I don't think he misses the point. He just doesn't find it a good one. I mean, you are of course entitled to your opinion, and to call his a mistake, but saying RMS "misses the point" of what he's been working on almost 30 years is a bit disingenuous.
And I personally disagree with the statement that BSD-like licenses are definitely better than GPL. As always, it depends. What would have happened to Linux without GPL? Would it support as many devices as it does? Would it be as widespread as it is? On the other hand, I agree that a BSD-like license has been a plus for PostgreSQL, by permitting the existence of various 'PostgreSQL enhanced' companies selling their versions and offering support.
What's ultimately more important than the license is the people working on the project and the community that forms around it. They are what ultimately lead to success or failure, not the choice of a license.
> What's ultimately more important than the license is the people working on the project and the community that forms around it.
Is it? Look at the router industry - licenses like zlib, the busybox license, and the "no copyright" sqlite licenses have enabled an entire industry to take open source software and use it for profit without having to invest developer hours to making it themselves. But they still sell that software, and take away their users freedoms, on their routers. Every time I buy a new one, I have to either look for one with an open source firmware already installed, or be very careful to make sure one I get is supposed so I can reflash it (if its even possible on some models) to reclaim my freedoms.
Tons of other embedded devices do the same. As long as Linux at least is gpl v2, they can get away with using the kernel as well as long as they don't modify it directly.
And who knows how much open source code Microsoft and Apple have in their OS installs that has been heavily modified without any of that work being contributed back to the source projects while those companies profit off license key sales?
> Every time I buy a new one, I have to either look for one with an open source firmware already installed, or be very careful to make sure one I get is supposed so I can reflash it (if its even possible on some models) to reclaim my freedoms.
Now, why do you do this? What motivates you? Is it ideology? Or do you have a practical reason for wanting that power over the appliance?
Now, if you have a practical reason, then you have proven what I have been arguing, namely that freedom is of economic value. I.e. freedom is important not because of some abstract value but rather what you can produce with it.
If we accept that, then in the long run, software with more freedom will outcompete software with less freedom.
One of the great things about software is that it doesn't wear out from over use.
So what if there are a bunch of router companies selling proprietary products consisting largely of (permissive) Free Software components? It does not harm to those projects; in fact, they probably get more bug reports and other fringe benefits. So what if someone's making money off those components; is that money somehow being taken from the authors? No. That would be a claim far more dubious than even the RIAA's.
We can ask "what if those components were copyleft?", and I can think of three answers:
1) All of this whizz-bang technology we have will also be copyleft, and hence Free Software.
2) We'd have completely unknown, secret, proprietary technology in its place. This would presumably be less capable, more locked-down, less secure and likely much further behind than what we have now (since it had to be written from scratch).
3) We wouldn't have that kind of technology, since nobody finds it profitable enough to develop.
Scenario (1) wouldn't be an inevitability if those components were copyleft. Those companies aren't "getting away" with using Linux; they're using it in a perfectly legitimate way. If you find that usage unethical (like I do) then strive to:
1) (ideally) Replace their technology with something better.
2) (acceptable) Avoid using it (we managed for thousands of years without issue).
3) (unfortunate) Put up with it, but support those who are trying 1 and 2.
This is the approach I take, for example with regards to social networks.
I don't understand your point. There are people taking advantage of both GPL and BSD. Some (not all) of the GPL violators can be made to comply, and none of the BSD "violators" can be made to comply, because they are not actual violators.
All of the previous statements do not alter at all that all the software pieces that are "taken advantage of" are huge value contributions to society at large and that their success is not dependent on the license choice.
If you are angered by "other people profiting with my free software" you will of course choose GPL to have an option to sue. If you don't care, you will choose BSD or GPL for another reason. Neither option will make much of a difference on the success of your project.
> There are people taking advantage of both GPL and BSD.
But whereas it is fine that they take advantage of the GPL, because they must publish their modifications and anything linking against it, with BSD they can make all the changes they want, link to whatever they want, and sell the result.
I get you are fine with that, but some of us aren't. Which is why we use copyleft licenses and try to explain out positions like this.
> are huge value contributions to society at large and that their success is not dependent on the license choice.
LLVM vs GCC - Apple effectively built the former ground up because the latter wouldn't budge on the GPL. There are a lot of parallels - be it busybox vs coreutils, or bionic vs glibc - where the license does matter, and what happens is businesses that are uncomfortable with providing freedom spend the effort (which is entirely in their right) to rebuild the whole thing under a different license.
The license does matter. And the GPL does help, because it means you either need to preserve your users freedoms or rewrite it yourself. It is a great motivator to drive the ideology if you can put great software in front of someone and say "all it takes is you behaving ethically, and you can avoid having to write this all by yourself". But the license really matters, it is why there is this conflict of interest in the first place and not everyone just uses the GPL without hesitation.
> But whereas it is fine that they take advantage of the GPL, because they must publish
As long as they comply. There are lots of people who don't comply and are sued where possible, which is not at all in all cases.
> LLVM vs GCC - Apple effectively built the former ground up because the latter wouldn't budge on the GPL.
And how is that better for the world, duplicating functionality and reinventing the wheel?
And what I've said twice (and now thrice) is that the license choice does not determine the SUCCESS of a project. People are a way more important factor.
As you said, you need that great software and awesome community, but you need it FIRST.
Yes, I'm aware of the monoculture argument. But anyway, the raison d'etre of the GPL is to preserve freedom, not to avoid monocultures, so, even if in this case we might be able to see a good outcome (even though that is still debatable as we have more than one compiler, gcc is big, but far from the only compiler) it's not actually relevant to the question posed in the general case, unless we think the only benefit of the GPL is to avoid monocultures...
And I personally disagree with the statement that BSD-like licenses are definitely better than GPL. As always, it depends. What would have happened to Linux without GPL? Would it support as many devices as it does? Would it be as widespread as it is? On the other hand, I agree that a BSD-like license has been a plus for PostgreSQL, by permitting the existence of various 'PostgreSQL enhanced' companies selling their versions and offering support.
What's ultimately more important than the license is the people working on the project and the community that forms around it. They are what ultimately lead to success or failure, not the choice of a license.