On Saturday, 22 November 2003, Lawrence E. Rosen wrote: 1. I think we *are* beginning to observe instances of the problem we are trying to solve. The EOLAS patents, for example, were such a threat that W3C formed a PAG that recommended reexamination (which the PTO just agreed to do). I've read emails from Bruce Perens identifying patents that have affected IETF standards. Perhaps others in cc have specific examples. On the other hand, what evidence can you adduce to suggest that there is some form of "natural immunity" for open source software? Well, yes but no, Larry. Eolas sues not the developers or distributors of Mozilla, but Microsoft, looking for the deepest pocket, as one would expect from a traditional patent plaintiff primarily concerned with recovering royalties. The W3C behaves as you and I want it to, and as we told our colleagues on the PPWG it would have to begin behaving: affirmatively challenging patents that threaten obstruction. That the members who were skeptical of that course of conduct when the Free World first raised the possibility in the PPWG in November 2001 found themselves doing it less than two years later is a sign of one part of the ecology's development: like Dan Ravicher's Public Patent Foundation, the W3C activity demonstrates, I believe, that one aspect of patent defense will be organizational, based on non-profit legal advocacy rather than market-based license-controlled activity.
But I'm not, as I said before, asking argumentative questions, and I wasn't leading up to "adducing evidence." I'm asking how to interpret negative evidence: the non-occurrence of actions against free software. In that context I think I hear you saying that it's an artifact: given more time, such suits will begin to happen in reasonable profusion and addressing them by license provision will be necessary. I don't say you're wrong, just that I'm not yet sure. 2. The escape hatch you describe can be prevented by appropriate license language. For example, many open source licenses define the licensee as "any entity that controls, is controlled by, or is under common control with" licensee. I've wondered if the EOLAS patent would be less a threat to open source browsers if the University of California, a passive plaintiff in that infringement lawsuit, were threatened with loss of its rights to open source software. If current language provisions like these aren't sufficient, can you suggest more effective language? You may be right here too, but it might be more complicated. I think the way a litigant intending disruption would probably circumvent the licenses would be to form a non-profit and assign the patents charitably. A party interested in royalties would either sell the claims to a patent-reclamation investor like Eugene Lang, or create an arms'-length investment vehicle. In the first instance, which I think concerns me most, I see no likelihood of convincing a judge that the donor "controls" in any relevant sense the non-profit to which it donated the claims. In the investment-entity case, any one particular free software licensor *might* be able to pierce the veil and show that the original entity should be found to have forfeited its copyright license under the patent retaliation clause, but the costs of enforcement and the uncertainty of success would have been multiplied substantially. Only the wealthiest free software licensor could afford that effort, thus making the mutual defense aspect of such clauses much less valuable. At any rate, that's the speculation that concerns me. Let's hear what other people think. Best. Eben
