Open Source Fights Back: 'We Won't Get Patent-Trolled Again'

ZDNet's Steven Vaughan-Nichols reports: [...] At KubeCon North America 2024 this week, CNCF executive director Priyanka Sharma said in her keynote, "Patent trolls are not contributors or even adopters in our ecosystem. Instead, they prey on cloud-native adopters by abusing the legal system. We are here to tell the world that these patent trolls don't stand a chance because CNCF is uniting the ecosystem to deter them. Like a herd of musk oxen, we will run them off our pasture." CNCF CTO Chris Aniszczyk added: "The reason trolls can make money is that many companies find it too expensive to fight back, so they pay trolls a settlement fee to avoid the even higher cost of litigation. Now, when a whole herd of companies band together like musk oxen to drive a troll off, it changes the cost structure of fighting back. It disrupts their economic model."

How? Jim Zemlin, the Linux Foundation's executive director, said, "We don't negotiate with trolls. Instead, with United Patents, we go to the PTO and crush those patents. We strive to invalidate them by working with developers who have prior art, bringing this to the attention of the USPTO, and killing patents. No negotiation, no settlement. We destroy the very asset that made patent trolls' business work. Together, since we've started this effort, 90% of the time, we've been able to go in there and destroy these patents." "It's time for us to band together," said Joanna Lee, CNCF's VP of strategic programs and legal. "We encourage all organizations in our ecosystem to get involved. Join the fight, enhance your own company's protection, protect your customers, enhance our community defense, and save money on legal expenses."

While getting your company and its legal department involved in the effort to fend off patent trolls is important, developers can also help. CNCF announced the Cloud Native Heroes Challenge, a patent troll bounty program in which cloud-native developers and technologists can earn swag and win prizes. They're asking you to find evidence of preexisting technology -- referred to by patent lawyers as "prior art" -- that can kill off bad patents. This could be open-source documentation (including release notes), published standards or specifications, product manuals, articles, blogs, books, or any publicly available information. All entrants who submit an entry that conforms to the contest rules will receive a free "Cloud Native Hero" t-shirt that can be picked up at any future KubeCon+CloudNativeCon. The winner will also receive a $3,000 cash prize.

In the inaugural contest, the CNCF is seeking information that can be used to invalidate Claim 1 from US Patent US-11695823-B1. This is the major patent asserted by Edge Networking Systems against Kubernetes users. As is often the case with such patents, it's much too broad. This patent describes a network architecture that facilitates secure and flexible programmability between a user device and across a network with full lifecycle management of services and infrastructure applications. That describes pretty much any modern cloud system. If you can find prior art that describes such a system before June 13, 2013, you could be a winner. Some such materials have already been found. This is already listed in the "known references" tab of the contest information page and doesn't qualify. If you care about keeping open-source software easy and cheap to use -- or you believe trolls shouldn't be allowed to take advantage of companies that make or use programs -- you can help. I'll be doing some digging myself.

This is not how "prior art" works

[angel'o'sphere]

Bad guy has a patent (how dubious it might be)
You infringe on his patent.
I can proof prior art.

You have a problem!
I have not.

To invalidate a patent, you need more than "prior art".
Perhaps that is why they wrote it in quotes.

You have to show it is a standard procedure, everyone is doing.

Re:This is not how "prior art" works

[Entrope]

You are entirely wrong [wikipedia.org]:

Prior art (also known as state of the art or background art) is a concept in patent law used to determine the patentability of an invention, in particular whether an invention meets the novelty and the inventive step or non-obviousness criteria for patentability. In most systems of patent law, prior art is generally defined as anything that is made available, or disclosed, to the public that might be relevant to a patent's claim before the effective filing date of a patent application for an invention.

Emphasis added. Prior art doesn't have to be "standard procedure".

Re:

[Sique]

Standard procedure just makes it more easy to argue Prior Art. It is a sufficient, but not a necessary precondition.

Re:

[2TecTom]

the real problem is patent periods have been extended to beyond what is a reasonable return. Just like copyright, patents are now used to stifle innovation and extort money from enterprise.

this is exactly how classism works, the entitled upper class corrupts the law in order to gain ownership of all that has value, welcome to economic slavery

Re:

[dfghjk]

False.

Get yourself an education before offering your ignorance to others.

Great, but the real solution

[bradley13]

This is great - just like you should never pay off ransomware, you should not feed the patent trolls.

However, the long term solution has to be fixing the patent system. Require genuine innovation, and an actual product. Non-practicing entities should have no right to file patent infringement claims (since they have no products being infringed).

And obviously: patents on software or processes should not exist at all, in any form.

Re:

[iAmWaySmarterThanYou]

This is great - just like you should never pay off ransomware, you should not feed the patent trolls.

However, the long term solution has to be fixing the patent system. Require genuine innovation, and an actual product. Non-practicing entities should have no right to file patent infringement claims (since they have no products being infringed).

And obviously: patents on software or processes should not exist at all, in any form.

Products don't get infringed. Patents do. A patent is a how-to guide for others to follow explaining how a new idea and how to implement it.

I can legally make the same product as you without your permission as long as I don't use your patented ideas. If we had product patents then I could not. Your product patent idea is far worse than the system already in place so be careful what you wish for.

Why do you think software patents are special? An idea is an idea. Just because "it's on a computer" doesn't

Re:Great, but the real solution

[ShanghaiBill]

Algorithms and math formulas are already not patentable.

Algorithms can be patented.

An example is the LZW compression algorithm used by GIFs. Before the patent expired, you needed a license to create a GIF.

Re:

[iAmWaySmarterThanYou]

https://patentlawyer.io/can-yo... [patentlawyer.io]

"In short, you can patent an algorithm, but the patent will not cover the algorithm itself; rather, it will cover the specific application of the algorithm in a process or device. This distinction is crucial as it impacts how you can protect and leverage your intellectual property in the tech world."

Re:

[ShanghaiBill]

"you can patent an algorithm, but the patent will not cover the algorithm itself; rather, it will cover the specific application of the algorithm in a process or device.

I'd love to see that translated into English.

IBM's patent on LZW covered more than just GIF and more than just image compression. IBM was able to enforce it for any use until it expired.

Re:

[dfghjk]

That quote is English. Perhaps it just doesn't have enough money in it to interest you.

You cannot patent an algorithm. That's a clear as English gets.

Re:

[retchdog]

We know ShanghaiBill has trouble with English and probably reading in general: https://slashdot.org/comments.... [slashdot.org]

Of course if the money and the plebs team up to change things, reading comprehension doesnâ(TM)t matter very much I suppose.

Re:

[ShanghaiBill]

That quote is English ... You cannot patent an algorithm. That's a clear as English gets.

From the quote: "In short, you can patent an algorithm, ..."

So "can" means "cannot"?

Re:

[retchdog]

hey there ShanghaiBill, this is just an example of a common elision used by English speakers either to be polite by softening their correction, or simply to save time communicating.

let me help you read it.

literate people acting in good faith would parse it as "what you erroneously call 'patenting an algorithm' is actually done by 'patenting the implementation or device embodying the algorithm'."

since you could not or would not, the contrapositive directly implies that you are either illiterate in English or

Re:

[pauljlucas]

Just a guess: if they worded the patent such that the algorithm compresses data, any data, then that covers images, text, whatever.

Contrast that with an algorithm that does, say, flood fill in a paint program. If you were to try to patent it, it might be harder to come up with generic wording that would cover applications outside of paint programs.

Re:

[billyswong]

The patent law doesn't require a product to exist for a patent to be valid. You may think the law is wrong, but that makes sense before software patent is a thing. A person may design something but fail to get fund to materialize the design. So patent law in the old days prevent competitors or big capitals "steal the idea" by starving the inventor.

Obviously, if pure software patent couldn't exist, then there ought to be less patent troll. I think "pure software are already protected by copyright" shall b

Re:

[drinkypoo]

However, the long term solution has to be fixing the patent system. Require genuine innovation, and an actual product.

The patent law doesn't require a product to exist for a patent to be valid

That's literally what the GP was complaining about. Your comment makes no sense at all as a response to theirs unless you were trying to prove that you didn't understand it.

Obviously, if pure software patent couldn't exist, then there ought to be less patent troll.

Pure software patents definitely shouldn't exist, but if they do, there's no reason that you should be able to get a patent on the idea of some software. You should have to in fact create and demo the software in order to get the patent.

Re:

[swillden]

However, the long term solution has to be fixing the patent system. Require genuine innovation, and an actual product.

The patent law doesn't require a product to exist for a patent to be valid

That's literally what the GP was complaining about. Your comment makes no sense at all as a response to theirs unless you were trying to prove that you didn't understand it.

billyswong didn't state his point very clearly, but it was a good point. The reason that patent holders don't have to produce a product is because it's perfectly possible for an inventor to come up with a novel and useful idea but not have access to the resources needed to make it into a product, and requiring that the inventor produce a product before they're able to litigate their patent would mean those with the resources to make products could swoop in and steal the idea, leaving the inventor with noth

$3,000 cash prize?

[dfghjk]

$3,000? What an insult. Attorneys handling patent litigation could bill that in a day or two. They're offering you the tip on their steak dinner (a week's invoices) in exchange for you providing them with the most crucial argument in a case that will pay them for years. Fuck off.

Also, here's the patent: https://patents.google.com/pat... [google.com]

And here's the first claim:

1. A system comprising:
a programmable network device adapted to host a plurality of network device applications;
a programmable cloud device ada

Re:

[drinkypoo]

I have been involved in very bare bones multi-customer cluster management that definitely predates this patent (early 2000s) but the company has long disappeared and never open sourced it and IÃ(TM)m sure it wasnÃ(TM)t secure so it would not be prior art based on that specific patent claims

So it had no security? Because unless defined, "secure" doesn't mean much. It could be argued that it means industry-average access control mechanisms, but other than that it doesn't have a meaning because nothing is perfectly secure.

Re:

[dfghjk]

"Every semicolon should be viewed as an AND, not an OR."
Sure. If you read my post otherwise, you misunderstood. I explained each clause, there are really only two. There can be no interpretation with OR.

"The patent is for a cloud system (a cluster of devices) that allocates multiple virtual (software) clouds and each set of resources for multiple networks/customers through API independent of the state of the underlying machine. So basically, it is a patent for a Kubernetes-like system."
This is COMPLETELY

Re:

[Registered Coward v2]

However, if this group needs to offer a $3,000 prize to someone who can teach them the invalidation argument then they are incompetent. They are not serious people.

I think their approach is very much like the argument for open source software, get enough eyes on the problem by people with various areas of expertise and you will build a good argument for invalidating a patent. It's not about getting paid, but using the community's depth and breadth to tackle the problem.

Re:

[dfghjk]

Except invalidating a patent takes a lot of expensive legal work. Their approach offloads hard technical work to free "community depth and breadth" while reserving the profit for themselves. It is really just exploiting free labor by appealing to OSS principles.

As I said, this case is trivially easy to understand regarding how you would invalidate it. The fact they would make an appeal like this shows them to be incompetent charlatans.

CICS for the win

[xanthos]

IBM was doing this more than 50 years ago. CICS Transaction Server (CICS TS) sits at the head of the CICS family and provides services that extend or replace the functions of the operating system. These services can be more efficient than the generalized operating system services and also simpler for programmers to use, particularly with respect to communication with diverse terminal devices. [wikipedia.org]

"Am I really doing this AGAIN?"

[Pseudonymous Powers]

LUCY: [holding a football and bending over Charlie Brown, who lies supine and defeated] Again, Charlie Brown. Again and again and again and again...

VMWare was around before this patent

[laughingskeptic]

We refer to VMWare instances these days as "on-site clouds". There is nothing in this 2013 submarine patent (filed in 2013, multiply amended waiting for "cloud" to become a valuable term) that couldn't be done on VMWare in 1999.