What angers me most is that we have to deal with software patents in the EU at all, despite the EU patent convention excluding from patentability "programs for computers as such".
Somehow, the "as such" has been interpreted to make the exclusion entirely meaningless, since algorithms in pure software running on generic hardware still get patented.
The farcical excuse is that software is not "software as such" and "is patentable if it provides a new and non-obvious 'technical' solution to a technical problem". But that is already a condition of patentability, so the interpretation is functionally equivalent to the exception being absent. Literally ignoring the law.
Without any doubt, open source and patent free software is among the factors that enabled progress in the past decades. But does this really mean that patented software doesn't have a place at all in our society?
Do you really think that if software couldn't be patented in the future there wouldn't be any downsides? Wouldn't the benefits of this decision rely on the patent release of code that was designed and developed with such an effort because it was certain that it could be patented?
> there wouldn't be any downsides? [..] the patent release of code that was designed and developed with such an effort because it was certain that it could be patented
I believe the downsides of software patents outweigh the upsides - simply too much monopolistic abuse is done with them, and developing anything new is a minefield where every two-bit "inventor" or troll or giant conglomerate from the last 25 years (or however long patents last in EU) will want a cut.
There is certainly software that patents would make more financially viable to develop (and assuming no state funding fills in the role of patents). But there is a much larger category of software that would have been developed anyway, that gets patented just to form a patent arsenal for extorting developers.
But that is a different question, on what the law should be. What the law in the EU is is that pure software is not patentable, yet it is being flagrantly ignored.
The problem is that basically nobody honest[0] is trying to patent software as a way to make money in their industry by licensing the patent. The honest people are either taking patents out for defensive purposes only, or not at all. The vast majority of software patents are deliberately designed to sound like they claim ownership over things which were state-of-the-art decades prior, so that a patent troll can extract quick settlements out of unwary defendants.
The one exception to this is MPEG, but even that is being destroyed by the one-two punch of maliciously-constructed patent pools with overlapping claims designed to get you to pay twice for the same patents... and AO1, which is threatening to both outperform royalty-bearing codecs and doesn't have the problems of three different companies trying to sell you the same thing.
The underlying problem is that - separate from any problems that patents might have in general - software developers don't use them, so the Patent Office is ill-equipped to actually adjudicate patentability and prior art in this particular subject field. In some alternate universe where CONTU hadn't massively expanded the scope of copyright by letting it subsume functional works like software, and everyone in the proprietary world was using patents, then maybe there could be an argument for it having a public benefit over no ownership at all. But as it currently stands, software patents are almost exclusively an extortion scheme.
[0] I'm going to include most proprietary software vendors in the "honest" category, too.
There's an extensive literature arguing that patents in general are bad, that I broadly subscribe to and recommend reading up on.
Against Intellectual Monopoly is one libertarian style book on the subject that you can read online.
However, if you're not quite ready for that step, I found persuasive the argument that patents help more (relatively speaking) when the boundaries are very clear. So if you give patents to people who discover a new theraputic use for a chemical compound. It's relatively easy for people to come up with a different compound and prove its not the same thing (proving that either are actually effective in what they claim to treat and without major side effects is a different conversation).
But anywhere there is ambiguity, like software, they become a tool for the bigger players to bully the littler players. See the conversation between Sun and IBM ceos where the Sun CEO tells him to get lost because he knows they dont infringe the patents he's claiming they do. IBM guy says, but are you sure I can't go and find something that will stick in our vast pile of vague software patents. Basically a protection racket.
Personally I've found European organizations and persons far more obsessed with patenting software and especially protocols than the US. One reason on the persons front is that it seems much more common to give employee engineer-inventors a share residual patent income, making them highly invested in seeing their patents deployed widely and enforced. Vs the US where its much more common for engineers to get a one time bonus.
This idea that the EU is special in excluding patentability "programs for computers as such" is nonsense: The US does so as well (except for the decade between state and bilski).
Somehow, the "as such" has been interpreted to make the exclusion entirely meaningless, since algorithms in pure software running on generic hardware still get patented.
The farcical excuse is that software is not "software as such" and "is patentable if it provides a new and non-obvious 'technical' solution to a technical problem". But that is already a condition of patentability, so the interpretation is functionally equivalent to the exception being absent. Literally ignoring the law.