by Dennis Crouch
Rain Computing, Inc. v. Samsung Electronics (Fed. Cir. 2021)
Rain and Samsung agree that this case comes down to claim construction. And, as typical, the patentee is attempting to thread the needle with a construction that is broad enough to be infringed, but narrow and specific enough to avoid invalidation. The district court construed one term narrowly — resulting in no infringement. On appeal, the Federal Circuit focused on a separate term — and found its indefiniteness rendered the claim invalid.
Rain’s US Patent 9,805,349 covers a method for delivering apps via a computer network using a webstore and server authentication. The claim requires “a user identification module configured to control access of said one or more software application packages.” If you have been following patent claim construction over the past decade: You know that the focus here will be whether this “module” limitation invokes 35 U.S.C. § 112 ¶ 6 as a means-plus-function term. And, if so, is it indefinite for failing to provide sufficient structural support in the specification.
Normally, patent claims are expected to spell out the actual structure of the invention — not just its function. However, § 112 ¶ 6 (now renamed § 112(f)) allows a patentee to avoid adding those specifics in combination claims by including an element “as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof.” The statute goes on to include an important caveat, the element “shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” The statute does not spell-out what happens when the specification lacks any corresponding structure — but the courts have filled the gap. A means-plus-function claim that lacks corresponding structural disclosure in the specification is deemed invalid as indefinite.
Means-plus-function claiming is fools gold — especially for anyone with a thin disclosure. The claims appear broad, but are narrowly interpreted and regularly invalidated. For years, patentees have moved away from using the term “MEANS” and replaced it with some other generalized word such as “MODULE” as used in this case. By skirting the MPF trap, these newly styled claims were more broadly interpreted and rarely invalidated as indefinite. However, that approach was collapsed by the case of Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015). In Williamson, the Federal Circuit held that claim terms that “fails to recite sufficiently definite structure or else recites function without reciting sufficient structure for performing that function” will also be interpreted as a means-plus-function term. Although the burden on invoking 112(f) is still on the patent-challenger, Williamson eliminated the heavy presumption that it doesn’t apply.
The further trick with Williamson: (a) If a particular term – such as “module” is structurally defined within the specification then it will typically not be interpreted as means-plus-function because it was particularly defined by the applicant. (b) On the other hand, if the term is not structurally defined in the specification, then it will invoke 112(f) and also be found indefinite for lack of structural support.
Here, the court first noted that the word “MODULE” “is a well-known nonce word that can operate as a substitute for ‘means’.” The Williamson case itself focused on the word module and found that it did not indicate any “structure.” The court drew the same conclusion in Rain Computing— finding that the claimed “user identification module” has no commonly understood structure and the specification included no structural disclosure.
Here, the specification does not impart any structural significance to the term; in fact, it does not even mention a “user identification module.”
Slip Op. So, we fall into the category (b) from above: Nonce word without structural support → therefore → Invoke 112(f) and invalidate. The court writes:
Nothing in the claim language or the written description provides an algorithm to achieve the “control access” function of the “user identification module.” When asked at oral argument to identify an algorithm in the written description, Rain could not do so. Without an algorithm to achieve the “control access” function, we hold the term “user identification module” lacks sufficient structure and renders the claims indefinite.
Slip Op. Thus, the claim is indefinite. This rendered the separate non-infringement related claim construction moot.
During prosecution the patent examiner had expressly stated that the limitation “does not invoke 112, 6th paragraph, or 112(f).” However, the examiner’s somewhat cryptic reasoning was that a 112(f) means limitation cannot exist within a method claim. On appeal, the Federal Circuit noted the examiner’s error: “Applicants are free to invoke § 112 ¶ 6 for a claim term nested in a method claim.”
= = = =
Note: In Synchronoss Techs., Inc. v. Dropbox, Inc., 2019-2196, 2021 WL 520047 (Fed. Cir. Feb. 12, 2021), the Federal Circuit encountered a claim a quite similar limitation: “user identifier module.” In that case the court also found that the term invoked 112(f) but was indefinite from lack of structural support.
(Fed Cir Clerks – note that FN 1 in Rain Computing does not accurately reflect the language of the claims in Synchronoss).
= = = =
Above, I noted the draw of means-plus-function claim language for “thin disclosures.” In this case, the disclosure was actually quite extensive, but the patentee appears to have wanted to shift the scope of coverage away from the original focus of the invention disclosure. This patent is one of nine US patent applications all within the same direct family and relying on the same original disclosure.
J. Doerre
I think my point was crystal clear. Moore extended Williamson by holding that if a claim element doesn’t recite in of itself structure for performing the function that even if known solutions are readily available that the claim elements still invokes 112(f).
And my diode example is spot on. It illustrates that under Moore’s reasoning that a functional description of a diode would invoke 112(f). Moore’s reasoning has nothing to do with modern innovation practice and invalidates probably 10’s of thousands of the claims of patents granted before Moore’s opinion.
Moore is an absolute disgrace. This was judicial activism to invalidate a patent that Moore didn’t feel deserved a patent under 103. Now we are all stuck with this extension of Williamson. J. Rich probably would have figuratively have spit in the face of Moore.
Violating the maxim of “Ends do not justify the Means” ==> Patent Profanity
Again.
A note for future references, the discussion of this thread has spilled over to two newer threads — due in part to the fact that editorial filters STILL constrain a full exchange on the topic.
This is the core issue:
A claimed function conveys structure when a POSITA knows how to build a machine to perform the function but it doesn’t convey structure when a POSITA doesn’t know how to build a machine to perform the function?
J. Moore admitted that there was code available on the internet at the time of the filing of the patent application that performed the claimed function. J. Moore’s opinion is a travesty of justice. The CAFC is using 112(f) as a means to invalidate claims that they feel are invalid under 103 but can’t put together a 103 argument. They are adding to their equity bag of tricks that already includes Alice.
Night Writer thinking the court is up to something when it’s been saying the same thing for at least 12 years. Here’s a full refutation of Night Wrtier’s position from Aristocrat in 2008:
Aristocrat’s principal contention is that the district court was wrong to hold that the patent’s disclosure of a general purpose, programmable microprocessor was not a sufficient disclosure of structure to satisfy section 112 paragraph 6. In particular, Aristocrat argues that computer-implemented means-plus-function claims do not require disclosure of a corresponding algorithm, as held by the district court. Instead, Aristocrat contends that the structure disclosed in the specification of the 102 patent, which was simply “any standard microprocessor base [sic] gaming machine [with] appropriate programming,” was a sufficient disclosure of structure under this court’s precedents.
In cases involving a computer-implemented invention in which the inventor has invoked means-plus-function claiming, this court has consistently required that the structure disclosed in the specification be more than simply a general purpose computer or microprocessor. The point of the requirement that the patentee disclose particular structure in the specification and that the scope of the patent claims be limited to that structure and its equivalents is to avoid pure functional claiming. As this court explained in Medical Instrumentation Diagnostics Corp. v. Elekta AB, 344 F.3d 1205, 1211 (Fed. Cir. 2003), “If the specification is not clear as to the structure that the patentee intends to correspond to the claimed function, then the patentee has not paid the price but is attempting to claim in functional terms unbounded by any reference to structure in the specification.” See also Biomedino, LLC v. Waters Techs. Corp., 490 F.3d 946, 948 (Fed. Cir. 2007) (“[I]n return for generic claiming ability, the applicant must indicate in the specification what structure constitutes the means.”). For a patentee to claim a means for performing a particular function and then to disclose only a general purpose computer as the structure designed to perform that function amounts to pure functional claiming. Because general purpose computers can be programmed to perform very different tasks in very different ways, simply disclosing a computer as the structure designated to perform a particular function does not limit the scope of the claim to “the corresponding structure, material, or acts” that perform the function, as required by section 112 paragraph 6…
Aristocrat also argues that, even if there is no disclosure of an algorithm in the patent, the disclosure of a microprocessor with “appropriate programming” is a sufficient disclosure of structure for means-plus-function purposes, because the evidence showed that one of ordinary skill in the art could build the device claimed in the 102 patent based on the disclosure in the specification. That argument, however, conflates the requirement of enablement under section 112 paragraph 1 and the requirement to disclose the structure that performs the claimed function under section 112 paragraph 6.
Although the examples given in the 102 patent might enable one of ordinary skill to make and use the invention, they do not recite the particular structure that performs the function and to which the means-plus-function claim is necessarily limited.
Whether the disclosure would enable one of ordinary skill in the art to make and use the invention is not at issue here. Instead, the pertinent question in this case is whether Aristocrat’s patent discloses structure that is used to perform the claimed function. Enablement of a device requires only the disclosure of sufficient information so that a person of ordinary skill in the art could make and use the device. A section 112 paragraph 6 disclosure, however, serves the very different purpose of limiting the scope of the claim to the particular structure disclosed, together with equivalents. The difference between the two is made clear by an exchange at oral argument. In response to a question from the court, Aristocrat’s counsel contended that, in light of the breadth of the disclosure in the specification, any microprocessor, regardless of how it was programmed, would infringe claim 1 if it performed the claimed functions recited in the means-plus-function limitations of that claim. That response reveals that Aristocrat is in essence arguing for pure functional claiming as long as the function is performed by a general purpose computer. This court’s cases flatly reject that position.
“when it’s been saying the same thing for at least 12 years”
This is a bit of a fallacy, with the insinuation that a span of time somehow provided veracity to what is being attempted.
This shows a lack of appreciation and a lack of understanding of history, and (again) reflects the selective bias that Random has, as he only gloms onto those things that affirm his preconceived viewpoint.
User identification module is so well-known in the art. It’s a great burden to the invetor to disclose every piece of information in the disclosure. The essence of the law should not be encourage the disclosure of well-know things, but should encourage new inventions.
How can “a user identification module configured to control access of said one or more software application packages” have no structure when read in view of a POSITA?
The court even acknowledges that there is code available for performing these functions available on the internet. I think the spec and claims were poorly written but that doesn’t mean the court can invalidate the claims.
The fact is that someone skilled in the art of software development in security and access would understand exactly what that module need to have in it. Yes, there are billions of solutions as there are with any modern technology even biotech.
Had I reviewed the specification, I would have said it has to have a description of that module with some more details on how it would operate. But this should not be taken as a 112(f). That is the core of the problem with this decision.
What claim element that performs the same function would suffice for it not to invoke 112(f)?
How can “a user identification module configured to control access of said one or more software application packages” have no structure when read in view of a POSITA? The court even acknowledges that there is code available for performing these functions available on the internet.
A function does not connote a structure. That’s like saying a machine for printing at a distance connotes a structure because we know about fax machines, sms phone texts and network tv. Knowing that the function can be performed still doesn’t connote any particular structure.
The fact is that someone skilled in the art of software development in security and access would understand exactly what that module need to have in it.
Well that’s not true, but that’s not the standard anyway. Are you suggesting there’s only one way to perform an authentication check? Even at the most basic level everyone would tell you that there are at least three factors of authentication – what you know, who you are and what you have (knowledge, inherency and possession) – not to mention each of those checks could be combined and carried out in hundreds of fashions.
Yes, there are billions of solutions as there are with any modern technology even biotech.
But this IS the problem.
What claim element that performs the same function would suffice for it not to invoke 112(f)?
“wherein the access is controlled by the steps of: receiving an input of a username and password, searching a database of usernames for an authentication answer, comparing the password to the authentication answer, and granting access in the event that the password and authentication answer match.” The fact that there are “billions” of other ways to control access to a software application is precisely the problem.
“A function does not connote a structure”
You mean like “screwdriver” or the THOUSANDS upon thousands of other examples that do exactly opposite of what you assert?
“A function does not connote a structure.”
To a POSITA it does. In fact you could not say anything more wrong in patent law than what you just said.
I have quoted from texts used at MIT in engineering that expressly say that functional language is used to represent known solutions and that functional language is used because there are so many solutions.
Random you have absolutely no idea what you are talking about. None.
The function is a place holder for all the known solutions that include what is enabled by the specification. That is the point of using that language and –read this 1,000 times– it is the only possible way to write a claim in modern technology that protects the invention.
I would love for you to speak to one of my inventors. He has a Ph.D. in physics and invented a consumer product that was licensed for millions and millions of dollars. He works for a large electronics firm. I wrote the patent application and the claims that were licensed for many millions of dollars.
Your ridiculous ideas Random would limit inventions to narrow solutions and not what an POSITA can easily build and understand. Yours is judicial activism by anti-patent judges like Taranto.
I don’t even know how to have a discussion with someone so ignorant as you are in regards to actual inventions and modern claim drafting.
To a POSITA it does. In fact you could not say anything more wrong in patent law than what you just said.
I have quoted from texts used at MIT in engineering that expressly say that functional language is used to represent known solutions and that functional language is used because there are so many solutions.
A solution and a structure are not the same thing. If you have a growing malady in someone’s foot you could treat the malady or amputate the foot and those are both “solutions” to the disease but they’re definitely not the same or even equivalent structural acts. But you elevate the butcher to the same level as the biochemist if he has the right kind of lawyer. I simply recognize that the law is not so blind, and unsurprisingly decades of precedent agrees with me.
Random you have absolutely no idea what you are talking about. None.
Well one of us consistently writes long screeds about how every written description and means-plus decision the CAFC makes is absolutely wrong, and it’s not me. This decision references a decision (Williamson) which references a decision (Ranpak) which all have a module invoking 112f and then invalidating the claims for lack of a disclosed structure. Were all of the judges involved (including the en banc of Williamson) flouting the law? Do all three of these patentees all have incompetent lawyers? Are you the only competent lawyer out there? And if so, can you point to me to one of your cases where you argued your 112f position to a court (rather than bloviating on the internet) and won? Why isn’t your superior logic prevailing? Is…is it the Google judges?
The function is a place holder for all the known solutions that include what is enabled by the specification. That is the point of using that language and –read this 1,000 times– it is the only possible way to write a claim in modern technology that protects the invention.
“the invention” is not “anything that the sum total of the art may or may not know at the time I file that would achieve this result.” “The invention” is what this particular inventor is positing for how to achieve something.
I would love for you to speak to one of my inventors. He has a Ph.D. in physics and invented a consumer product that was licensed for millions and millions of dollars. He works for a large electronics firm. I wrote the patent application and the claims that were licensed for many millions of dollars.
Wow, that’s cool. So, does he know his attorney can’t grasp basic disclosure requirements?
Your ridiculous ideas Random would limit inventions to narrow solutions and not what an POSITA can easily build and understand.
Hardly. This guy could have explained three factor authentication, and then there’d be a question as to whether what the defendant did constituted an equivalent. And if, as you claim, all authentication is really the same thing then this suit would have prevailed. So really your argument isn’t that the scope couldn’t be claimed, it’s that the attorney was too lazy to write a handful of paragraphs about how authentication is performed. So even if we take your dubious prospect that all authentication is the same and all solutions are known as true, that STILL doesn’t lead to the result you’re complaining about.
Perhaps I don’t know anything (just my expertise in this area but okay) but MY opinion is that in 2007 lay people could have told you that transferring data over the internet was ubiquitous, subscribing to remote access licenses or download update schemes to software programs was known, and therefore the features surrounding how the system determines that THIS person is who THIS data should be sent to (i.e. “controlling access”) is the only plausibly inventive part of the claim.
If you bothered to read the image file wrapper of the application, you’d see that the claim was anticipated but-for exactly those features relating to the user identification module, that the secondary reference and motivation for combining that module was attacked on appeal, and that that argument carried the day. So to be clear – the only reason this claim issued is because of the novelty of the thing you say everyone in the art could have made and would have appreciated the value of using beforehand. If only the PTAB could have been as wise as you. Or if the patentee’s position today – see pg. 8, citing the patentee’s argument that this could be done “with common software algorithms” and THE INVENTOR saying open source software could implement it – had been presented to the Board at the time. Isn’t it convenient for Rain how nonobvious the feature was when there’s a 103 rejection pending at the office and they managed to discover how completely known it was once there’s a 112 validity attack post-issuance?
It’s unsurprising the examination was hamstrung when it’s unclear what acts are performed by the “user identification module controlling access” at the client and what acts are performed by the “a server device authenticating the user” at the server. The way I would have interpreted that is that any field presented by the client asking a client for a username before unlocking access constitutes a user identification module, and any comparison steps taken by a server with that information constitute authentication, which would have let one cite what I can only assume would have been thousands of prior art applications. The examiner took a more limited view, and ultimately failed on a motivation to combine attack (lol) because of it. I agree with you that there are copious examples of authentication prior to this invention (I just recognize it’s not legally relevant to this issue). Given that, I’m sure that the success of the argument that steps (b)-(d) provided the claim with nonobviousness had absolutely nothing to do with the indefiniteness surrounding the functional language in steps (b)-(d), right?
>>If you bothered to read the image file wrapper of the application, you’d see that the claim was anticipated but-for exactly those features relating to the user identification module, that the secondary reference and motivation for combining that module was attacked on appeal,
That isn’t the issue. It is the combination of elements and what was done with the user identification module in that it was sent.
It may be invalid under 103, but a POSITA knows exactly how to write the user identification module as we know as there was source code available for free on the internet to do exactly that.
You are loopy and ridiculous. You have been blinded by the e vils of the CAFC making things up that have no correspondence to modern technology.
The fact is that claiming there is no function conveyed by a user identification module when J. Moore admits that there is code available on the internet to perform exactly that function is gross misconduct. Of course it conveys function.
What these vile judges have done is used 112(f) as another 101 to invalidate claims that they think are invalid but they can’t make the 103 arguments. This opinion is an absolute disgrace for the reasons I have stated.
And arguing with you is ridiculous as you don’t address the points I make but just keep repeating yourself with these “all” and “every” arguments that are ridiculous.
Do you get that a claimed function conveys structure when a POSITA knows how to build a machine to perform the function but it doesn’t convey structure when a POSITA doesn’t know how to build a machine to perform the function?
It is the combination of elements and what was done with the user identification module in that it was sent.
So just to be clear, your position is the entire rest of the claim was known, the module was known, but the module was transmitted to a client device, and that renders the claim non-obvious?
Question – how would the structure of the module need to change in order to accommodate that the module was differently located? Where in the specification is that changed structure? And if it isn’t in the specification, how did the art know of the structure when it wasn’t even obvious to the art to move the module?
Your position is ridiculous but even if it was accepted as true it’d still fail to carry the day. You’re trying to argue that one of skill knew what the modified structure would be when it was non-obvious before and there was no additional teaching.
It may be invalid under 103, but a POSITA knows exactly how to write the user identification module as we know as there was source code available for free on the internet to do exactly that.
That’s an enablement answer. This isn’t an enablement invalidation.
The fact is that claiming there is no function conveyed by a user identification module when J. Moore admits that there is code available on the internet to perform exactly that function is gross misconduct. Of course it conveys function.
I think you mean “structure” not function. And again, the fact that I know things can perform a function does not convey structure. I know the function of a device for travelling exists, that doesn’t convey either the function of a car, or a motorcycle, or a boat, or an airplane. I know all those things exist, and I know they all do that function, but the question is which structure is in the spec? None. Which structure outside the spec (which isn’t even enough statutorily) is being referred to? We don’t know. The statute requires the scope to be a structure in the spec.
Do you get that a claimed function conveys structure when a POSITA knows how to build a machine to perform the function but it doesn’t convey structure when a POSITA doesn’t know how to build a machine to perform the function?
haha no it does not. That’s the exact issue in this case, but ignoring that, logically it does not. Again, if I tell you to make me a machine for travel, do you know definitively (i.e. with reasonable certainty) whether I am talking about the structure of a car or the structure of a boat or the structure of a plane?
You need to learn to be more concise.
link to patentlyo.com
…the shards of Random’s glass house – step carefully.
“I have quoted from texts used at MIT in engineering that expressly say…”
Worth adding here, since Greg “I Use My Real Name” DeLassus is on his annual Lenten abstinence from posting, is that Night Writer’s technical position was met with derision from Greg, with Greg asserting that patent claims are NOT a technical thing, but a legal thing.
I answered Greg’s assertion by sharing that the parallel to Night Writer’s technical basis is well understood on the legal side by giving examples of the Ladders of Abstraction concept for the legal patent nature (including the leading patent claim writing guide of Slusky’s:
Invention Analysis and Claiming: A Patent Lawyer’s Guide
Book by Ronald D. Slusky
but Greg refused to respond.
And just in case a sane person is reading this…
The key to understanding this Random is ask yourself why isn’t the inventor entitled to all the embodiments? Why would an embodiment that is obvious to a POSITA of implementing a function in the claims included in the scope of the claims?
If it isn’t included in the claims that means that you have no protection as a design around will be easy. Your desire to limit people to a few concrete embodiments is mor on ic. It provides no protection. It does not reflect reality. It is offensive to every ounce of my intellectual integrity.
And—guess what?–you can’t defend it. You just yap out nonsense about “all this” and “all that” and how would they know which embodiment to implement. Only a mentally disabled person or a judicial activist would say such things.
Why would an embodiment that is obvious to a POSITA of implementing a function in the claims included in the scope of the claims?
That’s not what you’re advocating. You advocate inventing one thing, describing what the parts of thing do by function, and then getting a scope that includes ALL of that function. I could invent a digital driving route planner that utilizes a terrible algorithm for planning the route, claim a processor that plans driving routes (since that violates neither 112a nor 112b for you), and receive coverage for ALL possible algorithms that plan routes. Even good ones. Even non-obvious ones. Even ones not even enabled at the time. According to you, I simply have to think that “other algorithms out there might determine routes better” and that constitutes an enabled, possessed, definite scope.
The burden on the public is clear – You only came up with a bad algorithm, but now there is no incentive for anyone to disclose a better algorithm. They can’t monetize their superior invention no matter how different it is from yours because it achieves the same general end-goal. That’s not a claim on the invention, that’s a claim on the function or utility itself.
It is offensive to every ounce of my intellectual integrity.
I can’t decide between a teaspoon joke or a dripping faucet joke. Fortunately you’ll let me claim all of them. Why don’t we have everyone here take a shot on this, and then whichever one is the most witty is the one I came up with and I’ll sue the person who said it?
And—guess what?–you can’t defend it. You just yap out nonsense about “all this” and “all that” and how would they know which embodiment to implement. Only a mentally disabled person or a judicial activist would say such things.
Well my opinion is largely irrelevant but I can point you to some Supreme Court statements on the damage that overbroad claims do, if you’d like.
Random–No.
You don’t understand modern technology. The problem you are talking about is the reverse doctrine of equivalents, which the CAFC has done their best to bury so as to destroy claim scope.
You statements just don’t have any correspondence to a real invention. You are talking to someone that works with real inventors and has to write claims that will stand up in court and cannot be easily designed around.
If you just go through a few real-world examples, then you would get it.
And learn what structure is to a modern technology person.
I’ve quote from engineering books at MIT that tell us what structure means to a modern technology person. It means all the solutions that a POSITA knows even if it is billions of solutions.
That is how modern innovators think.
As if it matters when the feature is the alleged novelty. “Car” is a sufficiently structural term, but there’s a good argument for why “a flying car” or “a car for flying” would both invoke means plus, and it wouldn’t matter because if there’s no disclosure as to how to achieve the flying-ness of the car the claim is going to be invalid anyway.
Maybe this might be a good time to review Prof. Crouch’s (coined) term of Vast Middle Ground.
Is your position that either “a flying car” or “a car for flying” would be valid terms even without detailed specification support, or are you just spouting meaningless platitudes?
My position is that it is appropriate to understand what “Vast Middle Ground” means — especially as the emotions get riled about 35 USC 112(f).
After all these years, how can you even ask that question Random?
Good question Marty (but likely, not for the reason that you may be thinking of).
I’m not really asking that question Martin. My point being that if you don’t describe what makes the car fly, it doesn’t really matter whether you’re in means plus or not.
You ask a question (but not really), because you think that you are making a point — but the question — and answer — show that you do not GET the point at hand.
“After all these years, how can you”
more shards of glass houses… coming from the guy who refuses to understand the legal terrain and for years continues to float his asinine ‘theories.’
Pity the patentee didn’t claim it as user identification circuitry. That would’ve fixed everything.
The circuitry decisions are all based on pre-Williamson caselaw about a strong presumption agaisnt 112f and whatnot. I dont think they are good law anymore (though it hasnt been explicitly stated).
The bigger issue is “processor.” It cannot be that “processor” is *not* sufficient structure when recited in the spec ( for performing a specialized function once 112f is invoked), but *is* sufficient structure when recited in the claim. If, as the CAFC has held, a processor is not sufficient structure for specialized functions once 112f is invoked, then “processor configured to XYZ” in the claims must also invoke 112f.
“It cannot be that “processor” is *not* sufficient structure when recited in the spec ( for performing a specialized function once 112f is invoked), but *is* sufficient structure when recited in the claim.”
I’m confident that the CAFC will be able to make up a reason to reach their desired result.
“I’m confident that the CAFC will be able to make up a reason to reach their desired result.”
Yes, but this only leads to the type of error known as “the score board is broken.”
Sadly, there are many who refuse to grasp that THIS is a problem.
then “processor configured to XYZ” in the claims must also invoke 112f.
You’ve hit upon the problem of computerized functional claiming. When you go from the Supreme Court’s admonition that limited use of functional language may be proper in some circumstances to functional language being used for most of the limitations all the time in computer language, it’s not surprising it screws with the analysis.
…because Supreme Court admonitions are not themselves “screwy enough”…
(see post 5.1.1.1)
Method claims are inherently functional in nature. An applicant can’t stray into mixing method and structure or they run into issues with mixing statutory classes in the claim. You can’t limit the use of functional elements in a method claim otherwise there is no point in using the method format.
Acceptable to Random:
A method comprising receiving a packet from a server. (I can examine this!)
Unacceptable to Random:
A system comprising a module configured to receive a packet from a server. (I can’t examine this because I like to pretend like I don’t know what you’re talking about)
For what it’s worth, there is a difference between the two, as the first claim requires that a packet be received, while the second claim does not.
The equivalent method claim might look like: providing a module configured to receive a packet from a server. I am pretty sure he would not be crazy about this one either.
I’ll make it unmistakable:
A system comprising a module that receives a packet from a server.
This language is still allegedly “confusing” to the average examiner who wants to abuse 112(f).
Of course, this approach risks making “it [] unclear whether infringement of [the] claim [] occurs when one creates a system that allows the … [module to receive a packet from a server], or whether infringement occurs when the [module actually receives a packet from the server],” and thus “does not apprise a person of ordinary skill in the art of its scope, and [] is [accordingly] invalid under section 112, paragraph 2.” IPXL Holdings, L.L.C. v. Amazon.com, Inc., 430 F.3d 1377, 1384 (Fed. Cir. 2005).
At this point I am mostly just being obnoxious, as it would probably be fine under current jurisprudence. See, e.g., Mastermine Software, Inc. v. Microsoft Corp., 874 F.3d 1307, 1315-1316 (Fed. Cir. 2017) (“Though claim 8 includes active verbs—presents, receives, and generates—these verbs represent permissible functional language used to describe capabilities of the ‘reporting module.’ … [T]he claims at issue here merely claim that the system “possess[es] the recited structure [which is] capable of performing the recited functions.” (quoting Microprocessor Enhancement Corp. v. Tex. Instruments Inc., 520 F.3d 1367, 1375 (Fed. Cir. 2008)).
On the broader issue, the Fed. Cir. has made clear that “[f]unctional language may [] be employed to limit the
claims without using the means-plus-function format,” Microprocessor Enhancement, 520 F.3d at 1375.
That being said, Williamson appears to make clear that “§ 112, para. 6 will apply if … the claim term … recites ‘function without reciting sufficient structure for performing that function.'” Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1348 (Fed. Cir. 2015) (en banc section of opinion) (quoting Watts v. XL Sys., Inc., 232 F.3d 877, 880 (Fed.Cir.2000)).
It seems like there is a strong argument that “a module configured to receive a packet from a server” “recites ‘function without reciting sufficient structure for performing that function’,” Id., and thus would be construed under § 112, para. 6.
I also think it is always helpful to keep in mind during prosecution that the examiner’s opinion is not the one that will ultimately matter. Convincing an examiner that a functional claim limitation should not be construed under § 112, para. 6 generally just makes his job easier, as it gives him a broader claim to try to search for. Ultimately, though, a court or the Board will make up its own mind on whether a claim invokes § 112, para. 6.
A method comprising receiving a packet from a server. (I can examine this!)
This is suggested by all structures that perform the function of receiving a packet from a server.
A system comprising a module that receives a packet from a server.
This is drawn to a smaller scope – only CERTAIN structure that perform the function of receiving a packet from a server are included. How do we know which structures are those certain structures? We start by determining what structure in the specification performs this function (thats explicitly what the statute says to do). If the specification has no structure, then we don’t know which structures are the certain structures being claimed.
It has nothing to do with “confusing” – the law tells me to start at point A, and there is no point A, so I can’t do what the law commands. The law says when you can’t do what is commanded (in this instance) it is a 112b violation. It’s not much different from an antecedent basis problem – if you refer to “the” widget before you refer to “a” widget, I likely understand you intended to introduce the widget there, but that doesn’t change the fact you’ve violated 112b.
One of the aspects of the Act of 1952 was to amplify the very reach of method claims.
See 35 USC 100(b).
Note that the change is NOT a tautology (as those who refuse to understand the terrain insist), but rather, an amplification with incorporation of a term of art as but one component of the larger permitted area.
Combine this with those who refuse to understand that the US Sovereign has chosen NOT a “technological arts” test, but instead have chosen the broader Useful Arts, and one can easily see how those who are anti-patent align with the issues here.
When the functional result is an item of information, the spec should have to disclose enough for PHOSITA to generate the information or it’s functional equivalent. If you claim a means or module to provide a certain arithmetic, the spec should not have to cover the theory of numbers or how to make an arithmetic circuit, because those are widely known in the art. If your information result is a rare species, then more disclosure would be needed. The means plus function test is really just a nested 103 inquiry, isn’t it?
Result…?
What is the ‘result’ of an automated traffic light?
In theory, 112(f) can only save claims. If a claim that includes “module for function X” does not invoke 112(f), then it covers *ALL* structures of accomplishing that function X. This requires *heightened* written description. If there isnt even one sufficient disclosure of structure in the spec, then there certainly isnt enough to disclosure to satisfy written description (or enablement) for all possibles tructures.
That is, anything that fails because it invokes 112(f) must also fail if it doesnt invoke 112(f), because the scope without invoking 112(f) is larger than the scope invoking 112(f) and therefore requires more disclosure.
“In theory…”
And in practice, written description for the predictable arts is a dead letter.
… “the predictable arts”….
So do you include how to add numbers on a computer in your specification?
No, because a processor is structure known to POSITA to perform the function of adding.
Do you think a “module” is a structure known to perform the function of “control[ling] access of said one or more software application packages?” (Maybe it is, and you can show me its a standard function for whatever a “module” is, but that would also hurt the 103 argument).
Note: The question isnt whether POSITA could make a “module” do that. Its whether a “module” is structure already known by POSITA to do that function. That’s what you seem to be getting confused.
… where is my link to the Grand Hall experiment….
In theory, 112(f) can only save claims. If a claim that includes “module for function X” does not invoke 112(f), then it covers *ALL* structures of accomplishing that function X. This requires *heightened* written description. If there isnt even one sufficient disclosure of structure in the spec, then there certainly isnt enough to disclosure to satisfy written description (or enablement) for all possibles tructures.
That’s not true at all. It’s true that means-plus decreases the scope, but written description of general elements can be met by the common knowledge in the art (through judicial fiat) while written description of means plus elements must be met by the four corners of the disclosure (through statutory requirement). It shrinks what you have to disclose, but also shrinks the manners in which it discloses it.
That is, anything that fails because it invokes 112(f) must also fail if it doesnt invoke 112(f), because the scope without invoking 112(f) is larger than the scope invoking 112(f) and therefore requires more disclosure.
A specification that simply mentions a writing utensil supports a claim to a writing utensil but does not support means for writing. It would have to mention a pen before it supports means for writing. That’s because the art knows of a wide range of structures for writing, but absent a nomination to a particular one you can’t determine what the equivalents are for a 112f analysis.
RandomGuy:
The test for whether a claim invokes 112(f) is whether the claim recites sufficient structure for performing the claimed function.
The test for whether, once invoked, a spec is sufficient to support a 112(f) claim is whether the specification recites sufficient structure for performing the claimed function.
We agree on both of these tests, right?
Note they are the same test. If a claim has “writing utensil for writing”, then you only avoid 112(f) by arguing that “writing utensil” denotes sufficient structure to POSITA for performing the function of writing. If a claim has “means for writing” and the spec recites a “writing utensil,” then you save the claim by arguing that “writing utensil” denotes sufficient structure to POSITA for performing the function of writing.
Not sure why you think a class of structures would serve in the claim but not in the spec. The test is the same. And just note there is the same “equivalents” analysis for non-112(f) claims under DOE– so using an “equivalents” argument to say you need more specificity for 112(f) claims doesnt really make sense.
The test for whether a claim invokes 112(f) is whether the claim recites sufficient structure for performing the claimed function.
Not exactly. Whether means plus is invoked is a matter of intent. The language used is evidence of the intent. It is possible to refer to a class of items by the function they produce and not their structure and still be outside of means plus. As anon will tell you, 112f is an OPTIONAL manner of claiming. (I expect a /golfclap for that, anon)
The test for whether, once invoked, a spec is sufficient to support a 112(f) claim is whether the specification recites sufficient structure for performing the claimed function.
Yes
If a claim has “means for writing” and the spec recites a “writing utensil,” then you save the claim by arguing that “writing utensil” denotes sufficient structure to POSITA for performing the function of writing.
Except a writing utensil invokes no structure at all, it names a function. It can’t possibly fulfill 112f. But a claim to the element of writing utensil may meet both 112a and b if the art knows of structures that perform functions. Moreover, even if you name structureS you could cause a 112b problem because its still not nominating a structure. i.e. A spec saying a writing utensil could be a pen, pencil, or a tattoo rotary pen is just as indefinite as a spec that doesn’t define the writing utensil at all, because it just nominated 3 non-equivalent structures as the thing that performs the function.
Not sure why you think a class of structures would serve in the claim but not in the spec.
Because one can express an element grouping in any manner they see fit, and grouping elements by function can be known by the art to exist without the specification detailing any one of them.
This reads like Newman’s dissent in Williamson, and/or the holding in Flo Healthcare that was abrogated by Williamson. 112(f) is optional in that the drafter could have included structure into the claim rather than function; by drafting the claim to not include structure sufficient to perform a recited function, the drafter has taken the option to invoke 112(f).
Williamson: “§ 112, para. 6 will apply if the challenger demonstrates that the claim […] recites function without reciting sufficient structure for performing that function.” ”
(Note that is almost verbatim what I said; “The test for whether a claim invokes 112(f) is whether the claim recites sufficient structure for performing the claimed function.”).
If, as you say, “writing utensil denotes no structure at all,” then a claim with “writing utensil for writing” must invoke 112(f) under the Williamson standard. (I disagree, and think writing utensil invokes a class of structures, but its certainly arguable).
I’d be interested if you point to any post-Williamson cases that make the distinction you are trying to make.
ksksksks,
Great comments in this thread; agree on at least 99%.
I think the issue that RandomGuy is trying to raise is whether, absent the existence of 112, p. 6 / 112(f), elements that are not at the point of novelty could theoretically be expressed in the claim using functional language and construed to cover elements known to an ordinary artisan even if not disclosed in the specification.
The belief of many attorneys seems to be that this would be proper.*
Potentially supporting this belief, even in General Elec. Co. v. Wabash Appliance Corp., 304 U.S. 364 (1938), the Court noted that “[a] limited use of terms of effect or result, which accurately define the essential qualities of a product to one skilled in the art, may in some instances be permissible and even desirable, but a characteristic essential to novelty may not be distinguished from the old art solely by its tendency to remedy the problems in the art met by the patent.” General Elec., 304 U. S. at 372.
On the other hand, in Merrill v. Yeomans, 94 U.S. 568 (1876), the Court indicated that “in cases where the invention is a new combination of old devices, [the patent applicant] is bound to describe with particularity all these old devices and then the new mode of combining them for which he desires a patent[, and] [i]t thus occurs that in every application for a patent, the descriptive part is necessarily largely occupied with what is not new in order to [provide] an understanding of what is new.” Merrill, 94 U.S. at 570.
To the extent that prior to enactment of 112, p. 3 (now 112, p. 6 / 112(f)) in response to the General Elec. / Halliburton line of cases, functional claim language not at the point of novelty was acceptable and could cover elements known to the ordinary artisan but not explicitly disclosed, the plain text of 112, p. 3 (now 112, p. 6 / 112(f)) would potentially seem to eliminate this ability. However, it seems unlikely that Congress intended this result, given that Congress was intending to expand the ability of patentees to use functional claim language.
This issue is potentially relevant even with the existence of 112, p. 6 / 112(f) because it potentially raises an argument regarding when the saving exception should be triggered/applied.
Overall – really, really complicated issues that are far from settled (and which also have a ton of interplay with the ongoing eligibility mess), but I think RandomGuy raises an interesting question.
* (Some attorneys also seem to believe that functional claim language at the point of novelty is fine too and want to argue against application of the saving construction of 112, p. 6 / 112(f) without appreciating that it may just doom their claim.)
I posted this before seeing the most recent comments by you and RandomGuy. Re my last comment as to “an argument regarding when the saving exception should be triggered/applied”, I think Newman’s dissent in Williamson is one possible form of that argument.
The argument jumping off from Newman’s dissent would be that 112, p. 6 / 112(f) gives an applicant an opportunity to use the term “means for” or “step for” to invoke these saving constructions, but absent that functional language invoking the saving construction the functional language should be interpreted under the traditional case law, sink or swim.
I agree with most of what is said here. I am a big fan of the usefulness of Wabash and Perkin’s Glue in understanding the original functional language complaints.
I also agree it has a lot to do with eligibility, as today what is poorly answered by the written description test was more ably previously policed through a combination of indefiniteness and eligibility. A claim to a machine that was described only by functional result was indefinite, while a claim to the functional result itself was ineligible, and consequently one had to claim the structure of the machine. The WD test is a poor man’s backdoor for achieving the same result.
To reach the particular 112f issue that ks refers to, I agree that Williamson performs a bright line rule, but that ignores that there was 60 years before Williamson where elements were construed outside of means plus, which makes the categorical language in Williamson inappropriate. The federal circuit often gets itself in trouble with bright line rules that get overruled by the supreme court, see, e.g. Alice overruling “it has a physical thing so it must be statutory” or KSR with “there must be an express teaching suggestion or motivation” or Nautilus with “indefiniteness is only when it is insoluably ambiguous.”
I disagree with the suggestion that the rule is not based on what the applicant’s intent is. Williamson does make the presence or absence of the word “means” IRRELEVANT, which it would have to if the test was only whether the art would understand the term to be connoting structure. Williamson simply did away with the strong presumption introduced in 2004, not suggesting that it was beyond the inventor’s capacity to formulate an element bounded only by function. I agree the ultimate patentability of a claim including such an element (especially if it is directly connected to the novelty/nonobviousness of the claim) is dubious.
“understanding the original functional language complaints.”
It’s more than too bad though that you stop (driven by your preconceptions and interest in “the problem”) and forget to ALSO look at what Congress did to that ‘so-called’ problem.
As you do not – and you seem ONLY interested in perpetuating your own warped view, the actual law escapes you.
This is reflected in your celebrating of broken score boards, as you seem absolutely incapable of even recognizing what a broken score board is, or why such is the bigger problem.
Are you referring to a “shield only” versus a “shield and sword” (as as some would desire, a “sword only”)…?
That is, shield is meant to protect and enhance the ability of an applicant to obtain innovation protection, while a sword is used against an applicant to deny protection…
sending, to the user, a user identification module configured to control access of said one or more software application packages
To 99.9% of the patent attorneys practicing in the computer arts, the meaning of this language is plain as day. It simply describes the sending of a computer program (i.e., the user identification module) to a user. The computer program (i.e., the user identification module) itself is technology that has been known for untold decades.
Normally, patent claims are expected to spell out the actual structure of the invention — not just its function.
This is nonsense. There is no requirement that structure be claimed. All that is required under 35 USC 112 is that the “claims particular point out and distinctly claim[] the subject matter which the inventor … regards as the invention.” Physical items (i.e., not just things in the computer arts) are described using functional language all the time. What are the terms: light, radiator, switch, distributor, computer, remote control, container, lawn mover, hole punch, connector? They are description of physical things by what they do. In other words, they are words that define things by their function — not their structure. Are they “means plus function” language? What if one my claims recites a “switch” and my specification doesn’t provide sufficient structure — is the claim now indefinite?
The Federal Circuit concluded their opinion with the following statement:
Without an algorithm to achieve the “control access” function, we hold the term “user identification module” lacks sufficient structure and renders the claims indefinite.
Under the Supreme Court’s decision of Nautilus, “[a] patent is invalid for indefiniteness if its claims, read in light of the patent’s specification and prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention.” The Federal Circuit’s decision held that claims were indefinite yet contained no mention of Nautilus or the standard set forth therein. The reason being is quite clear — the Federal Circuit’s position is irreconcilable with Nautilus.
Those skilled in the art have exceedingly little uncertainty as to what constitutes “a user identification module configured to control access of said one or more software application packages.” Again, this is exceptionally old technology is this art. As such, this language easily passes Nautilus.
The claim went to the PTAB and the Examiner was reversed (apparently, the Board understood the language well enough), yet all the time and money the inventor spent to get this patent has been wasted because of the inclusion of a trivial (probably wasn’t necessary to overcome the prior art) and easily understood limitation.
This decision is just another in a long line of decisions by the Federal Circuit intended to make it more difficult to get patents directed to inventions (and particularly those involving computer software). When you read Federal Circuit and CCPA decisions from decades ago, the case law was written to avoid “traps for the unwary.” However, at least over the past decade the Federal Circuit has excelled in creating creating these traps. Where the patent system once catered to smaller inventors and wouldn’t invalidate a patent over minor issues, the current patent system excels in making it the “sport of kings” whereby any tiny flaw can doom a patent.
The result of this decision is that anytime anybody uses the term “module” or anything comparable (whatever that might be), patent attorneys will be forced to include an “algorithm,” which is totally unnecessary to make/use the claimed invention. However, this will drive up the cost of preparing the patent application and make it harder for the little guys.
There are about 350K issued US Patents that include the term “module” in one of the claims. How many of those would pass muster under this new decision? As it turns out, Samsung (as the assignee) has almost 8000 issued patents that include the term “module” in one of the terms. For example, USP 10,922,011 recites “a secure deletion control module configured to determine the at least one block to be erased.” I hope, for their sake, there is an algorithm for that module in the specification.
Below is something that is profound. The CAFC isn’t even trying to create a body of law so that patent attorneys know what they need to do to write a valid patent. The CAFC needs to be disbanded. I heard rumors of rumors that Trump was onboard for disbanding the CAFC. Maybe if the Ds get out of office, which may be never with their planned 30 million new Hispanic voters that vote 80 percent D. But if an R is elected again maybe they will disband the CAFC. J. Rich would be writing dissents to each one of these opinions calling his fellow judges purchased clowns. I guess what the CAFC is doing amounts to at least gross negligence. Add intent and I think it is much worse.
Wandering through >>>This decision is just another in a long line of decisions by the Federal Circuit intended to make it more difficult to get patents directed to inventions (and particularly those involving computer software). When you read Federal Circuit and CCPA decisions from decades ago, the case law was written to avoid “traps for the unwary.” However, at least over the past decade the Federal Circuit has excelled in creating creating these traps. Where the patent system once catered to smaller inventors and wouldn’t invalidate a patent over minor issues, the current patent system excels in making it the “sport of kings” whereby any tiny flaw can doom a patent.
Your problem is that I can buy a switch, a lawn mower, a hole puncher, a radiator, etc. from home depot, or I can look in a textbook or manual and find a structure that describes them. I cant buy or find any reference to a “module that controls access to XYZ.” (and if I could, then the lawyers didnt do a good job showing it).
The other issue arises when for 102/103, the “module” is the novelty, but then for 112 its something with definite, known structure. It cant be both. (not sure if that is the case here or not, but this issue arises a lot, where an element is argued both as being novel *and* as being a structure recognized in the art; if its recognized in the art, it cant be novel. the combination can be novel, but not the element itself).
I cant buy or find any reference to a “module that controls access to XYZ.” (and if I could, then the lawyers didnt do a good job showing it)
Let me repeat, under the Supreme Court’s decision of Nautilus, “[a] patent is invalid for indefiniteness if its claims, read in light of the patent’s specification and prosecution history, fail to inform, with reasonable certainty, those skilled in the art about the scope of the invention.”
I see nothing about a requirement that the item has to be capable of being purchased at Home Depot or described in some manual. The issue is whether those SKILLED IN THE ART have reasonable certainty as to the scope of the invention. In this instance, is the scope of the term “a user identification module configured to control access of said one or more software application packages” reasonably certain? Would one skilled in the art be able to identify one if he/she encountered it?
The other issue arises when for 102/103, the “module” is the novelty
Read the entire claim, this module was not the novelty. It was an ancillary element that could have easily been removed from the claim.
The other issue arises when for 102/103, the “module” is the novelty, but then for 112 its something with definite, known structure. It cant be both. (not sure if that is the case here or not, but this issue arises a lot, where an element is argued both as being novel *and* as being a structure recognized in the art; if its recognized in the art, it cant be novel. the combination can be novel, but not the element itself)
You are confusing indefiniteness and enablement — it is a classic error that I frequently see at the Patent Office.
A claim defines the scope of the invention — the hypothetical fence defining the scope of what the inventors are claiming as their invention. To be indefinite means that the where the line is drawn by this fence is unclear (i.e., not reasonably certain). Enablement is another issue altogether — it is whether the specification allows one skilled in the art to make and use the invention. In the computer arts, the mere description of an element is sufficient to enable one skilled in the art to make and use the invention.
Getting caught in the filter….
Your problem is that I can buy a switch, a lawn mower, a hole puncher, a radiator, etc”
Ah, but this statement only invites the problematic view of Random of the (determinative) SCOPE of these things, as any (and every) one of these things that you are using a broad label for are indefinite in that label as they do not (and cannot) cover the full scope of those things, given that new improvement inventions still occur FOR those things.
To “conquer” the (asserted) indefiniteness issue (in a turtles all the way down manner), one needs an EXACTING picture claim – and only an EXACTING picture claim.
As Wandering reflects at 3.4.2 (and as Dave Stein previously shredded Malcolm several years back), the ‘conundrum’ of “functional language” is NOT limited to the computing arts, and the anti’s that take aim AT the computing arts can not limit the use of their rhetorical
w
e
a
p
o
n
s
to only the computing arts.
This also feeds into understanding just how VERY vast is that Vast Middle Ground.
Still getting caught in the filters…
Your problem is that I can buy a switch, a lawn mower, a h0 le puncher, a radiator, etc”
Ah, but this statement only invites the problematic view of Random of the (determinative) SCOPE of these things, as any (and every) one of these things that you are using a broad label for are indefinite in that label as they do not (and cannot) cover the full scope of those things, given that new improvement inventions still occur FOR those things.
To “conquer” the (asserted) indefiniteness issue (in a turtles all the way down manner), one needs an EXACTING picture claim – and only an EXACTING picture claim.
As Wandering reflects at 3.4.2 (and as Dave Stein previously
s
h
r
e
d
d
e
d
Malcolm several years back), the ‘conundrum’ of “functional language” is NOT limited to the computing arts, and the anti’s that take aim AT the computing arts can not limit the use of their rhetorical
w
e
a
p
o
n
s
to only the computing arts.
This also feeds into understanding just how VERY vast is that Vast Middle Ground.
… nice to see that filters are STILL being differently applied, as the item that drew the George Carlin filter was the “cut and paste” from someone else who did not have to deal with the filter.
“The Federal Circuit’s decision held that claims were indefinite yet contained no mention of Nautilus or the standard set forth therein. The reason being is quite clear — the Federal Circuit’s position is irreconcilable with Nautilus.”
You come across here as appearing to be deliberately arguing in bad faith through how your argument completely ignores the existence of 112(f). It is the 112(f) issue that makes the claims potentially indefinite, and pretending it doesn’t exist isn’t a good look.
It’s not like the CAFC hasn’t built outs for you guys to have your open ended functional software claims. And it’s not like these particular claims were granted before Williamson. Save your teeth to be gnashed for a better case.
You come across here as appearing to be deliberately arguing in bad faith through how your argument completely ignores the existence of 112(f). It is the 112(f) issue that makes the claims potentially indefinite, and pretending it doesn’t exist isn’t a good look.
Indefiniteness is a 112(b) issue — not a 112(f) issue. The text of 112(f) is the following:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
Is there anything about indefiniteness in that statutory language? No. 112(b) speaks to indefiniteness. Therefore, Nautilus is controlling case law.
If the Federal Circuit is going to tank a claim for being indefinite, they should apply the controlling case law.