Richard Fontana

I think Bradley is making two arguments there: (1) a license is a better legal structure to use than something denominated as a promise or pledge, and (2) the Red Hat Patent Promise shouldn't be limited to the set of free software licenses that it currently specifies but should cover all free software licenses. Both are legitimate points worth exploring and debating, and in my opinion do not have obvious correct answers; when Mark Webbink crafted the Red Hat Patent Promise several years ago it was not intended to stay fixed in that form for all time (indeed, it was subsequently modified in 2007, when GPLv3 and LGPLv3 were added to the list of covered licenses).

The idea that something like the Patent Promise could *as a policy* be changed prospectively -- to take what I think is a relatively harmless example, suppose Red Hat were to remove the largely obsolete IBM-PL from the list of covered licenses -- is true regardless of whether it's structured as a promise (the Patent Promise indicates that the public can rely on this promise) or a license (which is a kind of enforceable promise not to sue).

My grand legal theory, which I've discussed in some past talks, is that FLOSS licensing law is largely created by the actions of hackers. Therefore I think lawyers who are dismissive of hackers in the way you describe cannot fully understand FLOSS, while hackers who are immersed in FLOSS culture may be more qualified to explain licensing issues than traditional commercial software lawyers. I have encountered the dismissiveness, but I hope it is not as extensive as you suggest.