[kelnos]

I'm a little confused what your argument is, then. You claim that software is not patentable in the EU, but then accept that some companies manage to get software patented in EU countries by writing their patents creatively.

So... software is patentable in the EU, at least in some circumstances. That's the end result, even if technically it's not allowed.

Unless these patents in question don't actually cover video codec software, which it seems they do?

The ultimate test, of course, is if these patents would stand up in court. But it seems no one has challenged them? Which indicates to me that either a) people believes they would hold up in court, and so they don't try to challenge them, or b) no one is sure, and is afraid to find out. Either outcome has the end result of "software can be patented in the EU sometimes".

[amiga386]

I'd say it's the difference between a country where stealing is a crime, but some people commit that crime anyway, and sometimes the stealing isn't detected, and sometimes it's detected but it's not worth the cost of prosecuting the thief... versus a country with laws like "steal all you want! it's not a crime at all! in fact we actively encourage it! stealing is great!"

jcranmer is trying to convince you that there is no difference between these two positions, because, technically, in both there is at least some stealing.

But they're not the same. Let's go back to the top-level question. Why is a US-based wrapper for a French library choosing to pre-emptively delete all binaries it has ever produced and completely shut down, out of fear of unstated legal actions, while the French library has been in active development for 25 years, and -- if software patents weren't explicitly illegal there -- would likely have been sued to oblivion by the same people who menace the USA.

[jcranmer]

Instead of trying to put words in my mouth, let me just say what I think:

I believe software patents are no more legal in the US than they are in the EU, and this is based on SCOTUS precedent in the US that makes the legal argument for software patents very sketchy. In a just world, that would mean that pretty much all of the MPEG-LA's patent portfolio should be worthless. But I suspect that in both the US and the EU, the fig leaf of "but I didn't patent the software algorithm, I patented the hardware running the software algorithm" is sufficient to get it granted and survive scrutiny to get it invalidated.

I don't have the time to look up the court case history of the EU--and I know that the EU is a civil law jurisdiction so the court case history isn't precedential in the same way it is in the US--but I would not be shocked if there have been cases that have upheld these software patents, despite the seemingly clear language in the treaties that software patents don't exist. Because the legal field is one wherein the "obvious" definition of software patent isn't necessarily the same as the legal definition.

The reality is that MPEG-LA positively asserts that it has a buttload of EU patents it can sic on your ass for violating, and were I being sued by them, I wouldn't be willing to bet that "but software patents are illegal!" is a winning defense.

[amiga386]

If you were to look it up, you'd find the European Patent Office (EPO) runs its own courts; boards of appeal, like the USPTO PTAB/BPAI. It set precedents on itself with appeal decisions and opinions. However, it is the member states (party to the European Patent Convention (EPC)) that enforce the patents they allow the EPO to issue in their name, and they all have their own laws and courts. There has been a recent development where some of the member states entered into a unitary patent court agreement.

The EPO has widened and narrowed the scope of what it considers patentable, and that gives people trying to purposefully get a patent on something that is clearly software either more or less leeway. Obviously I'd like the EPO to be even less forgiving.

But I don't think we're going to get anywhere. The EPO neither has perfect enforcement of its no-software-patents rule, nor is it easily hoodwinked with This One Simple Trick of saying it's a hardware device. The reality is in the middle, and neither of us has a good set of comparative statistics or case studies to contrast the differing EPO and USPTO approaches, so we should probably stop.

But I stick by my original claim: I believe the reason ffmpeg and VLC have not been sued to oblivion is because they're legally based in France, where the MPEG-LA's practises aren't tolerated. The support of a "no software patents" country/government/society encouraged them to develop ffmpeg without fear; the fear we see on display here today in the USA.

[jcranmer]

> I believe the reason ffmpeg and VLC have not been sued to oblivion is because they're legally based in France, where the MPEG-LA's practises aren't tolerated. The support of a "no software patents" country/government/society encouraged them to develop ffmpeg without fear; the fear we see on display here today in the USA.

I strongly doubt that's the reason. Until recently, it was quite easy to sue anybody in your favorite court in the US on the flimsiest of grounds, and even after SCOTUS said "you need more than 'they sell things here' to establish jurisdiction and venue," the patent-friendly judge in West Texas seems to be trying his hardest not to listen. No, I suspect the reason is much simpler: an open source project doesn't hold enough money to be worth the cost it takes to litigate (on the order of millions of dollars).