Science would be far better off if it didn't have literal timed monopoly clauses tied to inventions. Any derivative or improvement is literally legally banned desu
With AI there could be an insurgence of control measures. If not already, like tariffs and production in the country, like chip manufacturers.
>>17009602The alternative is worse in practice. For inventions that are expected to remain extremely valuable for a long time or in areas where it's harder to get a patent, companies will just not publicize the invention at all and then bury their customers in NDAs and anti-reverse engineering agreements.
>>17009602What in particular is bothering you? Patents generally don't cover scientific knowledge, just specific applications of it for a particular use.
>>17009602If there wasn't some form of protection for the inventor/originator of process/proof as soon as anything practicably applicable was discovered largest company with necessary resources and flexibility to switch tracks would swallow up most of the benefits in most cases. Yes there are plenty of abuses, most actually through mechanisms forced into laws by abusers themselves, but the idea itself is logical and I would say necessary.And no matter how protected invention is for a period of time you are free to use it in your own research during that time. And you can use results of that research unless it includes said invention or is clear derivate of the invention.
>>17009621That would be much better, as other companies would be allowed to invent the same thing.
They are explicitly capitalist you fool
>>17009681They already are. If two inventors invent same thing and can reasonably show that they weren't aware of the other both get rights. But once invention is out everyone is assumed to have knowledge of it.Also, it is allowed to achieve same effect through different means. For example if classic typewriter that uses ink ribbon and transfers ink from ribbon to paper using individual levers each having one or two symbols is already out but someone creates another mechanical typewriter that uses rotating head with letters and carbon coated nylon sheets instead of ink that person would get independent patent. Same result, different mechanism of achieving it.
>>17009743>If two inventors invent same thing and can reasonably show that they weren't aware of the other both get rights.No, that isn't true. There are rarely multiple reasonable ways of achieving something.
>>17009602TRVKE, intellectual property is communist nonsense, doesn't exist and anti-innovation.
>>17009795Mentality of a literal parasite, all money all stolen IP. Returned.
>>17009795>property is communist nonsense
>>17009845intellectual property is not private property you moronic communist.If we both think 2+2=4 in our heads right now, is that idea private property?
>>17009681>That would be much better, as other companies would be allowed to invent the same thing.Dude, if someone has invented something it is silly if someone else reinvents it. That second person wasted his time, failing basic litterature search.
>>17009602OTOH, they're a way to draw attention to an idea.e.g. I, an unqualified shitter at home, could invent something about nuclear reactors. Nobody with any ability to make such a thing happen is going to give me the time of day. But if I flesh it out and get a patent on it? Boom, now it's official, maybe somebody sees the thing, pays me some pittance and runs off to build it.>>17009750>There are rarely multiple reasonable ways of achieving something.Go look up "types of X machine" on wikipedia or something, there are often many ways to skin a cat.
>>17010938That, done. 100% this time.
>>17009602they only last 20 yearsbeing forced to find different solutions to the same problem with different tradeoffs is the heart and soul of good engineeringthat being said, software patents are total fucking bullshit
>>17009743LOL. Elisha Gray and Alexander Graham Bell are nowhere near equally well known and only one of them got wealthy off the device both invented at the same time.
>>17009602Patents are also retarded because you are not required to implement a thing you declare, just>it will work brois enough. The entire system feels like a paper pushing bullshit work scam.
>>17009602>Any derivative or improvement is literally legally banned desuFirst off, you are allowed to file a patent application for an improvement on any technology, also where a patent remains in force. That is quite common, especially in pharmaceutics.Secondly, many jurisdictions outside the US permits use of patented technoilogy for research and education.And in practice, if you have a patent application on a significant improvement, you can always negotiate a cross license agreement with the one holding the patent of the base technology.
>>17010760It's how it's done anyway, patents just prevent the result from being used. A thing is designed, then it gets researched if it infringes on somebody's patents. Searching through literature isn't practical, nobody does that, and only patent lawyers read patents. >>17010938>there are often many ways to skin a cat.Almost always there is one that is obviously better, and the rest got invented to circumvent a patent.
>>17009681You can already invent the same thing as long as people don’t litigate. Most patents end up useless though
>>17009602>property is commie
>>17011318>communist deceptively removes the word privateNo such thing as communal property, no such thing as intellectual property, only absolute private property rights exist.
>>17011335Intellectual property is private property.
>>17011336It's literally not private, we can both have the exact same idea at the same time, they are not 2 separate ideas but just one same idea, thats what the word idea means, and it is literally not private, it does not fit the definition of what private means.But in the next breath you'll say communal property exists, which it does not, stupid fucking communist.
>>17011342It is private property in practice, communism didn't invent provate property, in fact all of it was state property in soviet onion.I don't know what you mean what communal property is. Im not a communist, Im a fascist.
>>17011343Whatever bro you don't believe in private property anyway so you can fuck off socialist, state property doesn't exist. communal property doesn't exist. intellectual property doesn't exist. total private property victory.
>>17011336Private property allows you to have something I own this horse. That doesn't prevent anyone from getting their own horse. Intellectual property allows you to forbid others from having something. You can't have a horse. I don't necessarily have any horses either, but I can forbid you from owning one.
>>17011449You can have it, you just can't distribute and charge other people for it.You can right now scan any book and have it sitting at home without CIA raping you any time.
>>17011053>It's how it's done anyway, patents just prevent the result from being used.Compulsory licensing is also a thing, again depending on jurisdictions.>A thing is designed, then it gets researched if it infringes on somebody's patents.Freedom to operate analysis is only needed if you plan to produce it for sale (some variations depending on jurisdictions) but I have never heard you need a license for design alone.>Searching through literature isn't practical, nobody does that, and only patent lawyers read patents.Patent searching is now a separate profession, though final analymsis is typically issued by a patent attorney. AI is also getting in on this line of work and I guess more than 100 companies now develop AI solutions for the IP professions.t.European patent attorney (TINLA IANYL)
>>17010943Yea. Because people are shit and abuses of law are widespread.And that's just the tip of the iceberg. Today, with right lawyers and connection you can patent almost anything, I'm afraid that if intel tried to patent binary code with todays practices they would succeed.
Since when is socialism about preserving monopolies you little uneducated shit?! .)
>>17011734What's the recommended course of action if a group of completely unqualified people make a temu version of your stuff then knowingly directly infringe your patents? cease and desist them? lawsuit? fraud charges?
>>17012088If it's actually a temu version, you can't do shit since the international PCT process is such a pain in the ass that most people just file regional patents.
>>17012121It is within Europe not actually a temu version.
>>17009602Patent copyright
>>17012088Process varies between jurisdictions but you normally start by contacting your counsel, patent attorney or attorney at law. There will be a claims construction to see if the claims read on the product, and if it does, it normally starts with a cease and desist, and then progress through a warning of initiating legal proceedings and in the end a court case. In most cases the issue will be resolved long before reaching the press. There are of course exceptions such as drugs (the legal kind).>>17012121Not sure how this can make sense.Patents are territorial, you have to have a patent in a country to be able to enforce it in that country. So a German patent cannot be enforced in China. Companies normally have a patent in their home country and can enforce that without having to use the PCT system.You don't need a PCT application and even if you had one it cannot be enforced directly. The PCT system is more like a common starting point that also gives you another 18 months before you have to decide what countries you have to enter in the national/regional phase. And it is not that hard, I prosecute these routinely. You really need to have a closer discussion with your patent attorney.>>17012127Are you and the infringer both in Europe?
Most of these laws are to protect businesses and the rich. Ring the police as a poor person and ring the police as a global business and see how long it takes them to arrive.
>>17012308Yes both Europe, pretty sure they sold said stolen IP for a lump sum which I will now probably have to spend even more time and money trying to get back so I am going to have the company I co-own start a lawsuit whenever possible.
>>17012324>Most of these laws are to protect businesses and the rich.That is a common misunderstanding, IP lets you stand up against the big corporations. Giants such as Microsoft has less need for IP (though they file a lot), they have more to gain by weakening rights so that they can steamroll small companies and individuals.
>>17009621>NDAs and anti-reverse engineering agreements.These should also be abolished.
>>17012480No, that just isn't true.Microsoft/intel are a great example of why patents don't work. x86 was already known to be obsolete when I was a kid. But it stayed, and eventually killed intel. The reason intel was this huge company wasn't that they were the only company who could make good CPUs. It was the company who owned the instruction set, and Microsoft was the company that demanded it to run their software. So basically you had to get intel if you wanted to run the vast majority of software. But it didn't end there. x86 was old. It was devised for chips in those arcade machines from the 70s and 80s. It was devised in times when machine code was typed in by hand, so that each instruction had to do as much computing as possible. But it wasn't the 80s anymore, but it was what intel had patented. So basically all their CPUs were not actually x86 CPUs. They were CPUs with a more recent instruction set, with an x86 interpreter running on top. But that was how it had to be done, because that was what intel owned, and once mobile devices became common, it collapsed, and intel went some sort of bankrupt, or I don't know how exactly it ended up.
>>17012775It wasn't Microsoft but IBM with their PC revolution. Also intel used underhanded techniques to ensure their monopoly over the instruction set. Only reason why AMD has forever crospatent agreement with Intel is to stop department of justice from suing Intel for monopolistic position.And I wonder what would have happened if 6502 had worthy successors, since that's still the most efficient architecture with todays ultra low power critical infrastructure devices like pacemakers still use that architecture.
(1/2)>>17012775>x86 was already known to be obsolete when I was a kid. But it stayed, and eventually killed intel.What does this even mean??68K was a comfier processor but the decision to go with x86 was more about availability and 8-bit compatibilities for peripherals.>The reason intel was this huge company wasn't that they were the only company who could make good CPUs.For a long time, Intel was supreme in fab tech. That, and also the microarchitecture, depends on trade secrets, not so much patents.>It was the company who owned the instruction set,And AMD had a license. Pentagon demands second sourcing, and did so also when 6502 was hot tech.>and Microsoft was the company that demanded it to run their software.You forget MS also supported Alpha for a while>So basically you had to get intel if you wanted to run the vast majority of software.So basically no.>But it didn't end there. x86 was old. It was devised for chips in those arcade machines from the 70s and 80s. It was devised in times when machine code was typed in by hand, so that each instruction had to do as much computing as possible. But it wasn't the 80s anymore, but it was what intel had patented.Patents last 20 years these days (used to be slightly different in the old days) so all the old base patents had expired by then.
(2/2)>>17013041>So basically all their CPUs were not actually x86 CPUs. They were CPUs with a more recent instruction set, with an x86 interpreter running on top.This makes no sense. The ISA is defined by what the programmer sees, not the microarchitecture. And microcode (vertical and horizontal) was old, even 6502 had microcode.>But that was how it had to be done, because that was what intel owned, and once mobile devices became common, it collapsed, and intel went some sort of bankrupt, or I don't know how exactly it ended up.Intel lost the mobile market since they refused to license RTL/Verilog/VHDL code, so would be licensees turned to ARM which, as a bonus, was not tied to specific design rules. And Intel design rules are awkward.Then Intel went all in on DEI and hired danger hair "engineers" who failed to deliver, to the utter astonisment of management. They lost several desig nodes and TSMC won the fab market. Annual capex of 50 billion dollars and an army of scientists with PhD hired for skills tend to deliver,. and deliver they did.Intel also hit stagnation in management an failed to refres into x64, that was done by AMD to the everlasting shame of Intel.Summary: x86 was always hideous and iyt got worse, x64 was better but they failed to clean up everything, and stagnation has still not been evicted.
>>17013041>Patents last 20 years these days (used to be slightly different in the old days) so all the old base patents had expired by then.They just added a couple if new instructions once in a while, and Microsoft demanded those as well. >>17013042>This makes no sense. The ISA is defined by what the programmer sees, not the microarchitecture.As I understand it I'm not a chip expert) there is an entirely different WLIW CPU inside an intel CPU. It isn't the first time in history when something like this happened. Watt is known for his steam engines, but what he did was that he stalled progress for 40 years with his patents, while he milked as much money as possible from Newcomen's old design, that used vaccuum instead of pressure, (it let the steam in, then condensed it) which was hugely inefficient, but necessary, because Newcomen couldn't make a cylinder that could last the pressure, or a piston that fit tightly enough. All the designs were abandoned after his patents expired, and high pressure engines were adopted instead.
>>17013299>As I understand it I'm not a chip expert) there is an entirely different WLIW CPU inside an intel CPU.Sounds like a horizontal microcode machine, then again there are probably several state machines in the processor, such as for handling the stack and prefetching. Not quite VLIW but there was a time when people wanted to make VLIW machines that ran close to the inner microarchitecture but it turned out not to work that well. Among otehr things, horizontal microcode is very verbose and will bloat the caches.>It isn't the first time in history when something like this happened. Watt is known for his steam engines, but what he did was that he stalled progress for 40 years with his patents, while he milked as much money as possible from Newcomen's old design, that used vaccuum instead of pressure, (it let the steam in, then condensed it) which was hugely inefficient, but necessary, because Newcomen couldn't make a cylinder that could last the pressure, or a piston that fit tightly enough. All the designs were abandoned after his patents expired, and high pressure engines were adopted instead.This story has many interesting aspects. First of all, his patent was extended by an act of Parliament. That violates the fundamental principle of legal certainty and people had all rights to be annoyed about that. The third party interests are important in the patent system. Problem was, the British companies that had planned to use the patented tech once the patent ran out of time, just kept on waiting.The second aspect, which is often masked by the above, is that on the continent, people moved ahead with combustion engines and gas turbines, none of which depended on Watt's patent. Diesel, Otto, Laval and more, none of these names are British.
>>17012784>And I wonder what would have happened if 6502 had worthy successors,I also wondered about that, and the idea of a 65000 pops up on many /ISA/ generals. I think they could have made a quick and easy stop-gap by rolling the SWEET-16 into the 6502. That would have fixed the only two problems with 6502: pointer handling was not optimal and 16-bit processing was slightly more complicated than it had to be.>since that's still the most efficient architecture with todays ultra low power critical infrastructure devices like pacemakers still use that architecture.Yes, the remarkable reliability means it can give you a lilfe time guarantee that is more literal than for most cases.Perhaps it is time for a new /ISA/ general.
>>17012775>They were CPUs with a more recent instruction set, with an x86 interpreter running on topthere isn't a modern CPU for which this pattern isn't true for, even ARM does this shit, so i do not think it is actually evidence of anything
>>17014319Internal combustion engines weren't relevant yet. They were useless until electric engineering made electric ignition possible. >>17014435ARM doesn't do it, it's hardwired. At most it sends multiple instructions through different pipelines.
>>17014647>They were useless until electric engineering made electric ignition possible.What about diesel?
>>17014660Diesel came later.
>>17014647>They were useless until electric engineering made electric ignition possible.There were alternatives such as hot-bulb and this onehttps://en.wikipedia.org/wiki/Pyr%C3%A9olophore
>>17013299>and Microsoft demanded those as well.That is an excellent way to annoy their costomers and make sure the customers stay on old versions of their software.Alternative: doubt.
