I've thought long and hard about free software, copyright, and licensing before, and I've had reason to seek the opinion of Richard Stallman on the matter at the time. Today, however, I read something of his and decided it was time to offer some advice in return. A little later I read some Slashdot comments which made roughly the same point, so maybe it's not remarkably insightful. Whatever the case, here's what I had to say to Mr Stallman. I hope it didn't come across as too smart-alecky.

(Note that the link to Dan Gillmor's weblog is no longer meaningful, since the paragraph I quoted has since been modified on that page.)

Richard,

You have in the past offered correction in the use of language to myself and countless others. I'd now like to return the favour with regards to your comments published by Dan Gillmor at the following URL.

http://weblog.mercurycenter.com/ejournal/stories/storyReader$664

In particular, I refer to the following extract from the second-last paragraph of the main article.

"If the Kerberos developers had released Kerberos under the GPL, Microsoft could not have undermined it in this way."

I understand the "this" from "in this way" to refer to Microsoft's tactic of embrace and extend. Thus a more cumbersome way of finishing the same sentence might have been, "Microsoft could not have undermined it by the tactic of embrace and extend."

If I've misunderstood you so far, then I think you haven't made your point clearly enough. If I'm on the right track, then I disagree with your assertion. Had Kerberos been released under the GPL, Microsoft could have done exactly the same thing, and the Kerberos community would have been just as powerless and incensed. It would have meant extra effort on Microsoft's part, to be sure, as it would have been necessary for them to reimplement the protocol rather than modifying the existing code, but I find it relatively simple to believe that they have the resources to do that if they consider it worth their while.

If the Kerberos protocol had been protected with a *patent*, then I'd be more inclined to agree with you, although there may be ways and means for working around patents as well.

In short, I believe you have done the GPL a disservice in this particular instance by implying it can offer a protection which it cannot. My suggestion would be to rephrase the original claim in terms such as, "If the Kerberos developers had released Kerberos under the GPL, it would have been necessary for Microsoft to reimplement the whole protocol from scratch in order to undermine it in this way."

On the other hand, this might not be what you want to say, since not only does it make the GPL look vulnerable, but the next logical step up from this would be to make the protocol proprietary. That would make it necessary for Microsoft to reverse engineer the software in addition to reimplementing it, whereas reverse engineering GPL software is much simpler because of the availability of generally well-documented source.

I'm speculating somewhat here, since I don't share your exact views, but perhaps the best direction you can take on the issue is to emphasise the dual need for maximising freedom whilst simultaneously combating proprietarists. The GPL can't guarantee that Microsoft won't embrace and extend, but at least it obstructs their freedom-tampering as much as possible whilst offering freedom to everyone else.

This is the point that you could make: the BSD license creates freedom but does nothing to combat the enemies of freedom. Proprietary software and secrecy combat their enemies, but at the expense of freedom. Only the GPL offers freedom and a means of combating those who would take freedom from us. This is more or less what you say in closing, but I feel you need to be a little more candid about the limited power of the GPL as a weapon against embrace and extend.

Regards,
TFBW

[On the off chance that you wish to use any of the above in your writings, I waive my copyright on it.]