them in the industry. Rather than implementing the standard that exists, (and we'll discuss in the next post why the PR value of this would have been high, while giving them maximum implementation freedom), they have chosen to try to confuse the market by offering up their own standard. 

They are trying to maintain the aura of standardization, while maintaining maximum control of what they perceive to be their property.  After promising to publish the patent license in the early part of the announcement, it has since turned into a statement that the covenant not to sue implementers on top of the original patent license is all they will be doing.  So the legal ambiguity begins. 

While someone could implement the Microsoft Office XML specification royalty free and without fear of directly being sued because of the covenant, some might not be able to distribute their product because Microsoft's patent license can not be sublicensed and this conflicts with certain free and open source licenses.

This is similar to the sorts of games being played out around C# and the Common Language Infrastructure. Microsoft took these technology specifications through Ecma International as well, who then took them on to ISO for standardization.  Rather than encouraging the one other substantial implementation of these standards beyond Microsoft's .NET platform, it continues a cloud of legal ambiguity around the Novell mono implementation. 

A similar game will be at hand around document format standards if and when the Microsoft Office XML specifications complete the cycle at Ecma International and later at ISO.  In the mean time (2-4 years), they will loudly continue to hammer home that government and enterprise customers should trust them that they're building a standard which will be royalty free to implement.

They will have Jason Matusow front and center on this mission, based on his recent announcement about a move to Microsoft's corporate standards team.  I wish Jason well in his new role, but respect too much his ability to manipulate and frame the language of the debate as he did introducing the term "shared source" into the industry and spending a lot of time with governments, the press, and analysts selling them on the idea that most people didn't care about source code anyway and Microsoft's sharing programs were basically equivalent to free and open source software. For the past couple of years he has been driving home the "move to the middle" message.

A successful standard is measured by the number of implementations.  This is as opposed to a vendor proprietary specification designed to enable multiple complement products but prevent the actual substitution.  If there is only one implementation of a standard then it's an economic failure, regardless of the imprimatur of a standards organization.  Unfortunately in the market, confusion may reign for some time before customers realize the problem. 

Looking at Microsoft's original press release, it would appear that they're positioning the entire move as an archive format.  That is essentially what it will become -- an archive standard -- if they can't generate a serious second implementation.  So customers will be able to work with ODF for all active documents as that standard takes hold with multiple genuine implementations, and customers can still have confidence that during the transition, they will be access the Office 12 documents produced in the interim period. Of course this won't be the way Microsoft will position it going forward.

The number of ways Microsoft is being arrogant here is impressive:

• It's arrogant towards their customers with respect to the needs of the Microsoft business machine versus the needs of the customers. (They could support ODF then demonstrate they really can innovate better solutions in their own products.)
• It's arrogant in its survival instinct around protecting its market position by deliberately sowing confusion around document format standards. ("Our document format is so much better that it should also be a de jure standard so people that want a standards-based implementation can keep buying our one true implementation".)
• It's arrogant towards the very standards processes and partners at Ecma International, the new technical committee, and ISO that Microsoft is proposing it supports.  ("Trust us.  We'll have our proprietary specification through the process in 12 months.  Then ISO can approve it quickly too.")

It's really too bad.  It should have been played out very differently, which we'll look at next. 

The press links:

• CNet (22 Nov 2005) Microsoft's Standardization Move Divides Experts
• ZDnet (22 Nov 2005) Excellent blog entry by David Berlind.
• Nicolas Carr's (22 Nov 2005) Distrust and Verify
• eWeek (22 Nov 2005) From the Outside Looking In: Analysts, Developers on Microsoft, Open Standards
• eWeek (22 Nov 2005) A heated editorial response from Steven J. Vaughan-Nichols
• Boston Herald (24 Nov 2005) Open Debate Flares: Sun says Microsoft Vows can't Stand the Light
• eWeek (29 Nov 2005) Microsoft Drops the Office Open Standard Ball

Comics used with the kind permission of Tatsuya Ishida.