Rupert Goodwins' Diary

Friday 27/5/05

One of the single most annoying things about Microsoft is the way it misuses information. Today's kerfuffle over its habit of patenting the blindingly obvious is rebuffed by the company with a terse statement that "Studies routinely rank our innovations among the most significant across any industry. A study by the Massachusetts Institute of Technology in 2003, which provided an overall assessment of Microsoft's intellectual property, found that Microsoft continues to develop relevant patents and gave us one of the highest scores on the list of technology companies in that category."

Which study? Is that the one referenced in the MIT Technology Review article about the Patent Scorecard? Microsoft's not saying. It shows MS coming fifth in the computer industry, behind IBM, HP, Fujitsu and NEC: the score is basically the number of patents filed weighted by how many times previous patents from the same company have been cited 'in this year's batch'. There are no details of how many of those citations are from researchers outside the company — and since Microsoft has already submitted four times as many patents since the beginning of this year as it had in the whole of 2003, the current figures may be very different. But none of this is made plain by Microsoft, which is apparently content to adopt the legitimacy of 'an MIT study' without taking on the responsibility of identifying it or admitting what it might actually be saying. I could be completely wrong in identifying this study as the one that Microsoft means — Ingrid is on the case, and will get the complete story whether they want her to or not.

One is reminded of SCO's famous statement that a team of MIT mathematicians had been doing deep analysis of Linux and had found thousands of cases where it had been illegally derived from SCO's Unix. Thrilling stuff, until SCO refused to reveal any information about the individuals ("contractual obligations") or what they found ("You'll see it when we get to court") or what their connection was with MIT. MIT itself was baffled by the claim, as none of its mathematicians knew about any such analysis, and after much headscratching SCO 'clarified' that the team 'had links' to MIT, the details of which were to remain as much a mystery as everything else in the case.

In the case of Microsoft's XML patents, the mystery is what the innovation is. If I go back to the Swedish farm, could someone send me a morse code message when that's resolved? I could do with another few years out there...