Microsoft Open Specification Promise Battle

The Software Freedom Law Centre published its analysis on Microsoft’s Open Specification Promise and GPL. Gary Knowlton of Microsoft believe they got it wrong. In general I agree with Software Freedom Law Centre’s analysis. In fact, I came to the same conclusion before the lawyers there publish their opinion. I started by clearing my mind of previous conception of Microsoft and GPL first, then simply looking at how the a layman will read and interpret two licenses and check for basic incompatibility. I found a few that SFLC highlighted.  Yes, I do not agree 100% with their analysis. However, if one actually understand their stance on “freedom” their conclusion is inevitable.

SFLC’s beef with Open Specification Promise is not about the actual text, but the legal minefield that any GPLed software have to thread carefully if it relies on Open Specification Promise. The problem is the fact that someone, in the future, might accidentally updated a piece of Free software to use an “updated specification” which was not covered by “Open Specification Promise”. For  organized Free Software, such as Linux and FSF offerings, the check-and-bound process would had caught it. Not true for the mum-and-pup software software shop me and a lot of free software people are running. It is us who are the most vulnerable, and SFLC’s opinion piece is aimed at us.

Gary Knowlton’s piece is a advertisement piece. The overall tone is you can use GPL with Open Specification Promise. Or is it? Read very carefully and you find that he does not actually say you can use GPL with Open Specification Promise. There is no statement that say “I, Gary Knowlton, put my reputation up on line by saying you GPL people can use Open Specification Promise”. It simply attack SFLC for publishing its opinion. It does this by pulling in existing Microsoft’s FAQ (non-committal and it is a classic sledge-of-hand where the answer does not answer the question) and other publicly available information from Microsoft which at best, dodge this issue. These are information that  SFLC has when it wrote its opinion and which I agree with SFLC’s analysis.

Moreover, readers should not read anything about how other companies collaborated with Open Source. Some of the accusation, like the same pledge, if come from IBM, will be accepted whole-heartedly by GPL people is not true. IBM’s pledges and Open Specification Promise are miles apart in detail. If anything, if Knowlton spent some time doing his homework, he can tell the difference. [Or since he is a smart person, may be he just choose to brush aside the difference.]

However, credits where credits due. It is true that there are a lot of pledge, particularly patent and specification pledges, that covers only existing versions, with the understanding that under the normal course of business, future versions can be reasonably expected be added to the same pledges, but NO PROMISE. Never mind Microsoft’s track record with Microsoft Office format, I am even going to dig into this part of the history book. What I will say is while it is conceivably Microsoft’s pledge is good enough for a lot of open source software, it is not the case for GPL.