Margaret Wasserman <margaret@xxxxxxxxxxxxxx>: > I am afraid that your choice below won't mesh very well with why > companies have software patents in the first place. You're right, it doesn't. Unfortunately, we really cannot live with anything less than I have described. My personal wish that this reality were different is as fruitless as yours. > However, in doing this, MMI wants to ensure two things: > > (1) That they continue to protect their patent rights regarding the > use of this technique for other purposes (such as cheese graters and > IP phones). So, they want to limit the royalty-free terms to > implementation of the IETF's WRAPCTL protocol and/or to use on > Wireless Access Points. (I think that this would run amok of your > proscription against "restricting the area of application", right?) Right. Area-of-application rules mean that code reuse can be fraught with dangerous legal problems because of someone's well-meaning judgment call about what's in-area and what is out-of-area. Such legal exposure is death to us -- our project groups can't afford lawyers to fight those battles -- so our licensing requirements have to foreclose that entire category of causes of action. Otherwise our development and distribution methods, which depend on code reuse and redistribution being safe and friction-free, would seize up and croak. Myself, I'll cheerfully recognize that MMI's interest is legitimate in some sense -- I wish I knew of a way to solve this problem that doesn't have deadly poison side-effects, and have devoted a lot of think time to trying to invent one. No joy. > (2) That they maintain and enhance the defensive value of their > patent, by making it clear that the royalty-free terms do not apply > to anyone who sues them for violation of a different patent. (If I > understand correctly, the OSS community doesn't have a problem with > this concept, as long as no paperwork is required?). That's correct. > Are there any IPR terms that MMI could offer that would meet these > goals while also allowing the WRAPCTL protocol to be implemented in > OSS implementations? If we could find that middle ground, I think it > would be very valuable to the IETF and to the OSS community. I don't believe any such middle ground exists. I and others have tried to imagine it into existence and failed. We regret this, but it seems to be reality. > The sad truth, of course, is that (in the example above) it is far > more likely that MMI would not determine that the WRAPCTL protocol > violated their patents until longer after it had been standardized by > the IETF and implemented in many commercial and OSS implementations. > So, what would be do then? Suffer a lot. What your scenario demonstrates is that there is a fundamental and nigh-unbridgeable conflict between open-source development and the patent system. The open-source community is already well aware of this, thank you. We don't know what to do about it either. -- <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a> _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf