Can Microsoft be trusted on OOXML covenants?

analysis Developers wanting to use Microsoft's Office Open XML specification will need to brush up on their legal skills.

Academics studying Microsoft's efforts to promote its Office Open XML (OOXML) specification as an ISO standard have called on the software giant to clarify the language it uses on its licensing specifications, in order to make them easier for the technology community to understand.

OOXML, under consideration to become an ISO standard at a ballot resolution meeting later this month, can only be considered as a standard if offered on RAND (reasonable and non-discriminatory) terms.

To meet this criteria, Microsoft has made the 6,000 page specification freely available at zero cost. But further to this, Microsoft first released the standard under a "covenant not to sue", which has since been updated into a document called the Open Specification Promise (OSP).

The OSP is essentially a form of licence agreement designed to give software developers piece of mind that Microsoft won't come after them for patent infringement or other intellectual property litigation upon using the specification.

While Microsoft's intentions seem clear on the surface (a "covenant not to sue" is as indicative of the company's plans as any phrase), academics and legal experts have pointed out that the ambiguous legal jargon within both the covenant and OSP gives little piece of mind to the average developer.

Ronald Yu, a US patent agent and academic says that smaller developers should have some concerns about the vagaries of some terms used in the OSP.

Although terminology such as "required portion" is understood in the specialist patent and IP law community, Yu questions whether even the most technically well-informed and legally-sophisticated developer would be able determine exactly what things referred to within the 6,000 pages of the specification are affected by patents that aren't covered under the OSP.

If this coverage is not explicitly defined, he says, how can an organisation be sure Microsoft won't still be entitled to sue it?

The covenant, Yu warns, has never been tested in court with regards to IP rights, and it includes no mention of a forum (the court or tribunal and country) to resolve a dispute. It's not a release, he stresses, and nor is it a contract and therefore is something that needs to be treated with some caution.

David Vaile, executive director of the Cyberspace Law and Policy Centre at UNSW, says that Microsoft participants at a recent symposium on the issue found it challenging to explain how an ordinary person "or even an ordinary lawyer" could easily determine which parts of the specification were covered.

"This lack of certainty would mean a cautious lawyer may be reluctant to advise any third party to rely on the promise without extensive and potentially quite expensive analysis, and even that could be inconclusive," he said. "In turn, this could restrict its viability as a usable standard for less well-resourced users, including small developers and many public organisations."

Steve Mutkoski, regional director of interoperability at Microsoft and one of the lawyers on the team that drafted the licensing terms, describes the OSP as an "innovation" that enables a company like Microsoft to gain a quick and broad adoption of a technology without needing to negotiate 20- and 30-page agreements with every party that wants to use it.

Mutkoski says a reading of the FAQs at Microsoft's OSP Web page provides adequate clarification as to what the terms mean.

Further, he claims that Microsoft inherited or adapted the legal jargon in the OSP from similar pledges offered by Sun, IBM and Adobe. If the terms are a problem, he says, it's not a problem limited to the legal team at Microsoft but one that affects the wider industry.

But Vaile, while acknowledging the heritage of the terms, argues that the "everybody else is doing it so why can't we" explanation isn't ultimately a very helpful one for users.

"Just because other lawyers do it, it doesn't necessarily mean they are things that people should be willing to accept," he said.

"The ultimate goal is to reduce litigation, litigation risk, and legal complexity and expense. It is a higher goal to move beyond language that specialist high-end lawyers at major business partners are comfortable with to language which technical users or ordinary lawyers are also comfortable with in practical applications."

Vaile acknowledges that the Microsoft approach to its Open Specification Promise is similar although not identical to other approaches used by IBM, and to a lesser extent Adobe and Sun, and that these represent a substantial advance on the past practice of negotiating long case-by-case agreements.

An interpretation of Microsoft's Promises is made less straightforward by a lack of any specific identifier of exactly which elements in the 6,000 page OOXML Draft Specification are "necessary" in patent terms, and thus covered, as opposed to not "necessary" and thus still exposed, he said.

Another contributor to uncertainty is the reluctance of Microsoft to discuss the potential interaction of the OSP with Free or Open Source Software licences such as the GPL.

Mutkoski asserts that if the OSP is good enough for ECMA and good enough for the likes of Apple and Novell to consider their own implementations, that it should be good enough for the wider industry.

Vaile observes that ECMA's support, as a trade association that can potentially be influenced by business relationships, should not be seen as conclusive evidence as to whether a particular term is comprehensible to ordinary users of the document.

The likes of Apple and Novell too, he says, are able to "play the patent game as business equals" in that they have a patent portfolio to bargain with.

"Smaller players and non-IT firms, those who are not and perhaps cannot be in close business deals with Microsoft, are potentially at a disadvantage in not having either a relevant competing patent portfolio to bring to negotiations, or the legal resources to assess the level of risk," he said. "This is why a truly global open standard ideally needs to be dealt with in a way that removes obvious sources of uncertainty for smaller participants as well."