The silent treatment

CAN CENSORSHIP clauses buried in sneakwrap agreements really be used to silence product critics? I’m afraid so, and it appears that silence itself is one of the weapons that can be used to do so.

A reader recently asked for my help in an unusual situation. Eric Shulman is a computer systems engineer for the U.S. Air Force’s Space and Missile Systems Center, and his team had been re-engineering measurement software used with a launch vehicle. “We encountered a moderate bug in our embedded compiler, and we wanted to evaluate code optimization,” Shulman said. To do so, they ran benchmark tests, using three different compilers available at the Air Force center, including Microsoft’s Visual C++ Professional 6.0. “We weren’t benchmarking the operating system in any way, we just wanted to get an idea of the efficiency of our construct,” Shulman explained.

After running his test, Shulman decided to include the results in a technical paper he was preparing for an Air Force publication and for a presentation he will be making at a technical conference next spring. The problem arose when Air Force officials noticed that the Visual C++ EULA contained a term stating that “you may not disclose the results of any benchmark test … to any third party without Microsoft’s prior written approval.” As the EULA was phrased in a way that made him think it might be referring only to operating system benchmarks rather than Visual C++, Shulman was unsure if it really applied, but he decided to consult an Air Force attorney.

By now I hope all my readers understand that these EULA-based censorship clauses are almost certainly unenforceable in court. The Air Force counsel may have realized this because he told Shulman he doubted that he really needed to get Microsoft’s permission to publish. Nonetheless, the attorney pointed out that the existence of the EULA term put Shulman personally in an awkward position. “He advised me that if I didn’t ask permission and then Microsoft came after me, I’d be on my own because the government wouldn’t defend me,” Shulman said. “If I asked for permission, then I’d be covered.”

Given this situation and given that Shulman felt his benchmarks showed Visual C++ performing quite respectably, he decided it would be best to check with Microsoft. After all, he had almost a month before the deadline for the Air Force publication, so he assumed it would be a simple matter. He assumed wrong.

After searching Microsoft’s Web site in vain for the procedure to follow, Shulman’s phone calls to Redmond finally elicited a fax number where he could send his request. He sent the benchmark results for all the compilers along with his request, and — when he did not get an immediate response — followed up a few days later with a certified letter to Microsoft’s legal department. The weeks went by and the deadline for his paper approached, and all he got from Microsoft was silence.

Shulman ultimately had to submit his paper for the Air Force publication without the Visual C++ results. “Not being able to use the Visual C++ results hurt a bit, as they provided some interesting data points,” Shulman said. “I was still able to use the data (from the other compilers), as we had ascertained we did not need permission to publish those. It all just tells me and the Air Force that we can’t use Microsoft compiler products anymore. I am not sure this is what Bill Gates would want, but this is what is happening.”

Still hoping to receive permission to include the Visual C++ results in his spring presentation, Shulman at this point contacted me to see if I could find out what had happened to his request. I never did find out exactly how Shulman’s fax and certified letter had gone astray, nor could Microsoft spokespersons provide a better procedure for Shulman to follow for getting permission to publish. Nonetheless, after my query, diligent efforts by Microsoft’s PR folks did eventually yield contact information for a Visual C++ product manager who gave Shulman authorization to publish. Yippee.

So is all well that ends well? After all, while Shulman’s first paper appears without the Visual C++ results, at least they will now see the light of day. And neither Shulman nor I believe anyone at Microsoft actually intended to muzzle him. It’s easy to imagine what really happened to his fax and certified letter — it was a simple matter of no one in Microsoft’s bureaucracy wanting to touch them.

Even assuming that the legal eagles there understood Shulman’s benchmarks technically and recognized that they portrayed a reasonable showing for Visual C++, why would any of them risk being the one to actually sign off on it and getting called on the carpet by some big boss who felt Shulman’s numbers don’t further Microsoft’s goals? It was just far easier to keep quiet, and let his fax and letter fall through the cracks.

But let’s consider the real import of this incident. The really scary thing is how close Shulman came to being completely stifled, even when he had good news to tell about Visual C++. All Shulman wanted to do was share his research (paid for by our tax dollars, by the way). If he couldn’t get permission to publish without my help, what chance would he have had if his results had shown Microsoft’s product in a worse light? And if Microsoft can silence the U.S. Air Force without meaning to, what can it do when it deliberately sets out to silence a critic?

Yes, you could say that all that was lost here were a few numbers. But I say what’s really being lost to this creeping commercial censorship is an essential part of what this country is supposed to be about. And if a lot of people don’t start making a lot of noise, the silence will soon be deafening.