Society would be better off if patents were abolished.
But like many other things, it has been exploited, especially by big corporations. And the patents no longer work as intended. And the big problem is that theres so much money on the line that is it not easy to change. I would assume large corporations already spend absurd amounts of money on lobbying for the patent-rules to work in their favor.
Patents as a concept I am all for. It is a good way to make sure that new entrepreneurs can make it to market before "big-corp" steals their idea and just push them out. But they would have to change how they work to stop patent-sharks and the big exploitations by said big-corp.
Moreover, there are also many patent claims that are accepted and that refer to things that are either impossible or non-economical at the time when the patent is issued, but they are put in the patent to cover future devices that someone will be able to make after the current technological problems are solved.
In the distant past, most patent offices required a demonstration with a working device in which all the patent claims were embodied, so such bogus patent claims against future competitors could not be accepted.
Now, every year more patents are issued than anytime in the past, but most of them are just garbage that was not patented before just because everyone else would have been ashamed to attempt to file such patent applications.
The number of truly valid patents issued per year now is actually less than in the 19th century or at the beginning of the 20th century.
The same as for new movie scripts, it becomes more and more difficult to find something really original to which nobody thought before.
This reads an awful lot like the no true Scottsman fallacy. Care to elaborate?
Therefore I base my opinion only on anecdotal evidence.
I have read carefully a lot of patents. I have read completely a few hundreds and I have browsed through the pages of a few thousands. I believe that few people happen to read so many patents, unless their job requires precisely this.
The number of the patents that I have read has been distributed rather uniformly over the years since the start of the USPTO archive until today. Because the number of patents issued per year has increased steadily, that means that I have read a larger part of the early patents than of the most recent patents, which are probably too many for anyone to read, unless they choose a very narrow area of applicability.
Besides a lot of junk patents, I have also read a lot of very valuable patents, which marked important points in the history of technology.
However, all the really valuable patents that I have seen have been issued before 1980. I have read a few patents from the eighties that were somewhat interesting improvements of prior techniques, but I have never seen any really innovative patent issued after 1990.
The patents that I have read were not chosen randomly, but they were found either by searching for the history of various important technologies or by following the references to other patents from the patents themselves.
I believe that the patents that I have read are more likely to belong to the better patents, not to the worse. Therefore I think that the failure to find recent innovative patents in my sample is likely to be consistent with the low proportion of such patents in the entire set of recent patents.
Not all patents are junk, many present competent solutions to various technical problems, but nevertheless they do not contain anything really new, they are just new combinations of already known techniques, that could be found after some work by any competent team that would be hired to solve that problem. Such patents fail the criterion for patentability of being "non-obvious for someone skilled in the art".
There are also patents that do not cover actual inventions but discoveries of various useful things, e.g. some new alloy that was not predicted by some theory, but it was just found in a series of empirical experiments, or some bacteria that were isolated from some environment and that happen to contain a useful enzyme.
Such discoveries are very valuable, but they do not have the novelty that they had much earlier when nobody had yet clear ideas about how to search for new materials or microorganisms.
Now such searches are well understood and it is mostly a matter of having enough time and money to enable such a discovery.
Their function was always to create more property to capitalize. I'm not sure you could argue that this wasnt the intended effect. The idea that they ever spurred innovation was the "big lie".
There have been historical periods of explosive innovation that coincided with lax or no enforcement of intellectual property rules but never, as far as im aware, the reverse.
By requiring the source to be available with the software, licenses drove the free software movement.
No reason Linux couldnt exist. All software would effectively be FOSS by default.
GPL is here to make sure the source code doesn't stay secret, and it is an essential part of the Linux development process, the driver model in particular. In fact, I don't see how a community driven, widely supported, monolithic kernel could be made without a copyleft license.
Linux without GPL wouldn't be Linux, and out of the free OSes (BSD, GNU/HURD, Haiku, Minix, ...), Linux is the most successful.
I think patents make sense to some degree, but they should expire much quicker and should not be transferrable.
Software/algorithm patents? Yes.
At least one of the Sonos patents in question is of that nature: it merely describes a system of distributed loudspeakers where one loudspeaker can set the volume of other speakers in the group remotely. The claims completely omit any actual solution that accomplishes that.
On other hand I do think there is still need to protect innovation in more physical things. Like new manufacturing processes or material innovations like using novel chemistry.
I think there should be some line, probably non-obvious implementation.
So while there may be a difference on software and hardware patents in practice, this is not a valid difference in terms of what the legal provisions are.
To be fair, some of the patents that listed in another comment here as part of the lawsuit are quite precise in their claims. Their claims cover the actual algorithms for integrating another loudspeaker into a secure WLAN and for ensuring exact playback. That's the kind of detail level that I would expect from a patent. As a result, if you solve the same problem using a sufficiently different set of steps, you're fine.
There's a difference between the spirit in which a law or regulation is written and how it is applied in practice over time. In case of patents, this gap has become quite large. I think that this is why so many here want see changes.
But no, for all the evil patents are we must have them, because some guy found a new way to make milk foam. Jesus ducking christ hacker news, this is almost as funny as Paul Graham writing an essay about why VC is great.
But mainly, my intent was to point out that patents might be a necessary evil to fight off copycats. Without that protection, I cannot imagine how any hardware startup could be successful, because the established players will always have the factories, tools, and experience so that they can scale up faster as soon as a new desired product is found.
Startups and labs should be able to patent so that they have a chance to make it in the market.