Poor Microsoft. Even when they try to do the right thing (from the perspective of the FLOSS community), they get picked on unmercifully. The most recent case-in-point: the application of their Open Specification Promise to the Office Open XML 1.0 standard. Despite the irrevocable promise not to sue people for implementing OOXML, a recent analysis by the Software Freedom Law Center concludes flat-out that GPL vendors should not rely on the OSP and OOXML should not be approved as a standard.
The SFLC's concerns center around two issues in the OSP. First, it states "New versions of previously covered specifications will be separately considered for addition to the list." This appears to suggest that Microsoft could at any point decide that, say, OOXML 1.0.1 was not covered - though there are additional promises surrounding standards that are submitted to standards bodies. The issue is that it could become impossible for an open-source project to track continuing changes to standard at any time.
The second problem the OSP has is that the OSP only applies to code that is used to implement the specification. Re-use the code in another context, and it might become subject to a patent lawsuit. This is inimical to the free sharing of code between OSS projects.
Microsoft dismisses these worries. The most eloquent spokesman for Microsoft is Jason Matusow, who focuses on standards and interoperability work for the company. In a blog entry from January he says flat-out "There are NO intellectual property rights issues with Open XML." Though commenters on that post have drawn the conversation into a fairly technical dicussion, the bottom-line position of Matusow is simple: Microsoft has promised over and over again not to sue people for using this specification, and the relevant standards bodies are satisfied.
So who's right? In the end, this may be a "blind men and the elephant" issue. From Microsoft's point of view, they have made sufficient legally-binding promises to ensure that people can implement OOXML in their products. From the SFLC's point of view, those promises are necessary but not sufficient to guarantee desirable downstream freedom for the code. Since they're focused on different aspects of the issue, it's unlikely that the two sides will ever agree.
Do you think they ever will?Â