On Sat, Dec 11, 2010 at 11:08:11AM +0000, Rob Myers wrote: > >>It's not "paranoia". It's a recognition that the task has been > >>necessary once, has been very difficult even after only a few years > >>of contributions, and may be necessary again after many more years. > > > >May be. > > And OSM isn't the only major free/open project that has had to be > relicenced.
No, it isn’t, and there is an argument for allowing some breathing room, but I think it is unnecessary to leave it wide open. > >>The upgrade clause means that another arbitrary licence can be > >>substituted anyway. See what happened with the FDL and Wikipedia. > > > >I agree to the upgrade clause in the ODbL. I do not agree to the broad > >“free and open licence” of the CTs. > > The reason I mention Wikipedia is that it shows that is not > sufficient to prevent relicencing. I don’t necessarily want relicensing to be prevented, but I think doing it should be discouraged. The Wikipedia relicensing was similarly a big effort, and they actually sought the clause, which was time‐limited, to allow them to relicense. The FSF could have just said “no” (but they listened to reason, and ultimately Wikipedia was still freely licensed). It was a big step, and proportionally a lot of thought went into it. A lot of thought has gone into applying the ODbL to OSM (sadly not much of it went to the CTs). Then we just give OSMF blanket rights, and define some very open conditions for relicensing, and the sense of proportion is lost. > >>A good example of a very successful project that decided it was > >>cleverer than the future is the Linux kernel. It can only be > >>licenced under GPL 2.0. This means that software patents, DRM, > >>Tivoisation, SaaS, internet distribution and other challenges to the > >>freedom to use software that have emerged since GPL 2 was written > >>and are addressed in GPL 3 and AGPL 3 still affect the Linux kernel. > > > >I don’t see how that affects this. > > You don't see how an actual example of licence lock-in having > detrimental effects on a project's users is relevant to a discussion > of licence lock-in? > >The kernel developers (rather > >Linus) chose to license under GPL v2 only for their own reasons. The > >above issues are completely irrelevant. > > Their reasons, whatever they were, have had detrimental consequences > for future users. The *fact* that this has caused issues is entirely > relevant. How widespread is this really? The types of devices where this has become a problem also tend to be running Busybox which has a history of pulling people up for licence violations. It gives the manufacturers bad press, and we get to avoid these devices for the free software friendly competition (ok, so there wasn’t much competition in the TiVo space at the time). We got new licences to choose from that countered “Tivoisation” and software as a service issues. Let’s not also forget the large projects, most notably Apache, that use even more permissive licences (the old GPL vs BSD arguments, oh the flames). > >I have never proposed that we go with ODbL 1.0 only, and have always > >accepted the upgrade clause as part and parcel of the licence. > > That's probably because it is. > > >>Yes, an upgrade clause is (on balance) good, although some people > >>regard that loss of control as immoral in itself. > > > >Opening it even more in the CTs, by that token, is more immoral. I > >wouldn’t say it’s necessarily immoral, but I do think it is totally > >unnecessary. > > I have provided historical examples of project licencing and > relicencing and I have argued that they show this not to be the > case. I do not think arguing by counter example is sufficient proof here. Those historical examples were special cases in their own rights, and a large number of projects have also survived without the need to ever relicense. > >>But that already removes the control of individuals over the licencing > >>other individuals can use in the future. And OSM has already ended up > >>with the wrong licence once. > > > >Yay, more fear. > > Which part of what I wrote there is factually or logically incorrect? I didn’t say it what you wrote was incorrect. I implied that you were using the current “wrong” licence choice as a reason for leaving it wide open because of the fear that it will happen again. I’m not after the freedom to relicense here, I’m after the freedom for the data to be useful. I don’t believe the freedom to relicense plays a large part in the continued usefulness of the data, the licence itself helps more with that, and if it doesn’t, why are we moving to it again? Simon -- A complex system that works is invariably found to have evolved from a simple system that works.—John Gall _______________________________________________ legal-talk mailing list legal-talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/legal-talk
