pouët.net

http://www.pcgamer.com/company-claims-no-mans-sky-uses-its-patented-equation-without-permission/?utm_content=buffer1d7a7&utm_medium=social&utm_source=facebook&utm_campaign=buffer_pcgamerfb

So just came across this and was interested in how you can actually patent a bit of code like an algorithm for the universe generation or land generation or even map generation for that matter, like Diablo 3.

Wouldn't this code be similar to other generation code for other universe systems like pretty much all the "Elite" games and I would also assume Spore too.

I know a lot of you guys have either dabbled in these algorithms or have actually used them in demos like Elevated.

Obviously they're looking at getting money out of Hello Games, but this claim that they're actually using their patented code seems a little far fetched, wouldn't it be similar anyways, regardless? Essentially meaning that they would be borrowing their code regardless of them having seen it or not, because no matter what you do, someone trying to do the same thing will actually use their code.

Maybe it's me, but it seems like a coding trap.

Oh crap. aDDict 2 had supershapes in it :P

Meh, so what does this patent actually cover?

The formula itself or also all equivalents that result in the same output?

So what if I generate cos() and sin() via a (generic) look-up table? does the patent still hold then? what if I replace cos() and sin() by approximation polynomials, does it still hold then?

What if I go further and put all variations of the formula I'm going to use in one very big lookup table, does it still hold then? This would eliminate all direct usage of the formula from the code.

With a certain set of parameters they can generate a simple sphere. Does this mean that they can now claim ownership of all spheres because they can be generated with their formula.

IMO patents should only apply to the _way_ things are implemented, not to an abstract mathematical formula.

I'd say this is less bullshit than a lot of software patents are, since it refers to a specific thing that was clearly the invention of one person (the superformula / supershape, which probably has its best known demoscene appearance in San Angeles Observation - Johan Gielis even gets a credit in the nfo file). They're not just trying to lay claim to a vague concept like "shopping baskets on websites" or "building landscapes out of polygons".

The fact that the patent exists is a bit shitty... you supposedly can't patent maths because that isn't a physical invention, but make it a patent about doing a mathematical process ON A COMPUTAR and that's fine. But that's more the fault of the overall patent system than this one particular case.

All the claims in the linked patent include



Which I think the game does not do? Unless displaying it on a computer screen counts as "physical form"?

patents in code <3 ol

The wikipedia page for the Superformula claims the patent has been withdrawn in 2006, so this should be a none-issue? :)

What if simple genetic algorithm doing symbolic manipulation will reproduce this formula, who will be legally bullied then?

I had a comment ready when noone posted here yet, but deleted it again.
But it included them just trying some AttentionStunt to get their name in the press and that this patent is a joke if got through.
So if it was withdrawn in 2006 both must be true!

The source of the "withdrawn in 2006" claim (which was on the Wikipedia page for just over an hour before being deleted again) is this Reddit post from someone who pretty much admits that they don't know what they're talking about. (I don't either, but I notice that the patent numbers they quote don't match the ones given elsewhere.) YAY INTERNET FACTS

just what i thought.
and looking at->
http://www.pcgamer.com/author/anchalk/
the man is bought and paid for.
this is how games press works.

...huh?

1in10: are you suggesting that Hello Games placed the story?

When you come to think of it, the B SKY B claim was actually legit, so why shouldn't this claim be?

that seem to me most logical.
at least when i was working for games company the reviews were bought.
most safe way for all involved to do business.
it's the way of the world.

this all is hearsay of course and you shouldn't believe me.

I'll just un-lurk. My belief is that "Superformula" is no patentable. I believe that a specific implementation of it (maybe an optimisation to the process of creating a mesh from SF) is possible. But EPO and UKIPO, typically, don't like patent math. Even the Americans don't really allow it. My thinking is also that this is simply an extension to the super-ellipse formula.

At the last place I worked (and we did a LOT of patents) there's no way we would have tried to patent and algorithm or math. However, we would have tried to get one on a specific implementation.

Example; I have a patent on mip-map generation. Crazy? Not really. It's a patent on a specific piece of hardware.

Last word. Most patent offices are so busy they just rubber-stamp the stuff and let the lawyers sort it out.

kuiash: wait, what? The patent is granted, what is there to believe or not? It's also not on math,but on implementation on apparatus with memory and processing unit ;)

Yeah... the "on a computer" bit made me smile too. I'll do the same... but on a mobile device! Yeah!

As Steve Ballmer used to say: everything is a PC!

http://supershape.ahoi.in/

https://www.vertexshaderart.com/art/R9YtdbpwyPFwwKsix#

I find it very hard to believe Hello Games would really need the extra press nor would go to such lengths to promo their game.

(offtopic: hey @kuiash! nice to see you on pouet! :)