You're correct on the narrow point that the patent termination provision in
Apache's proposed new license is of little value to the Apache Software
Foundation itself. Since ASF owns no important patents and probably never
will, the right to terminate their patent licenses if they are sued for
patent infringement will never be exercised by them. I assume that most of
Apache's individual contributors also own no patents and will not benefit
from that license's patent license termination. I don't think the author of
the proposed Apache license really thought through that provision carefully,
but merely copied it from one of the other licenses you criticized.
That would be mistaken. I am willing to rely on our community of contributors to act in our stead, since they do have patents and *their* patent grants may be terminated in such a case. That is why the proposed 2.0 license is phrased as being from the Contributors. If none of the Contributors have any applicable patents, then the situation is no better and no worse than for the current license. The ASF would have to cancel the project and/or ask for help to invalidate the patent.
That is another reason that the ASF should consider using the Academic Free
License, which *would* effectively protect Apache and its contributors. It
terminates both patent and copyright licenses, rather than just patent
licenses, in the event of a claim of patent infringement "with respect to a
patent applicable to software."
It is my understanding that such a clause would not be GPL-compatible,
though you'll have to ask Eben for a more authoritative reading. The 2.0
license is designed so that a recipient's right to copy and modify the
software remains equivalent to the GPL even if the patent license is
terminated. I'm not sure if that will still be true with GPL 3, but we
can revisit that when needed.
....Roy
