4 ms·
But real inventions can occur in software too? Inventions that takes serious money, expertise and so on. Some examples: - Rust memory management - New models
by commitpizza 4y ago
But real inventions can occur in software too? Inventions that takes serious money, expertise and so on.
Some examples:
- Rust memory management
- New models in machine learning that enables amazing results
- Protocols
- anonymous_sorry 4y agoBut is there broader benefit to a legal system that enables the inventor to enforce time-limited monopolies on them? Rust's memory management was invented without any such incentive, and the world is richer for it.
- bombolo 4y agoSure they can happen, but in general they require a much lower investment to be done, and are easy to reinvent. Once I had an idea while sitting on the toilet, about how to use past log data to pre-emptively scale up, when the peak hour is about to come. I had the idea just randomly sitting on the toilet for a few minutes… it was already patented.
- asah 4y agoThis cuts both ways: it's also easy to "surround" software patents with new innovations that are required to be competitive in the marketplace. Then both parties cross-license and you're good. Yes the lawyers get paid but the cost is pretty low compared with software engineers, and the societal benefit is that these innovative move into the public domain.
- commitpizza 4y agoI don't understand your point, do you mean that this cannot happen in other fields? I imagine that this is the case for most stuff in the world. Ideas are not usually unique but it's the great execution of the idea that present the real challange.
- bombolo 4y agoWell do you think a new pill is done thinking 10 minutes? Just the testing procedures involve massive investment. I'm not in favour of patents at large by the way, but they make even less sense for software.
- oneoff786 4y agoSounds to me like all of that would be better off public or as a trade secret.
- nonrandomstring 4y agoTradition has held that these are ideas, not inventions, and thus obtain value precisely because they can and should be freely shared (which is the ultimate aim of patents anyway). Ideas may also require intellectual labour. But one may not receive the protection of the government for them to give temporary advantage over competitors. Ideas are a collaborative venture, protocols being the very epitome of this - since a "protocol of one" is a bit like a birthday party for one... a bit sad and pointless. If one wants to make money, by all means invest time in pursuing patentable inventions, but do not presume that mere ideas (most of which are "nothing new under the sun") should be afforded the same protection. The problem we have today is that the patent system is derelict. The goalposts have shifted to allow almost all and any silly idea to obtain a patent and the system itself is weaponised for extraction. It has not shown any will to reform, so abolition may be its ultimate fate.
- abduhl 4y agoI’m interested in what “traditions” you’re talking about here? Honestly, it is not even clear to me what you’re talking about when you say “these” in your sentence. Certainly tradition hasn’t held that Rust memory management is an idea and therefore not patentable. Same with new models in ML? Maybe there is some tradition for protocols, but even those are generally patentable but required to be licensed via FRAND. Your post seems to entirely misunderstand the patent system. “But one may not receive the protection of the government for [ideas] to give temporary advantage over competitors.” This is just flat out wrong. The purpose of a patent is literally to receive the protection of the government for your ideas in order to give you a temporary advantage over your competitors via a government granted monopoly on an idea. The bargain that’s struck is that the patent owner gives the details of their patented idea to the world so that anyone can eventually practice it or build off of it in exchange for a limited exclusionary protection in commerce.
- nonrandomstring 4y agoI'm sorry I can do nothing more to help with your misunderstanding of my good faith post. Please do the necessary research with regard to the history and purpose of designs and patents law.
- dbfx 4y agoI'm skeptical rust memory management is such a big invention by them and would bet some money the ownership model already existed before in papers at the very least. And even if not just imagine having to pay Dahl or Alan Kay to implement OO in your language, which was by every possible metric more innovative than rust's memory management. Even if it would have stopped java from being a thing it's just not worth it.
- anonymous_sorry 4y agoThose things got invented anyway though, without being patented. The right question, in my opinion, is whether there are important software innovations that would have been significantly delayed, or kept secret, or never discovered at all, if software patents weren't a thing. And if so, do they outweigh legal costs and chilling effects of patent trolling, and the inefficiencies of people having to engineer around patented ideas. Intellectual property is a fairly artificial concept. It's quite "big government" when you think about it - government grants a monopoly on the application of an idea. It is only worth preserving if it makes the country richer.