Means Plus Function; Nonce Words; and En Banc Petitions

by Dennis Crouch

Back in March 2022, I wrote an essay on the spectrum of structural and functional claiming, particularly commenting on the Federal Circuit’s decisions in Dyfan, LLC v. Target Corp., 28 F.4th 1360 (Fed. Cir. 2022) and VDPP LLC v. Vizio, Inc., 2021-2040, 2022 WL 885771 (Fed. Cir. Mar. 25, 2022). See, Dennis Crouch, Discerning the Purpose and Means of Williamson v. Citrix, Patently-O (March 30, 2022).

A basic question in these cases is whether the claim elements at issue recite sufficient “structure” to avoid being classified as a “means” claim under 35 U.S.C. 112(f). Patentee’s typically prefer to avoid 112(f) classification as the result is typically either (1) narrow claim scope or (2) invalid claims.  The classification is seemingly easy when a patentee uses classic “means plus function” (MPF) claim language without reciting specific structure, but patentees usually avoid that approach in favor of quasi-structural elements such as a “module,” “circuit,” or “system.” Prior to Williamson, the Federal Circuit  strongly presumed claim elements lacking traditional MPF language were not MPF elements.  Williamson eliminated that  strong presumption in favor of a weak rebuttable presumption that can be overcome with a demonstration “that the claim term fails to recite sufficiently definite structure or else recites function without reciting sufficient structure for performing that function. Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc as to Part II.C.1).  That case went on to focus attention on “nonce words” such as “mechanism,” “element,” and “device” that “typically” fail to recite sufficiently definite structure.  The problem here is discerning in any particular case what level of structure is sufficient.  Although the court used the word “definite”, it seems unlikely that they intended to tie the 112(f) here to the same reasonable certainty test used for 112(b).

In the recent cases of Dyfan and VDPP, the Federal Circuit pushed-back a bit on an expansive interpretation of Williamson.  In Dyfan, the court considered a wherein clause requiring certain actions and concluded it was implicitly tied to claimed “code” and therefore recited sufficient structure to avoid MPF designation.  In VDPP, the court indicated that the nonce word designation should be based upon “evidence that a person of ordinary skill would not have understood the limitations to recite sufficiently definite structure.”  The district court in VDPP had not required submission of “evidence” but rather had summarily concluded that the “processor” and “storage” elements were nonce words lacking structure.  On appeal, the Federal Circuit reversed — holding that those words provided sufficiently definite structure.

Now, the accused infringers in both of these cases have petitioned for rehearing: calling for either (1) panel rehearing or (2) rehearing en banc.  The court has not yet requested responses in either case. (No response may be filed to a petition for panel or en banc rehearing unless the court orders a response).

In Dyfan, Target argues that the law was settled by Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc as to Part II.C.1) and confirmed in Egenera, Inc. v. Cisco Sys., 972 F.3d 1367 (Fed. Cir. 2020).  The petition asks the following questions:

  1. For claims that recite computer software for performing functions, what
    are the criteria for determining whether software-implemented functional claim
    language is subject to 112(6)?
  2. Can the Court ignore parts of the recited software-implemented function, or ignore such function entirely, in making the determination as to whether the claim recites sufficiently definite structure?
  3. If no, and if the software-implemented function cannot be performed by a general-purpose computer without special programming can 112(6) be avoided by claiming a general-purpose computer term (such as code or program or processor) and a software-implemented result, if a person of ordinary skill in the art would understand how to write a program to achieve this result; or is the claim required to recite an algorithm to achieve the software-implemented result to avoid the application of 112(6)?

Dyfan En Banc Petition.

As you can see, the focus of the questions tie the case to to general-purpose computers and software-implemented inventions, and thus implicitly raises overlay questions of subject matter eligibility. A group of law professor led by Mark Lemley have also filed an amicus brief in support of rehearing in Dyfan.

There is a reason Congress limited functional claiming in 1952. And that reason applies with particular force in the software industry. Many of the problems with abuse of software patents can be traced to the fact that we historically allowed patentees to claim functions, not implementations. It is broad functional claiming that leads to assertions that every part of a complex technology product is patented, often by many different people at the same time. It is broad functional claiming that puts stars in the eyes of patent plaintiffs, who can demand huge royalties on the theory that there simply is no other way to implement the technology they have patented. And it is broad functional claiming that makes most of the resulting patents invalid, since even if ten programmers developed ten different algorithms to solve a problem only one of them could be the first to solve the problem at all.

Williamson reined in those abuses by setting a clear rule: patentees can’t avoid section 112(f) simply by picking a magic word other than “means.” The panel opinion directly contradicts Williamson. This Court should grant en banc rehearing to resolve the conflict

Professor Brief in Dyfan.

In VDPP, the defendant (Vizio) asks the following questions in its petition:

  1. May a claim term that connotes structure to one of skill in the art nonetheless require means-plus-function treatment because the structure it connotes is not sufficient for performing the functions the structure is recited as performing
  2. Must the proponent of means-plus-function treatment of a claim term that does not include the word “means” adduce extrinsic or other evidence beyond the language of the claim itself, in order to meet its burden of production in rebutting the Williamson presumption—or may the proponent satisfy its burden by showing that on its face, the claim does not recite sufficient structure?
  3. May the algorithmic-disclosure requirement that this Court has erected as a bulwark against purely functional claiming for computer-implemented inventions be evaded by reciting in a claim, rather than the specification, a generic “processor adapted to” perform claimed functions?

Vizio Petition.

What is missing here:

  • The court decisions lack some clarity as to what it means to “recite sufficiently definite structure.” Is this linked to the definiteness test of 112(b)?
  • It is unclear what “evidence” is required for a court to overcome the presumption — in particular, must a court always look outside the patent document itself?
  • How should all this be handled by the USPTO during ordinary prosecution?

133 thoughts on “Means Plus Function; Nonce Words; and En Banc Petitions

  1. 10

    Below at the termination of numbered comments post 7.2.2.1.2.1.1 (which is yet another asinine comment from Greg “I Use My Real Name” DeLassus (which is nothing more than a logical fallacy of appeal to authority of which Greg has not only repeatedly shown that he lacks in the content of his posts, but to which he has also ADMITTED to not understanding computing innovation and the patent practice associated therewith), Night Writer comes out and (again) expresses his (understandable) frustration with the comments on the subject.

    To me (and NW can correct if he cares), the number one source of frustration can be summed up by characterizing the comments of those who would proselytize AGAINST innovation protection for computing innovations as being simply inte11ectually dishonest.

    This inte11ectual dishonesty comes in several REPEATED forms.

    Several of these forms, of course, have been discussed by others, including yours truly.

    It is extremely notable that these cogent points are NEVER directly addressed by those whom I shall (generously) label as Anti’s.

    Let’s take for example the canard of “point of novelty.”

    This is often used (across a spectrum of legal points) as some “distinguishing” in order to MISAPPPLY the actual law as written by Congress.

    In each and every use of this canard, understanding BOTH what is attempted – and why such attempt is ill-founded – leads simply to a better understanding of how – and why – innovation protection must be understood as a welcoming of tomorrow’s innovation, as opposed to some wooden “only prior types of innovation can be protected.”

    One of these anti’s (Malcolm to those who missed his prior reign of absurdity prior to a year plus sabbatical and his rejoining this community under a new moniker of “The Prophet”), wants to celebrate confusion and conflation of Different Sections of Law with his “Also referred to as ‘the seamless web.’” – sure go ahead and try to find any reasoned legal work that uses that phrase that is not a propaganda piece and anti-ISM – at 5.2.2.6.

    It is the use of “point of novelty” in the very sense of a purposeful confused and conflated sections of law and “merging of concepts that is nothing less than an abject MISapplication of law – in order to subvert the actual meaning of law.

    NW’s posit to this is the simple – and unquestionable reply that fits anytime someone tries to parse claims inappropriately: “ as the claims are to be read as a whole.

    NW goes on to explicate that “nonce word” – which means “nothing word” cannot have “module” be rendered as attempted because – at its core – the word has a very specific meaning OF structure.

    Greg tries – in his inartful, and lacking understanding of THIS art way – to insert a degree of “SPECIFIC” structure as an unstated heightened requirement over the notion of “nonce.”

    But this is a moving of the goalposts – and an intrusion into the spectrum protected FROM the “FULLY functional” – optional – CHOICE of claiming in “means plus” OR “steps plus” language.

    As I have also pointed out (a couple of times now), when the Good Professor earlier ventured into this spectrum he (perhaps inadvertently) coined a phase that fits most exquisitely: Vast Middle Ground.

    The bottom line on the whole 112(f) CHOICE is that the Act of 1952 – in MORE THAN JUST 112(f) blessed the use of terms sounding in function, and that by entering into SOME structure by which function can be realized by a Person Having Ordinary Skill In The Art, 112(f) has NOT been chosen.

    FAR FAR FAR too many anti’s rebel against the use of terms sounding in function – for no doubt a plethora of their own personal reasons. Many of these overlap in having an overzealous desire for exacting picture claims. But make no mistake – these feelings simply are not in accord with the Rule of Law.

    1. 10.1

      “Several of these forms, of course, have been discussed by others, including yours truly.”

      You’ve discussed the general problem of habitual liars before? I’m very surprised to hear that.

      1. 10.1.1

        You’ve discussed the general problem of habitual liars before? I’m very surprised to hear that.

        No. No you are not. But this
        L
        I
        E
        is itself not a surprise, – to anyone, as I have long covered your propensities, even creating an acronym that found its way into the George Carlin list.
        A
        O
        O
        T
        W
        M
        D

        For some odd reason, you like to pretend that the last decade or so of my taking you to the woodshed has not been captured in the plain black and white of the pages of this blog.

        Do you think perhaps that your new moniker of “The Prophet” somehow expunges all those bE@td0wns that you had under the Malcolm Mooney, Malcolm, of MM monikers (or the nigh countless other s0ckP uppets that you employed at PatentDocs AFTER you whined about how such were “the worst thing ever?”

        I am just not certain why in the world you would think that your stale decrepit games — which have never worked — will somehow now ‘magically’ work for you.

    2. 10.2

      “FAR FAR FAR too many anti’s rebel against the use of terms sounding in function”

      Also known as “functional terms” or “abstractions” but let’s not get into the nuances.

      1. 10.2.1

        Ah Malcolm – your “understanding” of nuances is a study in contradictions.

        You just don’t have any such understandings.

        Clearly, functional terms abound in patent law — David Stein tore you into tiny tiny pieces on that point many times.

        But you be you – it only makes my points stand out even better.

          1. 10.2.1.1.1

            Oh, I am quite sure of it.

            He did it with non-computing functional language.

            It really isn’t difficult to see this Malcolm.

            You have ZERO cognitive point to add to pretty much ANY point of patent law — and you’ve relied almost exclusively on your vapid ad hominem short script like forever.

            It would actually be a bit stunning if you managed to mumble something halfway intelligent.

  2. 9

    Bottom line here – so sorry Malcolm and Paul — is that the Grand Hall Experiment simply destroys ALL of your pretending.

    Your desire for exacting picture claims are simply not what are required under the law as written by Congress.

    Deal with it.

      1. 9.1.1

        There is not enough calls to clean up what you leave behind, Malcolm.

        Here’s on idea though: reply cognitively and on point to the substantive points of discussion rather than leave behind your usual droppings.

        Granted, this may emulate your leave of absence of just over a year, but would that really be that bad? You were never more in line with how patent law actually is than during your absence.

  3. 8

    DC: “The court decisions lack some clarity as to what it means to “recite sufficiently definite structure.””

    Logic and algorithms have no non-abstract “structure” . The absurd idea that an “algorithm” (an unpatentable abstraction which has no structure) could somehow have “sufficient structure” in some cases was a magical band-aid created out of thin air by the Federal Circuit to keep the logic patenting gravy train on the tracks. This was noted in real time and, years later, we can see that nothing has been achieved other than confusion and the unnecessary promulgation of hundreds of thousands of bad patents (and litigation over metaphysical nonsense like “when does something without structure nevertheless have sufficient structure).

    As noted below, in the digital world, logic only “connotes” a definite structure when there is a “glossary” directly connecting that logic to a definite structure in the computing machine. Such glossaries are never provided by applicants.

    What might help reach a resolution of some sort is if the lawyers and the judges and the PTO at least explicitly admit what even the supporters of logic patenting admit: in the computing context, neither 112, 103 or 101 are applied in a manner consistent with their treatment in other contexts or historically. What’s happened is that exceptions were created out of thin air and those exceptions are continually being revisited to paper over their (predictable) nonsensical results.

    There is a reason that logic (like other abstractions) was left out of the eligible categories of patentable subject matter …

    1. 8.1

      >>Logic and algorithms have no non-abstract “structure” . The absurd idea that an “algorithm” (an unpatentable abstraction which has no structure) could somehow have “sufficient structure” in some cases was a magical band-aid created out of thin air by the Federal Circuit to keep the logic patenting gravy train on the tracks.

      This is at the level of a witch burning. No structure? So a machine with “no structure” can drive? Just shameful that you continue to push these ridiculous arguments to deceive the ignorant.

      No one but the most uneducated and ignorant buy arguments like this anymore MM. I suppose all those chip manufacturers are making chips with no structure. I suppose your computer brings up this website and allows you type in a comment with “no structure”.

      Just absurd nonsense. Please stop.

      1. 8.1.1

        “ a machine with “no structure” can drive?”

        Nobody said or even suggested such a thing. The discussion is about the definiteness of the description of art-distinguishing structure that needs to be provided in order for an allegedly “inventive” machine to be claimed and granted protection in our patent system.

        Your response here is the same tired response you (and also anon) have been floating for eons: a strawman about “structureless machines” which completely missed the point. It’s a non-starter for you but go ahead and pat yourself in the back anyway. You’re a very serious person and totally not some washed up old cable-consuming hack with “economic anxiety.”

        1. 8.1.1.1

          MM, I think this will be my last response to you–EVER.

          To suggest that the language of an algorithm does not connote definite structure and does not to one skilled in the art connote the metes and bounds of the claims is just ignorant nonsense. Try to read a book on algorithms. Just look at a upper-level computer science course on algorithms and read the book. I’ve quoted such books on this cite with quotes about how functional language is used to represent known solutions or how software/firmware/hardware solutions are equivalent. One skilled in the art has no problem with this.

          Just more ridiculous nonsense from you. Just sickening that people like you exist to intentionally muddle the waters with outright l i e s.

          Never going to respond to you again.

          1. 8.1.1.1.1

            Night Wiper: “an algorithm does not connote definite structure”

            That is correct, with the caveats noted above, and this isn’t even controversial. The only controversy is how do we deal with this fact form the standpoint of creating a system for protecting abstractions such as algorithms and logic. The utility patent system was certainly not designed to do this (hence the exceptions created out of thin air as a gift to logic patentees).

            1. 8.1.1.1.1.1

              logic patentees

              Please stop
              L
              I
              E
              I
              N
              G

              You are aware that the choice of wares – from soft to firm to hard – destroy this odd misinformation tactic of yours, eh?

              This is not new- you have known of this for at least a decade.

    2. 8.2

      George Carlin filter….?

      Your comment is awaiting moderation.

      May 20, 2022 at 11:16 am

      There is a reason that logic (like other abstractions) was left out of the eligible categories of patentable subject matter …

      [ADDED] Except for the fact that they were not. Software is a design choice of wares in the computing arts — -it is NOT the “logic” that you falsely try to claim it to be.

      IT may involve logic, but it is certainly not limited so.
      Just like – exactly like – your prior false goal post moving of claim terms to be ENTIRELY IN THE MIND, software is a machine component and is NOT something entirely in the mind.

      Your attempts at mischaracterization is nothing more than misinformation.

      [END OF ADD]

      Clearly, Malcolm, you rest all of your screed on feelings – feelings entirely dissociated from the real world in which these types of computing inventions actually work.

      The easiest rebuttal to you remains: feel free to abstain from any use of any item to which you would deny patent protection to.

      This should be easy for you – seeing as you do not seem to think that anything “real” is involved.

  4. 7

    Typical Lemley trash below. Where is the proof to any of these problems? Can Lemley give real examples? How many examples can Lemley give? This is just more of the abstract psychotic writing of Lemley that has resulted in the destruction of our patent system. We should all pause as we see the country being ruined beyond recognition by the policies of the far left. Lemley exemplifies the far left (elitism, made 10’s of millions burning down the patent system, tied directly to those that want to end patents with his wife a Google executive, rules for us but not for Lemley, and so forth). Everything the far left touches turns to trash. Before Lemley started burning down our patent system the USA was number one in innovation in every area of technology. Now we are no longer number one in many areas.

    The big lie with this case–again–is that they are not viewing 112(f) in terms of people skilled in the art. A person skilled in the art of computer science knows that “module” recites structure. If you go to the elance website, you will see 1000’s of specifications for programs to be written in a bidding process. Many use the word module in specifying how the code should be written. That is one example of the absurdities of the case law. Another one is with algorithms. A person skilled in the art of computer algorithms does not need an entire algorithm written down in a method format to understand the metes and bounds of the claims. A sentence or two describing how a known algorithm is changed is sufficient. Again–what we see are trash judges limiting patents driven by literally 100’s of millions of dollars from large corporations paying people like Lemley who produces trash articles that are not peer reviewed and where there is no consequences for unethical conduct.

    Let’s get rid of far left in patents. Look at this problem from the perspective of an inventor and those skilled in the art. Not some agenda-driven far left “professor” who is raking in 10’s of millions of dollars in burning down our innovation system.

    >>There is a reason Congress limited functional claiming in 1952. And that reason applies with particular force in the software industry. Many of the problems with abuse of software patents can be traced to the fact that we historically allowed patentees to claim functions, not implementations. It is broad functional claiming that leads to assertions that every part of a complex technology product is patented, often by many different people at the same time. It is broad functional claiming that puts stars in the eyes of patent plaintiffs, who can demand huge royalties on the theory that there simply is no other way to implement the technology they have patented. And it is broad functional claiming that makes most of the resulting patents invalid, since even if ten programmers developed ten different algorithms to solve a problem only one of them could be the first to solve the problem at all.

    1. 7.1

      The best thing that could happen is for an R to be elected president and to dissolve the CAFC and all their trash case law. It would only take one bill in Congress to end the CAFC and their trash case law.

      The CAFC is full of far left judicial activists who need to go.

      1. 7.1.1

        That would not be enough.

        The state of the current CAFC has two generators: as you note on appointments, but also from the brow-beating from above by the Supreme Court.

        Congress DOES need to reset the CAFC — and I hope that they do so.
        (with the current polarized environment – this is highly doubtful)

        But when they DO get around to it, I do also hope that they employ their Constitutional power of jurisdiction str1pp1ng of the non-original jurisdiction of patent cases from the Supreme Court and remove that body from the loop.

    2. 7.2

      If you go to the elance website, you will see 1000’s of specifications for programs to be written in a bidding process. Many use the word module in specifying how the code should be written.

      Right. I do not think that anyone is questioning whether “module” connotes structure. Rather, the point in dispute is whether “module” connotes sufficient structure to accomplish the claimed function.

      Imagine that I write a claim that recites “… a sphere that kills tumor cells.” Obviously “sphere” is a structural designation, but there are plenty of spheres that do not kill cancer cells (e.g., oranges, soccer balls, etc.). Therefore, while “sphere” is structure, it is not sufficient structure. Same for “module.”

      1. 7.2.1

        Except module was held to be a nonce word.

        One can see the ill intent with these judges in that rather than trying to shape the law to accommodate modern technology that are activists using the law to limit patents.

        1. 7.2.1.1

          Except…

          Not sure why you used that word – Greg’s point IS that “module” (by itself) IS a nonce word.

      2. 7.2.2

        Sphere is a general descriptive term from the English language to describe the shape of atom or molecules in your example.

        Module is a term that is a term of art used in field of software engineering. So the case holds that a term of art that has a known meaning for what structure connotes is a nonce word.

        Nice try.

        1. 7.2.2.1

          And just in case you decide to continue with your nonsense read WT comments below.

          The point is that why would “processor” configured to connote structure and not a “module configured”? To a person skilled in the art of information processing the position taken by the CAFC is clearly wrong and can only be reconciled by understanding that the CAFC is filled with trash judges that are using 112(f) to weaken patents.

          And the case for the algorithm is even stronger and I am someone that has taught algorithms at a top 5 university.

          1. 7.2.2.1.1

            Wiper: “I am someone that has taught algorithms at a top 5 university.”

            Beyond parody.

            Tell everyone what is the necessary objective structural difference between a module configured to correlate real estate locations with projected sale prices versus a module configured to correlate tumor locations with probability that surgery will be successful.

            What’s the structural difference? Tell everybody, O Wise Professor.

          2. 7.2.2.1.2

            [W]hy would “processor” configured to connote structure and not a “module configured”?

            Agreed. There is no good reason to distinguish between those two. I would treat them both as M+F (if one party or the other properly raises the issue to rebut the presumption). If one is not going to apply an M+F construction to one of those, however, then one also ought not to apply M+F to the other.

            1. 7.2.2.1.2.1

              Greg, the issue is actually more subtle.

              The process and module both connote definite structure. The issue is whether or not what is claimed connotes a family of solutions to one skilled in the art or not.

              I also find it odd that you think you can tell whether something in an area you know very little about connotes definite structure to one skilled in the art.

              The entire issue is that the CAFC is removing the person skilled in the art in their analysis.

              1. 7.2.2.1.2.1.1

                I also find it odd that you think you can tell whether something in an area you know very little about connotes definite structure to one skilled in the art.

                This is not a totally whacky critique, but here’s the thing—claim construction is a question of law, which means that it is made by the judge. Most federal judges are not computer scientists, which means that—by necessity—claim construction must be a task that can be performed by a non-expert.

                If you really think that only a computer scientist can accurately construe a claim reciting “module,” then that is as much as to say that you also believe that the U.S. patent system has never worked for software inventions, does not presently work for software inventions, and cannot work for software inventions. If you really believe that, why not just lie back and welcome Justice Stevens’ Bilski concurrence as the law of the land?

                But, of course, you do not believe that. If—as I believe to be the case—we both believe that the U.S. patent system should work for software inventions, then one must also accept that non-computer scientists can construe software claims.

                Speaking as a non-computer scientist, I notice that there are “modules” that calculate interest, and “modules” that route telephone calls, and “modules” that adjust engine speed, etc. Is this only one single software structure that can achieve all of these different outputs? That would be truly a remarkable invention—that “module”—were it so.

                We both know, however, that it is not one single software structure. One programmer creates the “module” that routes calls, and another creates a totally different “module” that adjusts engine speed. Evidently, then, there is not one single software structure or class of structures that can be recognized as an “engine speed adjustment module.” Even a non-computer scientist (such as myself) can figure this much out.

                With that in mind, it follows that—where the “module” sits at the point of novelty in the claim—it must either be regarded as invalidating for want of written description/enablement or given a §112(f) construction. That is the way it works for every other art field. If software is to be patentable on the same terms as every other art field (and it should!), then that is how software claims must be treated.

                1. No Greg – law does not work like that. If the judge needs assistance to get to the meanings of terms of art, he obtains it.

                  He does not get to say, “well, I don’t know that, so we are just going to use some common vernacular and call it a day.”

                  JFC – you are an attorney, eh?

                  Even if you are “in house,” you ought to know these things.

                2. That you push for Stevens (loss of majority opinion) and dissent dressed as a concurrence as some type of “acceptable” fall-back shows just how anti-patent you are.

                3. Claim construction is a mixed with law and fact Greg.

                  And the claims are supposed to be interpreted in view of the specification by a person skilled in the art.

                  And module is not at the “point of novelty” as the claims are to be read as a whole.

                  Additionally, the issue here is that “module” should never had been held to be a nonce word as it is not. The issue was there definite structure in the claim as a whole viewed through the lens of a person skilled in the art in view of the specification.

                  That is the issue. The fact that the CAFC shifted the issue and absurdly held that “module” is a nonce word indicates clearly that the CAFC is full of anti-patent judicial activists.

                  And again—pretty tired of ridiculous people. I work with real inventors where 90 percent of the patents I write are for real products. The ones that aren’t for real products are because they were prospective products that were turned down by management after initial development.

                  The CAFC is making no effort to adopt their case law to how real inventors work and to accommodate real inventions and the families of structures that can implement an invention in the CS/EE/mechanical fields.

                  When I show a claim that is tailored to get rid of the issues that the CAFC has artificially raised I get push back from Ph.D. inventors saying but that would be easy to work around as they could just do it with a A/D convertors and software instead or they could just use a different algorithm to accomplish task X.

                  From the real world of inventors the CAFC is a cease pool of filthy nasty judicial activists that want to burn down the patent system and are doing it in a dishonest fashion.

                  Reality!

                4. Pretty tired of this nonsense coming from people like you Greg.

                  I would challenge you to work on a real invention in the CS/EE/mechanical world at the forefront of technology like I do. Go through the families of structures and how functional language is used to capture known inventions. And so forth. Then imagine that your continued work from the corporation is going to depend on you being able to get a claim chart together that holds water for licensing. I get shxt from inventors all the time who are smart as they come and can figure out in a second how someone can get around a claim.

                5. >>This is not a totally whacky critique,

                  The totally wacky shxt often comes from you Greg. And how about stopping the ad hominin attacks where you act as if there is some presumption that my posts are “wacky.”

                  F U!

                6. [I]magine that your continued work from the corporation is going to depend on you being able to get a claim chart together that holds water for licensing. I get shxt from inventors all the time who are smart as they come and can figure out in a second how someone can get around a claim.

                  Why should the rest of us care about your stream of business? The patent system does not exist to keep you gainfully employed.

                7. Greg writes >>>Why should the rest of us care about your stream of business? The patent system does not exist to keep you gainfully employed.

                  Goodness. What I described is at the heart of the patent system’s purpose. To provide an incentive to invent. What I described is how the judges on the CAFC are purposefully making it more and more difficult to write claims that capture a real invention and protect that real invention.

                  So everyone should care because that is what the patent system is supposed to be for.

                  Goodness Greg. Seriously? Do you really have no clue? I really thought you were better than this. But you seem to have fallen into acting like MM. Stop reading his posts. They have poisoned your mind.

                8. Stop reading [MM’]s posts. They have poisoned your mind.

                  It was years ago that I stopped reading MM’s posts. They are not worth my time, or anyone else’s, come to that.

                9. Night Writer,

                  Quite frankly I am shocked that you are shocked as to Greg’s views (as biased as they are due to his Big Pharma proclivities).

                  He has been like this for several years now, as you no doubt likely recall his willingness to accept the Trojan Horse of “112 Reform.”

                10. [T]he issue here is that “module” should never had been held to be a nonce word as it is not… The fact that the CAFC shifted the issue and absurdly held that “module” is a nonce word indicates clearly that the CAFC is full of anti-patent judicial activists.

                  You have this all backwards. Your side of this argument is the one that wins at the CAFC. I am arguing in favor of the losing side of this argument.

                  I am nobody. I am not a Congressman. I am not a judge. I am not even an arbitrator. You should care less about what I think than you would care about the last piece of tissue that you tossed in the trash.

                  Be happy that your side always wins this argument. Getting angry at me merely for advancing the contrary point of view is just going to give you an ulcer. I merit your contempt—perhaps—but not all this anger.

                11. Somehow Greg wants to chastise Night Writer here….

                  WTF?

                  You have this all backwards. Your side of this argument is the one that wins at the CAFC. I am arguing in favor of the losing side of this argument.

                  Greg:
                  stop
                  digging

                12. If Greg has any friends out there (heck, any acquaintances), you might do Greg a favor and pull him away from his keyboard.

                13. [T]he claims are supposed to be interpreted in view of the specification by a person skilled in the art… The issue was there definite structure in the claim as a whole viewed through the lens of a person skilled in the art in view of the specification.

                  In every case in which a putative “nonce” word is contested in claim construction, the patentee and the defendant both get to put on expert testimony. The experts are all people skilled in the relevant art field. The judge is almost never an artisan in the relevant field, so the judge’s job is always going to be to weigh the credibility of the various experts presented.

                  You seen aggrieved that judges do not always see these terms as you do. But that does not mean that the judge is ignoring what the skilled person thinks. The judge hears from two different sets of skilled persons who tell the judge two diametrically opposite stories. There is no such thing (in the 112 context) as a single skilled artisan. There are just a vast heard of skilled artisans, some of whom think one thing and some of whom think another.

                  You seem to think that the judge should always credit the expert that you favor. Fair enough. I also think that the judge should always favor my expert. A grown up knows, however, that you win some and you lose some. There is no upside to stewing in it for years after.

                14. Another logical fallacy form Greg – presupposing that the ONLY reason why Night Writer feels strongly about his position is BECAUSE the cases are “in his position.”

                  Greg – this is plain error, as “the position” under evaluation is a LEGAL ONE – and not one that is dependent on case by case facts.

                  Sure, in a contested situation, there may well be advocacy ON contested facts – with attorneys on each side.

                  THAT is not the legal point under discussion here – HERE, the legal discussion is on a higher plane that “module” is not a nonce word because it is a term of art with structure as understood IN the art.

                  Those pushing for “nonce” simply do NOT have reality on their side.

                  Quite in fact, were the courts at all perceptive, they would be leveling sanctions against attorneys pushing for such treatment as “the word is only a nonce word.”

                  We ALL would be better off if the anti’s were sanctioned for their (known) false proselytizing.

  5. 6

    [W]hat “evidence” is required for a court to overcome the presumption — in particular, must a court always look outside the patent document itself?

    I suppose that you could imagine a situation in which a patentee wrote “as used herein, an ‘output maximizing circuit’ is any circuit that achieves output maximization, regardless of structure.” In such a circumstance, one would not need to look outside the four corners of the document to determine that “output maximizing circuit” does not convey sufficiently definite structure.

    In the real world, however, the patentee is much more likely to have written something like “those of skill in this art know what sorts of structures are encompassed by ‘output maximizing circuit.” In that context, I am inclined to say that the determination nearly always will require consulting evidence beyond the four corners of the patent document.

    The operative question is what the skilled person understands from the words “output maximizing circuit” (or whatever the relevant phrase is). The way that you answer that question is by asking some skilled people (i.e., expert witnesses). Inevitably, then, in the sorts of circumstances that actually arise in litigation, the party pushing for a §112(f) construction in the absence of the word “means” will need to adduce a witness to testify that the language does not convey sufficient structure. If the other party wishes to resist the §112(f) construction, they will also need to adduce a witness to testify the other way.

  6. 5

    Getting a rehearing on 112(f) re the scope of Williamson v. Citrix for functionally claimed software from the CAFC is, of course, unlikely.
    But the subject Dyfan v. Target amicus of law professors led by Mark Lemley is interesting. Because some it strikes me as more like an argument for “abstract” unpatentable subject matter rejections for claim “preemption” than 112(f)? I.e., “It is broad functional claiming that leads to assertions that every part of a complex technology product is patented, often by many different people at the same time.” “..patent plaintiffs, who can demand huge royalties on the theory that there simply is no other way to implement the technology they have patented.” Saying “no other way to..” sounds the same as saying that the claim is so abstract [or so overbroad for its spec] as to pre-emptively cover every possible way anyone else invents?

    1. 5.1

      BTW, note the last line of the petition: “..or is the claim required to recite an algorithm to achieve the software-implemented result to avoid the application of 112(6)?”
      Are there not some prior CAFC decisions noting the importance of having algorithms in the drawings or text of the spec for software implemented inventions?
      But in what situations could that be required?

      1. 5.1.1

        Right. Aristocrat Techs. v. Intern. Game Tech., 521 F. 3d 1328, 1333 (Fed. Cir. 2008) explains that an algorithm is the requisite structure to support an actual means+function claim in the software context. It makes sense, then, that an actual algorithm should need to be recited (where relevant) in a claim to avoid M+F construction under Williamson, but I have never seen an actual court require as much. Quite the contrary, “circuit” seems always to be treated as sufficiently definite structure, without more.

        1. 5.1.1.1

          The “turtles all the way down**” would quickly point out that any actual algorithm recited in a claim could be argued to be nothing more than “step for” language.

          Yes, “step for” is a real thing in the family of law to which the larger label of “means for” is often used as a shortcut.

          **this analogy also fits with those who have proposed having a “definitions section” be made mandatory. Definitions of words are — gasp — made up of more words.

    2. 5.2

      I.e., “It is broad functional claiming that leads to assertions that every part of a complex technology product is patented, often by many different people at the same time.” “..patent plaintiffs, who can demand huge royalties on the theory that there simply is no other way to implement the technology they have patented.”
      While there have been much talk about “weak patents,” arguments in the abstract are also weak. This represents a good example of an argument in the abstract.

      Computers perform functions. In a computer program, those are oftentimes the elements.

      If part of a claim to a vehicle includes “a transmission mounted to an engine using a movable hinge,” I dare say few would argue that these limitations are abstract. However, to one having ordinary skill in the art, there a thousands and thousands of way to implement that language. There are thousands of different transmissions and thousands of engines and an untold number of movable hinges. The fact that broad language is used does not make the claim language abstract.

      Now consider the computer language of “routing, to one of a plurality of servers, a request based upon a flag set within the request and a workload, respectively, of each of the plurality of servers.” The routing is a “function.” There are probably hundreds and hundreds of ways to implement this function. However, that doesn’t make it abstract.

      sounds the same as saying that the claim is so abstract [or so overbroad for its spec] as to pre-emptively cover every possible way anyone else invents?
      That same (il)logic can be applied to the claim language I discussed above with the engine, transmission, and hinge. The problem with many of the anti-software patent arguments are that they are presented by people who do not understand that the claiming conventions in software are little different than the claim conventions in electrical/mechanical devices. As such, when they say “so and so is really bad and we need to change it,” they do not appreciate that their proposed changes screws up everything else before their sole focus is on software.

      Regardless, no one is claiming things comparable to Morse’s claim 8 (i.e., “I do not propose to limit myself to the specific machinery or parts of machinery described in the foregoing specification and claims; the essence of my invention being the use of the motive power of the electric or galvanic current, which I call electro-magnetism, however developed for marking or printing intelligible characters, signs, or letters, at any distances, being a new application of that power of which I claim to be the first inventor or discoverer.”)

      1. 5.2.1

        The problem with many of the anti-software patent arguments are that they are presented by people who do not understand that the claiming conventions in software are little different than the claim conventions in electrical/mechanical devices. As such, when they say “so and so is really bad and we need to change it,” they do not appreciate that their proposed changes screws up everything else before their sole focus is on software.

        Completely agree.

        1. 5.2.1.1

          +1 I agree too.

          Just reading through these comments it is clear that people don’t understand scope of enablement–what is enabled is the issue.

      2. 5.2.2

        If part of a claim to a vehicle includes “a transmission mounted to an engine using a movable hinge,” I dare say few would argue that these limitations are abstract.

        sounds the same as saying that the claim is so abstract [or so overbroad for its spec] as to pre-emptively cover every possible way anyone else invents?

        That same (il)logic can be applied to the claim language I discussed above with the engine, transmission, and hinge. The problem with many of the anti-software patent arguments are that they are presented by people who do not understand that the claiming conventions in software are little different than the claim conventions in electrical/mechanical devices.

        The same logic cannot be applied because you miss the relevant feature of the distinction which is the purported novelty of the feature. The transmission example you gave is presumably a context limitation describing known things. It isn’t the possibly inventive feature. But the computer software is the possibly inventive feature, and describing that inventive feature by the result of its use rather than the structure of what it is has the effect of extending the monopoly beyond the invention, because it necessarily includes things that the Applicant has not thought of but, once they are thought of, would produce a result that is at least marginally similar to one that was thought up.

        For example, the limitation “[code for] calculating an optimal route and minimal travel time between an input location and a current device location” is entirely acceptable when driving direction apps are known to the art and the specification attempts to do something inventive with the calculated route and time (e.g. they are preliminary data that are later acted upon in a nonobvious manner), and entirely inappropriate when the art never had a driving direction app and the invention is an algorithm that calculates the route and minimal travel time. That allows a person with a terrible algorithm (e.g. I posit that cars travel on all roads at 15mph) to encompass excellent algorithms (e.g. Google Maps) despite not positing them and importantly not disclosing them to the public.

        To use a more physical example – If one invents the bow, the claim cannot be for a machine that fires a projectile. The lack of structural limitations is an attempt to extend the monopoly beyond the invention. For example, that claim would also cover a gun, even though the inventor did not posit and did not enable a gun and a gun operates under entirely different principles of operation than a bow, even though it also performs the same general act of discharging a projectile. But if one is inventing a firearms safety system and that system includes, as part of it, a machine that fires a projectile in a world that is well aware that machines fire projectiles, that functional language is appropriate.

        Regardless, no one is claiming things comparable to Morse’s claim 8

        As someone who examines computer software, virtually everyone is claiming comparable to Morse’s Claim 8. The allegedly inventive algorithm is either never disclosed at all, or disclosed but claimed functionally. But one of ordinary skill would discriminate in the manner in which a function is implemented.

        Take the previous example of a device calculating a driving route and a driving time. That disclosure is an abstract idea (“We should use the processing power of a computer in order to calculate an optimal driving route and time”) together with a non-limiting bad assumption disclosure (“Cars always travel at 15mph”). There is absolutely no evidence that the applicant ever posited using real-time data, for example. Or even historical data from a similar time the previous week. And yet the claim circumscribes those unposited algorithms. Current map apps offer the option of calculating based upon whether or not to use toll roads or expressways. There is no evidence the disclosure ever considered that either. But all of these unposited algorithms still accomplish the broad function of determining a route and a driving time.

        It doesn’t much matter if you view the problem as one of a failure to point out and particularly claim (defining something by the result of its use does not particularly describe what the thing is), or as a failure of enablement or possession (Applicant fails to show that they possess the full range of their functional language) or as an abstraction (Applicant merely asserts an abstract goal without concrete limitations on how to arrive at the goal). The problem is unquestionably the use of functional language as part of the nonobvious feature. That’s why the courts previously (and I say previously only inasmuch as they have not said it recently, not that it has ever been overruled) had a rule against functional language at the point of novelty. That’s why means-plus claiming exists at all – because of the Halliburton patentee using vague functional language to describe their contribution.

        It is, in actuality, the same problem as Morse, your current applicant just throws in a handful of context limitations first: A processor, memory, a few context limitations to set up the problem, and then, when the critical point of describing the new thing that this inventor contributes, they don’t limit based what they thought of at all, they limit based on the effect of its use. Morse’s claim wouldn’t have been any more valid if he had thrown in token “receiving message data” and “publishing the printed characters” limitations so long as he remained vague about the function his machine provided: “conveying information using electromagnetism to print at a distance.” That reads on the cell phone text message as well as the telegraph. Are we to believe that there is no harm in allowing Morse’s claim to reach the cell phone text?

        As General Electric v Wabash said: “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.”

        It is one thing to use functional language to describe a category of old things. It is quite another to attempt to enlarge the invention by using functional language in lieu of structural language for the newly invented thing. The fact that functional language may generate a broad but definite category does not mean that the category is nonabstract or fully possessed.

        You can’t invent the function of achieving a result. You can only invent a handful of posited means that will get you there. Those posited means are a valid claim. The function itself is not. How can you ever prove that you possess all the means for achieving a result?

        It’s virtually never appropriate to argue that claims are nonobvious because of functional language, and yet virtually every computer software claimset does just that.

          1. 5.2.2.2.1

            Yes, there is sometimes conflation or confusion of proper and allowable “functional” claiming with “result” claiming. But note they they can partially overlap when the function is being claimed so broadly that it “preempts” any way of achieving the desired result, such as a claim with only a bald functional statement of achieving the desired result.

            1. 5.2.2.2.1.1

              All claims preempt — that is what they do.

              You are simply trying to slice your bologna too thin.

              1. 5.2.2.2.1.1.1

                All claims preempt — that is what they do.
                Why don’t they get that?

                The claim term of “transmission” preempts thousands and thousands of different types of transmission.
                The claim term of “switching” preempts thousands and thousands of different types of switches.

                BTW, what does a transmission do? It transmits power from one thing to another. The functional nature of what it does is baked into its name. The same can be said about a switch.

                1. The funny thing Wt, is that Malcolm (and the cheerleading of Paul), as well as Random, has been answered a long time ago — it was Dave Stein as I recall that shredded these folk with numerous examples of non-computing claims directly on point.

                  These people have no real basis for their feelings at law.

        1. 5.2.2.3

          But the computer software is the possibly inventive feature, and describing that inventive feature by the result of its use rather than the structure of what it is has the effect of extending the monopoly beyond the invention, because it necessarily includes things that the Applicant has not thought of but, once they are thought of, would produce a result that is at least marginally similar to one that was thought up.
          The elements of software are its functions. These functions are almost always known. Just like the known elements of most electro-mechanical inventions. It is the combination of the elements in new ways is what is inventive. BTW, a “function” in software is little different than a “step” in an electro-mechanical claim.

          For example, my example claim language referred to a transmission (mechanical) and the a request (computer). While the “request” itself isn’t a function, the receipt and storage of the request are functions. Just like there are 1000s of ways to implement a transmission, there are 1000s of ways to implement “receiving a request from a client computer.”

          It’s virtually never appropriate to argue that claims are nonobvious because of functional language, and yet virtually every computer software claimset does just that.
          Steps of a method are functional language. Instead a “receiving a request from a client computer,” let’s say our claim language is:
          “receiving a supply of gasoline;
          nebulizing a portion of the supply; and
          directing the portion into a cylinder of an engine.”

          These are all functions (of a fuel injector). This is functional language. This is why I previously wrote that “the claiming conventions in software are little different than the claim conventions in electrical/mechanical devices.” I write claims for both software and electro-mechanical devices. The claiming conventions are use for both are nearly the same.

          Please explain to me the difference between a method step and “functional language.”

          Also, what is the difference between: “welding A to B” and “A is welded to B”? There isn’t much difference between the two. One is an act (i.e., function) and the other is the result of the act. However, it will be difficult to get one without the other. Regardless, there are thousands of different ways to accomplish the act — including ways that have NEVER been done before and would be patentable in their own right. Are you saying that language such as this cannot be used to distinguish a claimed invention over prior art?

          You can’t invent the function of achieving a result. You can only invent a handful of posited means that will get you there.
          So if my invention, as a collection of elements, include ‘A welded to B,’ then must I describe the specific technique used (e.g., TIG, MIG, GTAW, etc.)? However, since each type includes many different types of variations, do I also have to disclose the variation used? We can go down this rabbit hole pretty deeply before it ends.

          Any language used is necessarily ‘abstract’ because words are abstract. The term “transmission” is abstract in that it could have tens of thousands of variations. If I claim a transmission, am I required to list the specific transmission needed?

          It doesn’t much matter if you view the problem as one of a failure to point out and particularly claim (defining something by the result of its use does not particularly describe what the thing is), or as a failure of enablement or possession (Applicant fails to show that they possess the full range of their functional language) or as an abstraction (Applicant merely asserts an abstract goal without concrete limitations on how to arrive at the goal).
          Your first point misstates 112, 2nd. Claims do not have to “describe what the thing is.” Examiners mess that section up all the time. The operative phrase is “distinctly claim.” You need to distinguish the invention — not describe it. Describing the invention is what the specification is for.

          As for “possess[ing] the full range of their functional language,” can you identify the case law regarding that. How, for example, can I show possession of the rule range of “welding A to B”?

          As for “Applicant merely asserts an abstract goal without concrete limitations on how to arrive at the goal,” this takes us down the same rabbit hole I mentioned earlier. I claim “welding A to B” and you argue that you haven’t described any “concrete limitations.” I then claim “welding A to B with TIG” and then you argue that there are thousands of ways to do that so the claim language is still abstract. I keep adding on limitations, but you keep arguing (rightfully) that there are still hundreds of different ways of doing that. We can keep going down that rabbit hole for a long time. The end result would be claim language that would be TRIVIAL to design around — which is the EXACT RESULT attempted to be achieved by the anti-patent crowd. This is what they are looking for. The want more and more and more and more claim limitations so that infringement is a very unlikely scenario. This is why we push back so hard because we can see where this is headed if taken to its logical end.

          1. 5.2.2.3.1

            Wt,

            Not only are you exactly correct, this has been explained nigh countless times already to Random.

            It
            does
            not
            matter
            to
            him

            He will repeat his inane script no matter what.

          2. 5.2.2.3.2

            The elements of software are its functions. These functions are almost always known. Just like the known elements of most electro-mechanical inventions. It is the combination of the elements in new ways is what is inventive. BTW, a “function” in software is little different than a “step” in an electro-mechanical claim.

            You’re conflating different legal concepts. A function is the effect or the result of using means. Means are a series of actions or structures, some of which are non-equivalent. It’s certainly possible to confuse means and functions when you lack expertise in the area or when you are intellectually lazy, which is why I like to use broad concepts that lay people understand. One can conceive of functional language that relates a bow and a gun (they both perform the function of firing projectiles) but even lay people would understand that the structures of a bow and gun are not equivalent, as they rely upon entirely different principles of operation. And yet the average attorney who posts to this board appears to be under the misapprehension that if they’re just a good enough writer they can make the inventor of the bow also be the inventor of the gun. They are mistaken, which leads to a bunch of bad lawsuits (which as a lawyer myself I am rather fond of) but also tragically often results in a true inventor being left without protection because their attorney never actually disclosed, let alone claimed, what the actual invention was.

            The law prevents a person who posits a structure from being given dominion over nonequivalent structures. That is simply the rule, and people fail to understand it, but I assure you there is always a rejection to be made on claim language that is not tailored to the means disclosed.

            While the “request” itself isn’t a function, the receipt and storage of the request are functions. Just like there are 1000s of ways to implement a transmission, there are 1000s of ways to implement “receiving a request from a client computer.”

            The number of ways of doing something is largely irrelevant. The question is whether one of skill would find them to be materially different ways. This has been explicitly stated by courts in a variety of contexts. Take Lizardtech, where the court used the example that disclosing a single “fuel-efficient engine” does not support a claim to all fuel efficient engines. Take Morse, where disclosure of natural law did not give rise to every machine that used it in some manner. There can be 1000 ways of doing something that one of skill would find equivalent to each other and they would all be one means. Conversely, there could be 2 ways of doing something that the art would find to be materially distinct, and those are different means.

            If someone has disease in their arm I can cure the body of disease by amputating the arm. Or I can come up with antibiotics to cure the disease while keeping the arm. Curing the disease is a function that has at least two distinct means. There is no disclosure one could make regarding amputation – no matter how detailed the amputation disclosure and no matter how much boilerplate language that “the invention is the function of curing the disease and the amputation is merely one example embodiment of it” – that would produce a valid claim scope covering using antibiotics to cure the disease. You cannot scrivener your way into a valid scope covering things the art would find to be materially different. Some lawyers think they are smart enough to do so. These are the worst lawyers in the room.

            These are all functions (of a fuel injector).

            I don’t like to get into concrete examples of things that I lack expertise in. Regardless, the rule is the same for all contexts – Either those of skill in that art would find materially differences or not. It is not – as some like to say – turtles all the way down. At some point the language becomes specific enough that one of skill would recognize that everything that falls within that language constitutes materially similar (equivalent) things.

            Also, what is the difference between: “welding A to B” and “A is welded to B”?

            Nothing when the art already knows of the act of welding and the invention is not about welding, but that is not the concern we’re talking about. We’re talking about when someone allegedly strings a new series of acts together for the first, nonobvious time.

            One is an act (i.e., function) and the other is the result of the act.

            Except in a world where welding is novel and nonobvious, one cannot describe the function of welding by saying “welding” – Telling the art “You perform welding by welding” is circular. The manner in which you describe welding IN THE FIRST INSTANCE is to talk about the acts which make up welding. Those acts have to be described with sufficient particularity that one of skill can make and use the invention. Then the question becomes “does disclosing these acts mean this person is fully in possession of all the manners of welding” (i.e. is this technique the only manner of performing the function) and the answer to that is almost universally no, because science is generally inductive and therefore one can never conclude that they have filled the scope of the function.

            The first person who did welding was not the last word in welding, any more than the Wright Brothers were the last word in flying, or Morse was the last word in printing at a distance. The law requires one to possess the full scope of their claim. Disclosure of one manner of performing the function is virtually never enough to assert a claim to the function. Kitty Hawk was not the invention of the function of powered flying. Kitty Hawk was proof that one posited structure had a certain type of functional utility, which others have since expanded upon to reach the modern understanding of flying, which is still incomplete with respect to the function of flying – as the full scope of the function of flying remains outside of our collective grasp today. The Wright Brothers didn’t posit the modern airplane manner of flight, so their attorney better not write a scope that covers all flight, as they clearly didn’t possess it. It is sufficiently great to describe the Wright Brothers as who they were rather than aggrandize them into what they were not.

            Regardless, there are thousands of different ways to accomplish the act — including ways that have NEVER been done before and would be patentable in their own right.

            Which is why no claim can ever be non-obvious because of the function of welding. If we consider the other situation – where welding is known – the reference to the function of welding is permissible, but by definition cannot be nonobvious because the art knows of welding. There may be nonobviousness in the combination (i.e. nobody would think to weld in this situation) and there may be nonobviousness at some other point (i.e. the welding is simply background for a full description of some other nonobvious feature), but THAT other feature can’t be described by function either for the same reason welding couldn’t – you can’t describe the function by restating the function and you can’t use a particular disclosed means to support functional language when disparate means could exist.

            So if my invention, as a collection of elements, include ‘A welded to B,’ then must I describe the specific technique used (e.g., TIG, MIG, GTAW, etc.)?

            As above, this question can’t be answered without describing the state of the art and describing the specification disclosure. In an invention where welding is nonobvious, not only must your specification describe the technique but your claims must be limited to the technique. You obviously cannot enable or possess techniques that you did not disclose if they were previously nonobvious to the art. Do you dispute that a claim must meet 112a over its full scope? This is a direct analogy of the Lizardtech case – The patentee had a first claim that limited by technique which was found valid and not infringed and a second claim not limited by technique that was invalid under 112a.

            However, since each type includes many different types of variations, do I also have to disclose the variation used? We can go down this rabbit hole pretty deeply before it ends.

            You always have to disclose when it would make a material difference to the art, that’s not even really in debate. I don’t know anything about welding but let me give you a simple concrete example – If a TIG technique results in a structure with integrity and a MIG technique results in a structure that falls apart, and you claim welding not limited by technique, guess what? You’re unenabled over the full scope of your claim and you’re going to get a 112a rejection. Now you’re not required to enable every possible embodiment, but at some point the problem will be large enough to tank your claim. So OF COURSE the question of how specific you need to get and what you need to disclose is fact specific and based upon what the art needs disclosed. A rule that the art knows of ways to weld and therefore you can always refer to the function of welding invites rejection when disparate welding techniques are materially different with respect to the claim scope.

            The fact that the layer you need to get to is not a priori easily discernable is cause for caution in writing your specification and in constructing your claim set. But most attorneys seem to take the exact opposite approach – they equate difficult to discern with not being a requirement and throw caution to the wind. Let me give you some free advice – if your client made a machine by welding using a TIG technique and never tried a MIG technique, you might want to ask him if they know of any reason a MIG technique might yield different results, and you might want to include the TIG technique as an exemplary embodiment regardless of what they say. And you want to do this with every step in the construction of the machine – you want to know what they did, why they did it, and what would happen if they acted differently because you’ll find out that someone else is going to provide evidence that you could act differently. And you want at least one claim that actually covers the exact steps they did. Because otherwise you are relying upon inductive assumption, and it is definitional that inductive logic will fail you.

            If your lucky that person beat you to the punch because the problem will be found and manifest by the office in a 103 rejection, because if it post-dates you’re going to sail through the office having not received a 112a only to have your claims invalidated by a court. Few examiners even search postdating references let alone fashion those style of rejections.

            Claims do not have to “describe what the thing is.” Examiners mess that section up all the time. The operative phrase is “distinctly claim.” You need to distinguish the invention — not describe it.

            No, attorneys simply ignore the statutory requirement to point out and particularly claim because the CAFC has not given it teeth in the past 60 years. But the Supreme Court has routinely held it to have teeth (see General Electric v Wabash, Holland Furniture v Perkins Glue, Nautilus v Biosig). The CAFC has not developed the most recent Supreme Court guidance that requires reasonable certainty, but if Wabash and Perkins Glue are any guide intentionally claiming the result of use rather than the structure posited injects an unreasonable amount of certainty into the claim scope. But I will agree with you inasmuch as that is just my interpretation of what the court will do based upon 100+ years of precedent, and the court has not actually done it. You’re correct that 112b is the least forceful of the statements I made there.

            As for “possess[ing] the full range of their functional language,” can you identify the case law regarding that. How, for example, can I show possession of the rule range of “welding A to B”?

            Most people like to cite Ariad. I prefer to cite Lizardtech, especially because its in computers and thats what I do. Although Lizardtech touches on two independent claims, the way they would be constructed in most claimsets is an independent and a dependent. There is an dependent claim limited by technique that was not infringed, and there is an independent claim that is not limited by technique that was found invalid under 112a because it was not limited by the disclosed technique. In fact, the fact that the defendant could infringe the nonlimited claim without using the technique described in the specification proved the overbreadth of the nonlimited claim and formed the basis of the 112a invalidity.

            I claim “welding A to B” and you argue that you haven’t described any “concrete limitations.” I then claim “welding A to B with TIG” and then you argue that there are thousands of ways to do that so the claim language is still abstract. I keep adding on limitations, but you keep arguing (rightfully) that there are still hundreds of different ways of doing that. We can keep going down that rabbit hole for a long time.

            Well abstraction is based upon what the claim is directed to, so this is not a great hypothetical to use when you posit that the welding is conventional. But yes, there are an infinite number of layers between something that is an abstract result and something that is a concrete means, so you can “keep going down the rabbit hole for a long time” as you deride it. That doesn’t change the fact that the hole does end – At some point you reach a scope that only encompasses materially equivalent means to the means disclosed in the spec. You know such a scope exists, because it is the scope that the claim would encompass if the claim expressly used means-plus language.

            Which brings me back to computer software – It simply is not true that the art finds all algorithms that perform the same function to be equivalent. Search schemes exist, but do you think Google finds equivalence between a search algorithm that searches a database in 400ms and one that searches it in 300ms? That is literally hundreds of millions of dollars to them. Do you think that an algorithm that averages a few percentage points of error in calculating driving time is the same as an algorithm that has, e.g., 30% error? The former is incredibly helpful, the latter is less useful than human intuition. The notion that you could invent and disclose the latter and tart it up with enough fancy legal language to make a valid claim that reaches the former is insulting to the function of patent law and unsurprisingly is not tolerated. No amount of excellent legal scrivening can turn a molehill into a mountain.

            And you guys wonder why computer claims keep getting found invalid whenever they’re tested in court. It’s because you keep having clients that come up with mediocre algorithms. And that’s okay, you can have mediocre algorithms! Embrace the fact that not everyone is Edison. But then the lawyer gets ahold of it and decides that they are going to aggrandize the mediocre algorithm into a scope where even someone who comes up with a vastly superior algorithm is an infringer that has no incentive to disclose and no ability to compete in the market. All because the lawyer came up with the brilliant idea of labelling the actual invention an “embodiment” of some overbroad functional scope or, worse still, never discloses the actual invention at all in lieu of the functional scope. Every inventor is Edison and every 20-page spec is both the first and the last word one of skill will ever need to read on a functional result. Please. It simply does. not. work. Just faithfully convey the algorithm posited and take a scope to that effect, and if that scope is not financially valuable to you then don’t take a scope at all. You’re not entitled to aggrandize your invention until its fiscally viable for you, you’re only entitled to a scope that you actually posited and disclosed the full breadth of.

            1. 5.2.2.3.2.1

              The ladders of abstraction have more than two rungs.

              It really is as simple
              As that

        2. 5.2.2.4

          To use a more physical example – If one invents the bow, the claim cannot be for a machine that fires a projectile.

          No no no – that’s indefinite.

          What type of bow? What type of string? What type of release/recoil mechanism?

          And don’t even get me started on the arrows…

          This “invented the bow” simply cannot stand.

          No Patent For You (in the Seinfeld character tones)

          :bigsmile;

          1. 5.2.2.4.1

            What type of bow? What type of string? What type of release/recoil mechanism?

            Exactly. Presumably if you just say “bent wood” there would be types of wood that would be unable to be bent in that matter.

            So what are you going to do when someone figures out how to bend the wood, claims it in a patent and you make the mistake of suing him? Now you’ve got a clear scope of enablement problem, and you better hope not all of your claims are nonlimited in the type of wood.

            Hopefully your attorney was smart enough to provide and claim an embodiment that included what you actually did, and then explained why any steps on (to use your term) the ladder of abstraction to move away from a picture claim are justified. For example – If your bow was only made with oak, you might want to try making a bow with birch. Your attorney might want to explain why it is you think the steps that allowed for oaken crafting would also apply to any other type of wood. Ditto with strings or any other features. Then, you should have a series of dependent claims that apply the inductive logic that you are using to take the picture claim into an independent claim. Then you’re protected when any of your inductive logic steps fail – Turns out birch wasn’t enabled? that’s okay I have a dependent claim that specifically mentions oak.

            Further, if all else fails you can always take one claim and make it a means plus claim and be sure to have at least some protection for what you disclosed.

            I dunno why people find this surprising – if your bow used oak but you don’t disclose that you used oak and it later turns out that someone else filed first and used birch, you wouldn’t be surprised that a 103 rejection would tank your claimset. So why would you be surprised if you don’t mention oak and it later turns out birch is unenabled? You were wrong about enabling and possessing the full scope of “a bent [non-limited type] piece of wood” just like you were wrong about it being nonobvious in the prior hypothetical.

            This problem DOES EXIST and it HAS A SOLUTION when the solution is properly employed. The correct thing to do is employ the solution, not pretend the problem does not exist.

            This “invented the bow” simply cannot stand.

            It is true that it is highly unlikely that a single person invented the bow. It’s far more likely that one person realized that if you bent a particular type of wood and used a certain type of string you could propel a stick, and other people iterated on different ways to bend different types of wood and used different strings, etc. The full scope of the bow as we define it today was absolutely the work of hundreds or thousands of people standing on each other’s shoulders. The notion that we have 999 people who take nothing for their disclosed advancement until the guy who disclosed that you can bend oak and string sheep’s guts has his exclusivity period run is a huge disincentive to the advancement of the field and antithetical to the concept of patent law.

        3. 5.2.2.5

          As someone who examines computer software, virtually everyone is claiming comparable to Morse’s Claim 8.

          I truly pity anyone whose application finds its way to your grossly inaccurate world view of a docket.

        4. 5.2.2.6

          “ It doesn’t much matter if you view the problem as one of a failure to point out and particularly claim (defining something by the result of its use does not particularly describe what the thing is), or as a failure of enablement or possession (Applicant fails to show that they possess the full range of their functional language) or as an abstraction (Applicant merely asserts an abstract goal without concrete limitations on how to arrive at the goal). The problem is unquestionably the use of functional language as part of the nonobvious feature.”

          Also referred to as “the seamless web.” All of the statutes are serving the same ultimate purpose so it’s not surprising to see this kind of merging of concepts at the margins of each statute. And just beyond “the margins” is where logic patenting definitely resides.

          1. 5.2.2.6.1

            To try to rewrite here your points and interesting conclusion so that I might better understand it, that: If you view all of the following unpatentability challenges* as intended to serve the same ultimate purpose [overclaiming?], it’s not surprising to see merging of concepts [or overlap] at the margins of each statute.
            * 1. failure to point out and particularly claim the invention. [Defining something only by the result of its use does not particularly describe what the invention is.] And/or no full spec description-possession of the invention as later claimed.
            2. failure of enablement (applicant fails to show that they possess enough examples or ranges to show the full genus claim range.)
            3. An “abstraction” (applicant merely asserts an abstract goal without any invented limitations on how to arrive at the goal). The the use of functional language as part of the nonobvious feature.

            1. 5.2.2.6.1.1

              Not sure who’s being addressed here, PM, but it looks like you get it.

              The rub with the current (and long-standing) “convention” for claiming “new” machines that carry out logic operations is that the point of novelty is the logic operation, that element is described in functional terms, and there are ZERO actual distinct and distinguishing structures described, and no working examples (neither detailed descriptions of new machines reduced to practice or even bug-free code for a specified operating system) are specified or deposited.

              In short, logic claiming in the computer context only results in a granted patent because the courts and the PTO have created exceptions at every turn. Now the “industry” is predictably addicted to this overly generous set up but at the same time certain members relentlessly complain about how “it doesn’t make sense”. Of course it doesn’t “make sense”! It’s not supposed to “make sense.” The exceptions exist to avoid the sensible result which is that abstractions like logic and algorithms are not protectable with utility patents and the context (whether in the courtroom, the classroom, a hospital or “on a computer”) are irrelevant.

            2. 5.2.2.6.1.2

              To try to rewrite here your points and interesting conclusion

              Oh, horse
              S
              H
              I
              T
              Paul — there is nothing interesting in Malcolm’s tripe that has been regurged for at least a decade.

            3. 5.2.2.6.1.3

              1. failure to point out and particularly claim the invention. [Defining something only by the result of its use does not particularly describe what the invention is.] And/or no full spec description-possession of the invention as later claimed.

              The Supreme Court has never endorsed what the CAFC calls the Written Description test. What the Supreme court did prior to 1952 was identify claims that was insufficiently precise compared to the disclosure – one example of which is talking about the result of use of the invention rather than the invention itself – and reject those as failing to point out and particularly claim the invention, and conversely identify things that were not inventions at all. So when an inventor described a combination and said that the combination was as good as animal glue, the court said that claim language that was as good as animal glue did not particularly point out the combination (as it described other glues besides those disclosed) and a claim to the function of being as good as animal glue is not an invention. In other words – saying the invention was the function was not an invention, and saying the invention was the thing described in the specification resulted in a failure of the claims to point out and particularly claim.

              The overbreadth protection afforded by the “point out and particularly claim” logic disappeared after the 52 act because the CCPA did not enforce it. But that eventually became untenable so written description was created to enforce the linkage between the disclosed embodiments and the claim scope. But a key distinction between WD and the failure to particularly claim/noninvention (or 112b/101) dichotomy is that the dichotomy would consistently be rejected using just intrinsic evidence and logic, while WD requires extrinsic evidence. People complaining that the judicial exceptions sometimes function as super 112a rejections without evidence fail to realize that that was the rule for most of patent law – the reviewer (whether it was the court or the PTO) had the authority to identify on their own that the scope was overbroad of the disclosed embodiments.

              2. failure of enablement (applicant fails to show that they possess enough examples or ranges to show the full genus claim range.)

              Failure of enablement is tricky because everyone will simply use the Morse statement – I have told you the principle and therefore one of skill would be able to apply it to any context. That is especially powerful when the invention is computer logic. It is an effort to come up with one (usually mundane) manner of achieving a result and then handwaving oneself into being the inventor of the least obvious result. Failure of enablement is easier to apply in mechanical examples with completely different principles of operation. Its very easy to understand how the bow disclosure doesn’t enable a gun. It’s less easy to understand how a first algorithm for estimating driving time doesn’t enable all algorithms for estimating driving time. It ends up being a swearing contest between experts.

              3. An “abstraction” (applicant merely asserts an abstract goal without any invented limitations on how to arrive at the goal). The the use of functional language as part of the nonobvious feature.

              A subset of abstractions are these problems here. There are totally concrete ineligible subject matter. But some subject matter is ineligible purely because it is so broad that it encompasses distinct means. Functional language as it is often employed is this latter type. That’s not to say that it ALWAYS has this problem, as one can claim a function and limit the functional achievement to a particular means (after all, that’s what means-plus claiming does).

      3. 5.2.3

        The problem with many of the anti-software patent arguments [is] that they are presented by people who do not… appreciate that their proposed changes screw[] up everything else before their sole focus is on software.

        “Screw up” in what sense? What is the evil that befalls society as a whole (not just EE patent professionals) if a more aggressive reading of Williamson v. Citrix were applied to functional recitations in software claims?

        1. 5.2.3.1

          if a more aggressive reading of Williamson v. Citrix were applied to functional recitations in software claims?

          Well obviously the only two results would be (1) it would require applicants to actually disclose an algorithm for the function that they allegedly possess, and all that writing leads to hand cramps and

          (2) It would be impossible for claim scopes to reach nonequivalent means, so competitors who came up with what a POSITA would call a distinct manner of achieving the functional result would not be an infringer.

          You can’t have (2). That would lead to, you know, market competition and stuff.

          1. 5.2.3.1.1

            Round and round we go with the “what does non-obvious logic look like?” question. The PTO has had decades to develop a system for claiming algorithms in a formal searchable manner but they’ve done squat.

            Best part: “hey, let’s let computers be inventors of do it on a computer patents” is being peddled by the same crowd of born yesterday goofballs.

            1. 5.2.3.1.1.1

              Please Pardon Potential rePeat due to Count Filter….

              Your comment is awaiting moderation.

              May 19, 2022 at 8:52 am

              What your point is – is anybody’s guess.

              I mean, your “feelings” are clear — as they have always been. But what legal point are you trying to make? That the PTO has power to set innovation protection limits to the computing arts?

              Is that what you are trying to say?

          2. 5.2.3.1.2

            Well obviously the only two results would be (1) it would require applicants to actually disclose an algorithm for the function that they allegedly possess, and all that writing leads to hand cramps
            The steps of the method are the algorithm. Computer software claims rarely recite a single step.

            However, if you want each step of the method to have its own algorithm, that isn’t the law. If it was the law, then what is to prevent one from arguing that each step in the algorithm (an algorithm is, itself, a series of steps) should each have their own algorithm. This leads us down the same kind of rabbit hole I discussed earlier. Each step in the algorithm needs its own algorithm, and each step in those algorithms need their own algorithm, and each step in those algorithms need their own algorithm and so on and so forth. Its a nice deep rabbit hole that will produce unenforceable claim language — exactly what the anti-patent crowd wants. Moreover, there is no reason why that language could be applied to:
            “welding A to B”
            let’s say I add:
            “positioning A adjacent to B;
            tacking A to B;
            filling gaps between A and B with a bead weld.”
            Well, each of those steps can be broken down into more steps and those steps can be broken down into even more steps. Again, this is a pretty deep rabbit hole.

            You can’t have (2). That would lead to, you know, market competition and stuff.
            There is a difference between market competition and reproducing the invention without infringing with only a minor variation because of an aggressive reading of Williamson v. Citrix.

            1. 5.2.3.1.2.2

              If it was the law, then what is to prevent one from arguing that each step in the algorithm (an algorithm is, itself, a series of steps) should each have their own algorithm. This leads us down the same kind of rabbit hole I discussed earlier.

              You don’t need an algorithm for conventional subject matter because of the supplemental rule. But that only applies when it is conventional subject matter. If your invention is about calculating driving time, you need to disclose an algorithm that takes the data from the previous steps and explains how it processes it to generate the output of how long it will take. You can’t rely upon the supplemental rule there because the argument is that this is novel and nonobvious to one of skill.

              And, to the extent we’ll get into it again, you can’t claim it functionally because a functional claim is overbroad of the algorithm you posited. You can’t write a bad algorithm and intentionally describe it vaguely using functional language in an effort to circumscribe good algorithms that there is no proof you ever posited.

              There is a difference between market competition and reproducing the invention without infringing with only a minor variation because of an aggressive reading of Williamson v. Citrix.

              That’s a misstatement of law. Construing something in means-plus protects the patentee from someone using equivalent means, which would protect against minor variations. But you don’t want that – you want patent protection for non-equivalent means that there is no proof that the patentee thought up. “I figured out one way of making this function happen and I am not content with minor variations on my one way, I want protection even from entirely different means so long as they achieve the same functional result.” That’s not protecting the discovery, that’s protecting the market value of the discovery: You aren’t content with your burger keeping burgers with minorly different seasonings and fat contents off the shelf, you also want hot dogs off the shelf too. When someone has a grilling need, you want to be the only legal solution.

        2. 5.2.3.2

          if a more aggressive reading of Williamson v. Citrix were applied to functional recitations in software claims?

          What is the difference between:
          (1) “a module configured to route, to one of a plurality of servers, a request based upon a flag set within the request and a workload, respectively, of each of the plurality of servers.”
          and
          (2) “a processor configured to route, to one of a plurality of servers, a request based upon a flag set within the request and a workload, respectively, of each of the plurality of servers.”
          and
          (3) “routing, to one of a plurality of servers, a request based upon a flag set within the request and a workload, respectively, of each of the plurality of servers.”

          Current case law says that (1) is MPF and (2), (3) are not. Should (2) be MPF? What about (3)? If you think (3) should be MPF, what do you think about?
          (4) “welding A to B”
          and
          (5) “A welded to B”

          If (3) is MPF, then so should (4). If (4) is MPF, then why shouldn’t (5)? For (4) or (5) to be an MPF is what I meant by “Screw up.”

          1. 5.2.3.2.1

            None of it makes sense because it based on trying to limit patents by judicial activists.

            The big problem is trash judges using their untrained prejudiced minds to determine whether structure is recited when a person having skill in the art should be making this determination.

          2. 5.2.3.2.2

            Ok, but then one only reaches the point of “screwed up” if you insist that #3 deserves a M+F construction. It seems to me that if we do not cross that line (and I certainly would resist crossing that line) we can still—consistent with a more aggressive application of Williamson—give #2 a M+F construction and nothing has “screwed up”

            1. 5.2.3.2.2.1

              we can still—consistent with a more aggressive application of Williamson—give #2 a M+F construction and nothing has “screwed up”
              So your position is to draw the line between (2) and (3)? Personally, I don’t see much of a line to be drawn there, but so be it.

              The difference between (1) and (2) is that a “module” (accordingly to the Federal Circuit) connotes no structure whereas a processor does connote some structure. A “means plus function” is construed to be ANY means capable of performing the function. When I recite a “processor” for example, I have limited the function to a particular structure. Granted, there are millions of different types of processors, but that is still limiting. When I use “a processor configured to …” I am not claiming ANY means.

              Again, the presumption is that MPF is not invoked when “means” is not used. A “more aggressive application of Williamson” removes that presumption is leaves it up to the whim of the court to determine what is MPF and what is not. We’ve seen what leaving it up to the court to determine the meaning of a phrase (e.g., “abstract idea”) leads to — an unworkable mess.

              Law should be predictable and easily-discernable lines makes that so.

              1. 5.2.3.2.2.1.1

                Personally, I don’t see much of a line to be drawn [between (2) and (3)], but so be it.

                That is the line between a machine claim and a method claim. This is hardly an arbitrary divide.

                [T]he presumption is that MPF is not invoked when “means” is not used. A “more aggressive application of Williamson” removes that presumption is leaves it up to the whim of the court to determine what is MPF and what is not.

                That would not be a “more aggressive” application of Williamson. That would be a frank repudiation of Williamson. Williamson most definitely reaffirms that §112(f) is presumed not to apply when “means” is not used. Williamson v. Citrix Online, 792 F.3d 1339, 1349 (Fed. Cir. 2015).

                Rather, an aggressive application of Williamson would mean that one scrutinizes all words that are not “means” or “step” (including “processor” and “circuit”) as if they could—in this context—be nonce words. There should be no “list” of nonce words, in other words, where one gives special scrutiny to “module” but a pass to “circuit.” Both “module” and “circuit” should be scrutinized on equal footing.

                1. That is the line between a machine claim and a method claim. This is hardly an arbitrary divide.
                  But what difference does it make for a MPF claim? This is from 35 USC 112(f): “An element in a claim for a combination may be expressed as a means or step for performing a specified function.” The “means or step” covers by machine/device and method.

                  Rather, an aggressive application of Williamson would mean that one scrutinizes all words that are not “means” or “step” (including “processor” and “circuit”) as if they could—in this context—be nonce words.
                  You mean that no one would really know for sure whether 112(f) applies or not unless they take their claims to the Federal Circuit. As I wrote earlier, the law should be predictable and easily-discernable lines makes that so. The less predictable patent law becomes, the less patent law promotes the progress of science and the useful arts. Along the same lines, the less predictable patent law becomes, the less likely inventors will avail themselves of it.

                2. But what difference does it make for a MPF claim?

                  Congress added §112(f) to deal with functional claim limitations. In the context of a method, the steps—which must be verbs—are the structure. Section 112(f) was never meant to apply to structural limitations, so §112(f)’s applicability to method claims is necessarily attenuated. By contrast, in the context of a machine, manufacture, or composition, verbs in the claims are the precise circumstance that §112(f) was intended to address.

                  [N]o one would really know for sure whether 112(f) applies or not unless they take their claims to the Federal Circuit.

                  For better or worse, that is always true of all patent claims. Meanwhile, if one wants less unpredictability along the M+F dimension, one is always free to draft one’s claim explicitly to invoke §112(f), or else avoid using verbs in your apparatus claims.

                3. By contrast, in the context of a machine, manufacture, or composition, verbs in the claims are the precise circumstance that §112(f) was intended to address.
                  “A welded to B” 112(f) or not?
                  “welding A to B” 112(f) or not?
                  “distributing coolant to an engine” 112(f) or not?
                  “a distributor configured to distribute coolant to an engine” 112(f) or not?

                  else avoid using verbs in your apparatus claims
                  Said by someone who doesn’t understand how electro-mechanical devices are claimed. Again, I’ll repeat what I wrote earlier, “… who do not understand that the claiming conventions in software are little different than the claim conventions in electrical/mechanical device”

                  Oh, and let me get this straight, language that is used as part of a method claim doesn’t get 112(f) treatment yet nearly identical language found in a machine claim does get 112(f) treatment? Is that what you are saying?

                4. “A welded to B” 112(f) or not?

                  “Welded” is an adjective, not a verb. No §112(f) treatment.

                  “a distributor configured to distribute coolant to an engine” 112(f) or not?

                  Do artisans in this field recognize “distributor” to convey a class of structures? I expect that the answer is yes, although I am not certain of that. If so, then no §112(f) treatment under even an aggressive application of Williamson.

                  Said by someone who doesn’t understand how electro-mechanical devices are claimed.

                  A fair point. I do not understand how electromechanical devices are claimed—at least not as well as you do. Still and all, there is only one statute for all art fields. If you are trying to say that there are special rules for your art field that do not apply to mine, do you care to show me where in the law I find these special rules?

                  Oh, and let me get this straight, language that is used as part of a method claim doesn’t get 112(f) treatment yet nearly identical language found in a machine claim does get 112(f) treatment? Is that what you are saying?

                  Yes. I am surprised that this is engendering pushback. To be clear, if a method claim says something like “… (2) conveying the result of step 1 to an output maximization means…” then that boldface phrase should get §112(f) treatment (and the result show not change if “means” is replaced with “module” or “circuit”).

                  “Conveying,” however, is part of the structure of the claimed method. It makes no sense to apply §112(f) to structural limitations. That is not the purpose for which §112(3) was added to the 1952 act.

                5. Greg,

                  Please stop.

                  That hole that you are in is only getting deeper with your frantic shoveling.

  7. 4

    How should all this be handled by the USPTO during ordinary prosecution?

    I think that this one is fairly easy. Williamson tells us that a claim is presumed to be M+F if it uses “means,” and presumed not to be M+F if it does not. A presumption is “only… a device for allocating the production burden.” Texas Dept. of Community Affairs v. Burdine, 450 U.S. 248, 255 n.8 (1981). That means that if the claim does not use the word “means,” then the examiner must first raise the issue of M+F and adduce evidence before the claim should be given M+F treatment.

    As you note, however, “112(f) classification… typically… narrow[s] claim scope… .” The examiner’s duty is to give the claim its broadest reasonable construction, so invoking a doctrine to narrow the claim scope makes no sense in the examination context. Therefore, the Williamson analysis is probably irrelevant to the examination context in all but the rarest of instances.

    1. 4.1

      I’m surprised, Greg, that this comment of yours has elicited not one response at all, up to now. For me, it seems eminently helpful and refreshingly pragmatic. Perhaps the reason for the absence of comments about it is that nobody sees any reason to argue with it.

      No other jurisdiction has a statutory provision about claims in M+f form. No other jurisdiction makes such a meal of claims written in M+F form. All other jurisdictions manage just fine. Greg, your comment strikes me as the voice of sanity.

      1. 4.1.1

        Perhaps the reason for the absence of comments about it is that nobody sees any reason to argue with it.

        Sure – let’s go with that. After all, that’s why you agree with my points, eh?

    2. 4.2

      >The examiner’s duty is to give the claim its broadest reasonable construction,

      A few points:

      1) the Examiner is also supposed to examine for 112(b). It’s a little hard to see how a M+F claim element can be “definite” if there is no structure. Put differently, 112(b) should create a presumption that structure exists, therefore the C&C standard should apply.

      2) Even wrt to 103, it’s still the broadest “reasonable” construction. It’s not clear how an Examiner could perform that analysis wrt M+F claims w/o looking at the disclosed structure.

      3) The big problem wrt Williamson is the added uncertainty (on top of ‘equivalents thereof’ uncertainty). The old rule, imho, was better b/c at least you didn’t have to litigate “M+F or not” every time.

      3.1) The added uncertainty cuts both ways. For example, you should account for Williamson in invalidity opinions.
      And, ironically, I’d argue it’s pro-Troll b/c uncertainty and litigation costs their biggest weapons.

      1. 4.2.1

        It’s not clear how an Examiner could perform that analysis wrt M+F claims w/o looking at the disclosed structure.

        Two brief responses:

        (1) Obviously, the examiner should read the whole spec (rarely though this may happen in actual practice), so the examiner should be acquainted with disclosed structure in any event.

        (2) None of what is said in #4 above applies if the applicant actually writes a “means for” claim. The reasoning of #4 only applies in a world in which the applicant has not written “means for,” but we are concerned that—under Williamson—M+F construction may still apply. In those circumstances, if the broader reading is achieved with a non-M+F construction, then the examiner should apply a non-M+F construction. It really cannot be unreasonable not to apply a non-M+F construction where the applicant has not used “means” or “step.”

        1. 4.2.1.1

          It really cannot be…

          Actually Greg – you have inserted a logical fallacy as the point of the discussion is that YES – it CAN “really be”

          It is simply not the “all in”‘that you want.

          As to the “all out” that I would want, well, one of our views actually accords with what Congress has done (sans the patent profanity of the courts).

  8. 3

    Worth repeating this general observation, given as the good professor references his March article as describing a spectrum….

    link to patentlyo.com

    Not sure why Vast Middle Ground has fallen “out of favor” in any narrative of spectrum, but that phrase clearly has a much bigger impact that many of those opposing patents for the computing arts may desire.

  9. 2

    What a mess this has created. I actually just spent one hour yesterday going over all this case law in an internal CLE.

    I’ll say it again. The CAFC is about judicial activists trying to limit patents and not about trying to create a workable coherent body of law. The judges on the CAFC–by and large–are trash.

    1. 2.1

      I think it’s the laws that are trash. The current law splits “everything” into MPF or not. If you meet some set of rules (decided by the court), you fall into MPF; if you don’t meet those rules, you’re outside of MPF.

      The problem is that many inventions fall outside of MPF but have to be claimed functionally. Take a memory controller. You can’t claim the actual hardware, because modern memory controllers have hundreds or thousands of gates. So, you have to claim the function, eg, as such:

      A memory controller configured to perform:
      function 1;
      function 2;

      function x.

      Ostensibly, this isn’t MPF, but it really is. It’s claiming the function with only a limited reliance on the “structure”.

      I personally can’t see a difference with:
      A computer system configured to perform:
      function 1;
      function 2;

      function x.

      Or, “A computer system with a processor, memory, and software, where the processor, upon execution of the software, causes the computer system to perform: function 1;…”

      There is an artifice in MPF case law that allows a general purpose computer to have a “function” that’s described by an algorithm.

      But these types of artifices and the fact that many inventions have to be claimed using functional language cause the court to go through machinations to “fit” language into these poorly-defined buckets. Those buckets are created by the structure of laws we currently have.

      Change the law, and you could prevent some of these machinations.

      In my opinion, the laws should be rewritten to address this (and the debacle of Alice/101).

      1. 2.1.1

        Your “ with only a limited reliance on” is enough to fall outside of the PURELY functional notion that the Law IS.

        In other words, you have entered the realm of Vast Middle Ground.

        That the computing arts ARE affected oft times appears to be the “real” driver, while the fact of the matter is that this affects all types of innovation arts (yes, even baby seats for cars).

        Further, as the Law (as is) writes this as an optional choice by the applicant, it should be abundantly clear that TO invoke 112(f), an applicant KNOWS to use the ‘magic words.’

        As they say, this is not rocket science (those wanting otherwise should be evaluated for just WHY they want otherwise).

      2. 2.1.2

        So, you have to claim the function, eg, as such:

        But these types of artifices and the fact that many inventions have to be claimed using functional language cause the court to go through machinations to “fit” language into these poorly-defined buckets. Those buckets are created by the structure of laws we currently have.

        Yeah it’s too bad there’s not a language or something that would concretely identify the commands followed by the device in order to achieve the functional result.

        It’s obviously amenable to description, so when you say they “have to be claimed using functional language” what you really mean is that you believe that the applicant should be trusted to come up with their own vaguely descriptive language of the result to set the scope of their contribution, rather than artisan being presented with and being able to distinguish between equivalent and non-equivalent steps. That’s not so much a “have to” as a “being to lazy to do otherwise.”

      3. 2.1.3

        “The problem is that many inventions fall outside of MPF but have to be claimed functionally.”

        Yes, your clients haven’t invented anything. I’m being generous here. Please put down the patent crack pipe.

        1. 2.1.3.1

          LOL – as usual Malcolm, you are invited to entirely abstain from anything that you do not feel was invented.

          Was that the reason for your just over a year absence?

    2. 2.2

      Night Writer thanks for your irrelevant ad hominem attack on judges as “trash.”

      When an apparatus claim recites a “[INSERT FUNCTION] element” but doesn’t otherwise specify what such an element actually IS, it is many times entirely appropriate to limit the claimed “element” to whatever structure is disclosed (plus equivalents). Doing so is also “coherent.”

      It seems that whenever you don’t like the state of the law you (1) reach for an ad hominem attack and (2) mischaracterize the current state of the law (rather than characterize it faithfully but then offer some principled, substantive reason for why the state of the law should be different).

      Many of the posters on this blog consistently and repeatedly argue in this fashion.

      1. 2.2.1

        Litig80r: There was no substance to your post. The issue is what does it mean to recite sufficient “structure”. The CAFC’s determination, for example, that module does not cite structure is manifestly ridiculous. Go to elance or any site where one can contract out software and the term module is used literally 100’s of thousands of times and has a specific meaning in the specification of the code desired.

        That is just one example. I’ve given many, many examples of why the judges are trash in that they decide cases based on their prejudices against patents rather than applying the law.

        You ignore all the examples I’ve given and play the loss of memory game where each time you claim I’ve never cited any evidence.

        Please try harder. You have proven yourself to be totally worthless as anything but a cheerleader for weaker patents.

        I could spend a hour and go through the case law on which words have been deemed by the trash judges not to recite enough structure but then you would forget it by the next post. The other issue is that the trash judges just w a n k o f f in their chambers and decide these issues on their own as if they are experts.

      2. 2.2.2

        It seems that you Litig8or (1) just forget everything ever posted before. (2) Defend the CAFC no matter how ridiculous their case law is. (3) Make ridiculous personal attacks with no evidence. (Typical of your sort where you accuse me of what you do.)

        So please just don’t respond to my posts. You are as worthless as the trash judges.

      3. 2.2.3

        “then offer some principled, substantive reason for why the state of the law should be different”

        Are we reading the same blog? Night Wiper with a principled substantive reason for … doing what exactly?

        LOL

        1. 2.2.3.1

          Please Pardon Potential rePeat due to Count Filter…

          Your comment is awaiting moderation.

          May 19, 2022 at 8:56 am

          ^^^ Iron Rule of Woke Projection

          also known as
          A
          O
          O
          T
          W
          M
          D

        2. 2.2.3.2

          MM, do you ever have any substance to your posts?

          Read through the comments. There are principled reasons throughout the comments explaining why the CAFC’s use of 112(f) to invalidate claims is clearly wrong and judicial activism.

          1. 2.2.3.2.1

            MM, do you ever have any substance to your posts?

            That’s an easy question – the answer is No.

  10. 1

    May a claim term that connotes structure to one of skill in the art nonetheless require means-plus-function treatment because the structure it connotes is not sufficient for performing the functions the structure is recited as performing[?]

    This is the real nub of the dispute. There is no doubt that “circuit” conveys some structure, but does it convey structure sufficient for performing whatever is the functional recitation conjoined to “circuit” in the claim (e.g., “output maximizing circuit” or some such)?

    I have yet to see a case where the CAFC considered that “circuit” was not enough, but surely there are circuits that are capable of “maximizing output” (or whatnot), and circuits that are not. Therefore, it seems clear to me that “circuit” by itself is not sufficient structure to accomplish the recited function, and ought therefore to be given a §112(f) construction when it is employed in contexts like “output maximizing circuit” or the like.

    1. 1.1

      If all that you have (physically) to perform a function, then you logically have enough structure to perform the function.

      You might me aiming for something other than “sufficient structure.”

    2. 1.2

      Hi Greg,
      I do not think that parsing a phrase in a claim the way you did is just. In your example, you are separating “circuit” from “output maximizing.” Then you find that “circuit” does not provide enough structure, and “output maximizing” is functional. Divide and conquer. But paradoxically, you admit that there are likely “output maximizing circuits” available.

      You could in the same way find that a “link capable of pulling a trailer” could be indefinite. Link has structure, but not all links can pull a trailer. Pulling a trailer is functional. The term could be indefinite depending on the level of disclosure. Nevertheless, I am convinced that “links capable of pulling a trailer” are commercially available, and that a normal engineer is capable of understanding what this group of words refers to.

      1. 1.2.1

        In your example, you are separating “circuit” from “output maximizing.”

        Sure. If I write “output maximizing means,” then the only way to do the analysis is to separate “output maximizing” from “means.” Therefore, I approach “output maximizing circuit” in the same manner. If you are telling me that “circuit” works very differently than “means” here, do you care to unpack that a bit? I am not clear why we should treat the putative nonce word differently than we would treat the bare recitation of “means” when we are trying to decide—in the first instance—whether §112(f) applies or not.

        1. 1.2.1.2

          Yes, Greg, this is exactly what I did. If “output maximizing circuits” are known circuits, then those circuits are what the phrase refers to in the claim, in the same way as “screws” in a claim refer to screws. In contrast, if a person to which the invention pertains cannot identify what “output maximizing circuits” are, then you need to look at the term in the specification.

          1. 1.2.1.2.1

            If “output maximizing circuits” are known circuits…

            Got it. Thanks, that helps.

            I definitely agree that if the functional recitation of “output maximizing circuit” indisputably corresponds to a definite class of circuit structures (in the same way that the functional recitation “screw driver” corresponds to a definite class of tool structures), then it would be unfair to apply a M+F construction to “circuit” in that context.

            It is not clear to me, however, that this happens to be the case in most “****** circuit” claim construction disputes. Rather, it seems to me that in all those cases the plaintiff’s expert says “everyone knows the structure of this class of circuits,” while the defendant’s expert says the opposite. There never seems to be evidence for these disputes beyond the bare say-so of the experts. What is the fact finder supposed to do with that?

            If one required the patentee to list structural features of “******* circuits” in the spec, the fact finder could compare the accused device against those pre-specified features listed in the patent text. That is reasonably objective, and a deal more fair than just saying “I liked the defendant’s expert’s smile” when deciding these things.

            1. 1.2.1.2.1.1

              If one required the patentee to list structural features of “******* [__] in the spec,

              And BOOM – just like that Greg (having previously admitted his lack of knowledge in the art) steps directly into Wt’s warning of what NOT to do.

              I am more than certain that the irony is lost on Greg.

        2. 1.2.1.3

          Everything is a “putative nonce,” because….

          “Words.”

          (it’s turtles all the way down)

    3. 1.3

      >>>ut surely there are circuits that are capable of “maximizing output” (or whatnot), and circuits that are not.

      Not the way it is analyzed. Frankly that is silly.

    4. 1.4

      Exactly. What is the relevant distinction between an output maximizing circuit and a means for maximizing output? In both cases, the defining characteristic is the ability to perform the function of maximizing output (unless one gives the term “circuit’ undue structural significance). Although this ship has sailed, it would be better to police this under lack of written description. Assuming the specification describes one output maximizing circuit, it seems a stretch to say that a claim should cover all possible circuits (past, present, and future) that perform the function of maximizing output. Permitting claims that broad would allow people to claim an abstract collection of functions lacking any structural significance other than a passing reference to arbitrary structure like a general-purpose computer.

      1. 1.4.1

        What exactly IS that structure of a “general purpose computer?”

        If you are going to push for “exactness,” you better provide that same level of detail so that inventors may properly distinguish.

      2. 1.4.2

        NSII, you can also ask what is the difference between a thermometer and means for measuring the temperature. Same thing right, so let’s look at the specs what type of thermometer is disclosed. A thermocouple with appropriate electronics. All right then, the claim cannot cover a thermistor, because allowing that would make the claim too broad.
        I think your concern of overreaching claim scope would be better adressed by enablement. Nowadays, it is ridiculous to ask applicants to list all the known ways to provide a thermometer so that this term is not interpreted too narrowly.

        1. 1.4.2.1

          Why would it be ridiculous to require a list? What about lists is so clearly ridiculous? To my mind, lists are tedious, but not actually difficult. If they simplify claim construction, the cost of added work for the patentee is probably worth the benefit of added clarity to the affected public.

          1. 1.4.2.1.1

            Because you cannot patent ideas, the goal of patent prosecution is to get as close as possible to patenting an idea without stepping over the line. A list would might take all of the imagination out of a patent claim, and require it be something more definite.

          2. 1.4.2.1.2

            Being excessively tedious is a solid argument. In some technologies would require thousands of pages of disclosure and impose immense cost, especially on small inventors. In addition it creates a substantial risk that the applicant will overlook an embodiment and thereby grant a competitor a free license. Such a patent system would fail at “securing to inventors the exclusive right to their discoveries” and thus would discouraged patenting and the associated “progress in the useful arts”.

            The invention should be disclosed sufficiently for one of ordinary skill to understand and practice it. Any further burden on the applicant should be minimized.

            1. 1.4.2.1.2.1

              In some technologies [it] would… impose immense cost…

              Would it really. Once one had such a list, it could be cut-&-pasted from app to app. Meanwhile, §112(f) promises named embodiments and equivalents. The longer your list of named embodiments, the broader the reach of those equivalents. I think that a list rule would be both less costly and less potentially eviscerating than you are imagining.

              The real loser from such a rule would not be the inventors, but rather the patent professionals. Such a rule would sharpen the contrasts between the attorneys who know their stuff and the ones who are merely phoning it in. Clients would suddenly have a much clearer picture who is really worth their salt, and some attorneys do not welcome what would be revealed.

              1. 1.4.2.1.2.1.1

                cf
                Once one had such a list, it could be cut-&-pasted from app to app

                with
                Such a rule would sharpen the contrasts between the attorneys who know their stuff and the ones who are merely phoning it in

                As if merely cutting and pasting is somehow differentiable from phoning it in.

                This merely exhibits that Greg is — yet again — speaking from his arse.

    5. 1.5

      Helpful thread. I was frustrated during the debates over Tillis’ proposed amendment to §112 that would have applied §112(f) to not only nonce words, but all functional terms. Inventors would have been limited to picture claims. That seems to be where this petition is headed. Outside of pharma, functional terms are unavoidable – “link”, “hinge”, “fastener”, “regulator”, “low pass filter”… You will never write a claim that approaches the full scope of your invention without functional terms.

      The infringers and their supporters like Lemley and Tillis would limit us to only the embodiments we can list. Then the incumbent can have their engineers reduce our invention by replacing one of the elements in our claim with a structure we neglected to list. The corporations and their lawyers actually believe that is pure innovation – reading a patent and getting around it. Eliminating functional terms makes that job very easy. They can learn from our disclosure – and frequently from our prototypes and commercial products – and then make their own tweaked version without taking a license.

      1. 1.5.1

        > limit us to only the embodiments we can list.

        Which, obviously, will greatly increase the already punishing cost of getting a patent.

        Worse, it will hit the pro se community particularly hard. The patent bar can lessen the sting with a combination of boilerplate + automated term-expander / thesaurus software.

        1. 1.5.1.1

          [I]t will hit the pro se community particularly hard.

          That is a feature, not a bug. Anything that dissuades solo “inventors” from involving themselves with the patent system is doing a favor to all parties concerned.

          1. 1.5.1.1.1

            Could Greg be any more — at the same time — tone deaf and willfully 1gn0rant of how Efficient Infringement aims to ruin the innovation protection system of the US?

            His Big Pharma bias is on (the rather odd) proud display – and this is certainly NOT a feature, but a bug.

    6. 1.6

      “the structure it connotes”

      The term “it” in the above phrase is standing in for “claim term” and we are talking about claiming LOGIC for “processing” data..

      There is no “connotation” of “structure”. It’s ones and zeros and the “structure” depends on the “language” which (surprise!) is never specified in terms that would lead to unambiguous structure. Hence the ridiculous term “connotes.”

      1. 1.6.1

        Your lack of understanding of the computing arts is well known, Malcolm.

        Your “point” here is easily disabused with the Grand Hall Experiment.

        Reality though betrays you — computers differently configured are in fact different.

        How many different ways do you need this put to you? The televsion show House was just not correct, two groups of three resistors (series versus parallel), “soft” “firm” and “hard” are mere design choices of wares?

        Ridiculous? That would be you.

Comments are closed.