> The FSF could have advocated for moving software out of copyright and into patents, where source code disclosure could be mandated as a condition of patent grants and we'd be flooded with public domain software today.
I think this is the one thing I do agree with. If there's something I think the FSF got wrong it's their staunch aversion to software patents. I understand the knee-jerk reaction to patent trolls, and I do agree that from a liberal intellectual property stance it's hard to rationalize patenting something that's not a physical invention and that software patents are in some regards akin to patenting the fastest route from your house to the grocery store. I concede we have had some pretty bogus software patents like "an object oriented operating system", but I think that as society matures in its understanding of software it will only get better, not worse, and sniffing out bogus software patents and there will only become more and more prior art. I also think that any serious company needs to patent their inventions. Once you do that the invention becomes public domain and you have the rights to license it in any style you see fit, just as you do with copyright, or, copyright still applies and you have the power to lay claim to all forms of your invention, not just your reference implementation. I think the FSF's focus on copyright is a result of Stallman stumbling into the discovery that he could hack the copyright system and he just ran with that. And of course it is much harder to patent something because it has to be novel and copyleft has that nice.. viral effect. The reality is that most people aren't doing novel things with software, they're just doing things with software. The FSF gets more exposure if it can apply in the later case.
I don't understand why we need to tear down the FSF to incorporate approaches that include freedom-focused patenting into the software licensing discussion. It's also not a new talking point. The GPL covers mutual patent assignment and I believe the FSF has advocated for people to donate patents to different software projects etc. There is a very real issue with patents and free software and there has been litigation in the past between patent holders and open source software projects that have implemented software off of a patented specification. The FSF doesn't exist in a vacuum.
I find it way more understandable than the massive rant we're all talking about and it makes a way more coherent argument to boot.
My personal opinion is that the best solution (if I specifically had the power to alter history such that CONTU's 1978 final recommendation was different) would have been to establish a sui generis right for software distribution monopolies that lasted 10 years and required source-code escrow. This is cribbed from the existing sui generis right for IC designs. Copyright would be explicitly barred from covering software, but it would still cover embedded creative works within the software.
Under this regime, a videogame would have it's engine code covered by software-rights, but it's art assets covered by ordinary copyright. This would mean that, for example, old videogames would be freed from the constraints of having to relicense middleware and engines. We wouldn't have massive preservation hazards like, say, Adobe Flash Player that wind up rendering entire art scenes inaccessible until manually remade by their copyright holders. By the time the software is abandoned, anything a decade older would be free to preserve and adapt.
(If you're wondering: this is the licensing model id took with their old game engines: once the engine was no longer being licensed, they just dropped GPL'd source code. You still needed to buy DooM if you wanted to play DooM, as the source drop didn't come with any levels or models, but if you just wanted to play with the game engine or build your own game with it, you were free to do so.)
Video codecs are one of the few areas of software where patents are preferred over copyright protection; or at least they were until H.265's three patent pools broke the licensing model and everyone fled to AOM AV1. MPEG is also trying their hand at another royalty-free codec (even though ISO's patent disclosure policies make it difficult to do so). So I wouldn't be surprised if the standard shifts to Free software implementations of unpatented codecs with widespread hardware implementations.
Part of the apprehensiveness people have over patents is that they're much broader. This is also why creative works are born copyrighted (you only need to register to sue) but inventions aren't born patented. You have to apply for patents and prove to multiple countries' patent offices that your invention is novel. But patent offices do a terrible job at this; their incentive is to approve everything so they can collect filing fees and let the victims of patent trolls bear the burden of proving obviousness to a court. That's why I would really prefer some kind of "copyright-lite" to a patents-based approach.
I think this is the one thing I do agree with. If there's something I think the FSF got wrong it's their staunch aversion to software patents. I understand the knee-jerk reaction to patent trolls, and I do agree that from a liberal intellectual property stance it's hard to rationalize patenting something that's not a physical invention and that software patents are in some regards akin to patenting the fastest route from your house to the grocery store. I concede we have had some pretty bogus software patents like "an object oriented operating system", but I think that as society matures in its understanding of software it will only get better, not worse, and sniffing out bogus software patents and there will only become more and more prior art. I also think that any serious company needs to patent their inventions. Once you do that the invention becomes public domain and you have the rights to license it in any style you see fit, just as you do with copyright, or, copyright still applies and you have the power to lay claim to all forms of your invention, not just your reference implementation. I think the FSF's focus on copyright is a result of Stallman stumbling into the discovery that he could hack the copyright system and he just ran with that. And of course it is much harder to patent something because it has to be novel and copyleft has that nice.. viral effect. The reality is that most people aren't doing novel things with software, they're just doing things with software. The FSF gets more exposure if it can apply in the later case.
I don't understand why we need to tear down the FSF to incorporate approaches that include freedom-focused patenting into the software licensing discussion. It's also not a new talking point. The GPL covers mutual patent assignment and I believe the FSF has advocated for people to donate patents to different software projects etc. There is a very real issue with patents and free software and there has been litigation in the past between patent holders and open source software projects that have implemented software off of a patented specification. The FSF doesn't exist in a vacuum.