Precedential BPAI Opinion Rejects Functional Claim Elements as Indefinite and Not Enabled

logoEx parte Rodriguez, 08-0693 (BPAI 2009) (Precedential) fd080693.pdf 

In its first precedential decision since February 2009, an expanded BPAI panel has applied the Federal Circuit's 2008 Aristocrat decision in rejecting applicant's means-plus-function (MPF) claims as indefinite for failing to provide any corresponding structures in the specification other than a general purpose computer. From Aristocrat:

For a patentee to claim a means for performing a particular function and then to disclose only a general purpose computer as the structure designed to perform that function amounts to pure functional claiming. Because general purpose computers can be programmed to perform very different tasks in very different ways, simply disclosing a computer as the structure designated to perform a particular function does not limit the scope of the claim to "the corresponding structure, material, or acts" that perform the function, as required by section 112 paragraph 6.

In Aristocrat, the Federal Circuit made clear that reference to "appropriate programing" was not sufficient to transform a general purpose computer into a specific structure for 112p6 purposes. Rather, the court indicated that an algorithm for operating the computer should also be provided in the specification.

Here, the Rodriguez claim 10 included several MPF elements such as "means for generating a random system configuration file." However, the specification did not include an algorithm for generating the file apart from a statement that "[a]ppropriate software coding can readily be prepared by skilled programmers." Based on that failure the claim was found invalid as indefinite.

The Appellants have failed to disclose any algorithm, and thus have failed to adequately describe sufficient structure, for performing the functions recited in the means elements contained in claim 10 so as to render the claim definite. Accordingly, claim 10 is unpatentable under 35 U.S.C. 112, second paragraph, as indefinite. Aristocrat, 521 F.3d at 1333.

The remaining claims did not use standard MPF claim language. However, the Board found that the remaining claims all included elements that were properly interpreted as MPF because they claimed functional elements without any "noun denoting structure." (Citing Lighting World & Linear Tech). The claim elements in question include a "system configuration generator," "system builder," and "simulation verification environment." (et al.). These are the same functional elements found in claim 10 – and thus were found indefinite for failing to provide any corresponding structure in the specification.

In many ways, this MPF structure requirement runs parallel to the written description doctrine. It may well be that one of ordinary skill in the art could easily generate a random system configuration file without undue experimentation. That potential fact is irrelevant for MPF definiteness analysis. Rather, the question for definiteness of a MPF claim is whether a corresponding structure is described in the specification.

The Board did also rejected the claims for lack of enablement under Miyazaki because the functional claim language was not "commensurate in scope with an enabling disclosure." As the use of MPF claims has dropped, this third portion of the opinion may be the most important.

Notes:

  • The patent application is owned by LSI and entitled "Automated random verification of complex and structurally-variable systems."
  • Rejected Claim 1 is listed below. 

  1. 1. An apparatus comprising: a system configuration generator configured to generate a random system configuration file of a structurally variable and complex system; a system builder configured to build a system level netlist in response to said random system configuration file; and a simulation verification environment configured to verify said structurally variable and complex system in response to said system level netlist, wherein said simulation verification environment is configured to provide automatic random verification of said structurally variable and complex system in response to said random system configuration file.

97 thoughts on “Precedential BPAI Opinion Rejects Functional Claim Elements as Indefinite and Not Enabled

  1. “BTW – Does Google get you into the pay only databases of knowledge, or are those “genuinely UN-published” works?”

    It does if you pick up sufficient haxoring skillz to get into all “pay only” databases. Though I doubt Nat has those skills.

  2. Nathanael,

    Are you done crapping all over the various threads? Do you feel better? It’s amazing how wrong you are – did you google your legal knowledge? Find it at groklaw? Your noise to signal ratio is off the charts.

    Your post immediately above has so many errors in legal logic that it would be best simply expunged. Have you heard of a case called KSR? You do know that PHOSITA doesn’t actually search, do you not? Why are you conflating PHOSITA with a real person in the third paragraph? BTW – Does Google get you into the pay only databases of knowledge, or are those “genuinely UN-published” works?

  3. “PHOSITA is a legal construct with a double edge sword who has ALL knowledge of pre-existing work, not only in the art field of immediate note, but ANY art field that may share an answer to the questions being pursued.”

    This is actually dead wrong, and there are numerous cases explaining that the places where and extent to which the PHOSITA searches for prior art depends largely on the field.

    More importantly, in the fields of computing, a real person can effectively know all pre-existing, genuinely published work, by using Google. A reference to a well-known unit available “off the shelf” is certainly fine and needs no further elaboration, but that’s not the sort of stuff we’re looking at in this application.

  4. “The settled law is that describing the functionality and not the specific algorithm for implementing an element is generally sufficient to enable one of ordinary skill to write the software to implement the function. ”

    If that’s the settled law, it’s simply dead wrong on the facts, and if I ever got involved in such a case, I’d demand a jury trial on the question of whether this false statement was true, and I’d bring in expert witnesses.

    Written as a computer programmer. There’s a vast difference between saying “a system which presents X” and giving an actual flowchart or pseudocode algorithm. The former is defining a problem, the thing you *want* solved; the latter is actually presenting an instance of a solution.

  5. @Noise above Law

    I hear you. You may be right about the law for all I know. I have no reason to mistrust you.

    Then don’t come to us actual programmers saying patents bring disclosure. If the law disconnects enablement from real life, any real benefit of disclosure is reduced to a legal fiction that has no bearing on real life.

  6. “Then you’re two steps ahead of most of the stuff I examine ;)”

    That’s good to know, plurality. I guess it’s time to raise my rates again. :)

    What do you think, Ed?

  7. PHOSITA is not real life.

    PHOSITA is a legal construct with a double edge sword who has ALL knowledge of pre-existing work, not only in the art field of immediate note, but ANY art field that may share an answer to the questions being pursued.

    The double edge sword of PHOSITA cuts from the power given to Obviousness on the forward edge and the power given to applicants on the trailing edge to not have to (in fact recommended NOT to) include in applications the level of detail that PHOSITA can come up with, with all its power of experimentation (but not undue experimentation) and creativity (thanks KSR).

    PoIR – don’t conflate real life with real Law. There is a reason why others gently admonish people to have a foot in each world – technical and legal. you would be wise to heed such advice.

  8. From Red Monkey


    Even if an algorithm were disclosed, each step in an algorithm is functional, i.e. each step is essentially “doing X”.

    For example, suppose support for “means for generating a random system config file” amounted to a flowchart saying “generate a random number and retrieve config file corresponding to that random number”. Would that satisfy Aristocrat? After all, those steps are just as functional as the single step they purport to support.

    You could of course break these steps down further, but ultimately, it seems like these atomic steps are just as functional as the more complex step that they built.

    The question really is in how much detail must an algorithmic step be broken down before it can be regarded as supported under Aristrocrat?

    As usual you hit issues on the nail.

    Software development starts with writing a functional description of what is to be done. This description lacks details. The actual development is a process of progressively fleshing out the details with more atomic functional elements. At some point you are atomic enough to gain the ability to write actual code. This is when the experimentation with the software starts, when you have code you can run.

    So if the requirement is to enable a PHOSITA to recreate the invention without undue experimentation, how much detail do you need to provide? Programmers normally want to see details that translate directly to actual code. Anything less and the amount of experimentation will be about the same as if you start from scratch.

    This may not be the law but this is how it works out in real life.

  9. The withdrawal of the rejections of claims found to be indefinite is understandable, but the withdrawal of claims found to be merely unsupported by the specification, namely claims 11-18, makes far less sense. Further the withdrawal seems to be in conflict with MPEP 2143.03 (II) which requires applying art to limitations which allegedly do not find support in the claims.

  10. “If you claim an apparatus functionally, with no claimed structure, and it’s possible that every “unit” or “module” listed could be software, then clearly, the claim encompasses at least one embodiment that is nothing more than software.”

    But I don’t do that. And I generally avoid the words “unit” and “module” for at least those reasons.

  11. BigGuy: How is that nonsense? If you claim an apparatus functionally, with no claimed structure, and it’s possible that every “unit” or “module” listed could be software, then clearly, the claim encompasses at least one embodiment that is nothing more than software.

  12. “BigGuy, let’s agree to disagree on this one. At least two of us here have been getting a lot of this from the patent office recently to know how the patent office interprets the section.”

    I’m surprised I haven’t seen this particular flavor yet. I’m still getting the “your apparatus might use software in some embodiments, so it’s software per se” nonsense.

  13. BigGuy, let’s agree to disagree on this one. At least two of us here have been getting a lot of this from the patent office recently to know how the patent office interprets the section.

    Perhaps your interpretation would be a good rejoinder though.

  14. You know, this doesn’t seem to change much to me. I have been told since I was a whipper snapper that if you are trying to claim a method and especially software, you’d better have a t least a flow diagram. I don’t do much in software except signal processing so I may be missing subtleties, but has anything really changed?

  15. Speaking of the BPAI – has anyone checked if the number of dissents and concurrences has varied since they changed the credit metrics for them? Seems like I never see a BPAI decision with a dissent these days (not that a dissent is worth all that much to Applicant anyway)

  16. “The two sections conflict. If a structure is defined in terms of what it does, it is not limited to a particular structure. One section says give it no weight. The other requires that it be evaluated and considered.”

    They only conflict if you read way too much into “particular” structure in 2111.04. That’s not what Section 2111.04 is about. In a claim that reads “… a processor configured to do A, B, and C, whereby the processor’s user can access porn more quickly,” the “whereby” clause is not limiting. That’s all that 2111.04 is saying.

  17. The two sections conflict. If a structure is defined in terms of what it does, it is not limited to a particular structure. One section says give it no weight. The other requires that it be evaluated and considered.

    “Computer adapted to perform function A” would be anticipated by a generic computer under 2111.04; but would be given full patentability weigh under 2173.05g.

    Both sections rely on Fed. Cir. decisions.

  18. Here is the other section on functional claiming:

    2173.05(g) Functional Limitations [R-3]

    A functional limitation is an attempt to define something by what it does, rather than by what it is (e.g., as evidenced by its specific structure or specific ingredients). There is nothing inherently wrong with defining some part of an invention in functional terms. Functional language does not, in and of itself, render a claim improper. In re Swinehart, 439 F.2d 210, 169 USPQ 226 (CCPA 1971).

    A functional limitation must be evaluated and considered, just like any other limitation of the claim, for what it fairly conveys to a person of ordinary skill in the pertinent art in the context in which it is used. A functional limitation is often used in association with an element, ingredient, or step of a process to define a particular capability or purpose that is served by the recited element, ingredient or step. >In Innova/Pure Water Inc. v. Safari Water Filtration Sys. Inc., 381 F.3d 1111, 1117-20, 72 USPQ2d 1001, 1006-08 (Fed. Cir. 2004), the court noted that the claim term “operatively connected” is “a general descriptive claim term frequently used in patent drafting to reflect a functional relationship between claimed components,” that is, the term “means the claimed components must be connected in a way to perform a designated function.” “In the absence of modifiers, general descriptive terms are typically construed as having their full meaning.” Id. at 1118, 72 USPQ2d at 1006. In the patent claim at issue, “subject to any clear and unmistakable disavowal of claim scope, the term ‘operatively connected’ takes the full breath of its ordinary meaning, i.e., ‘said tube [is] operatively connected to said cap’ when the tube and cap are arranged in a manner capable of performing the function of filtering.” Id. at 1120, 72 USPQ2d at 1008.<

  19. Znutar, thanks. Here is the section of the MPEP:

    2111.04 “Adapted to,” “Adapted for,” “Wherein,” and “Whereby” Clauses [R-3]

    Claim scope is not limited by claim language that suggests or makes optional but does not require steps to be performed, or by claim language that does not limit a claim to a particular structure. However, examples of claim language, although not exhaustive, that may raise a question as to the limiting effect of the language in a claim are:

    (A) “adapted to” or “adapted for” clauses;

    (B) “wherein” clauses; and

    (C) “whereby” clauses.

    The determination of whether each of these clauses is a limitation in a claim depends on the specific facts of the case. In Hoffer v. Microsoft Corp., 405 F.3d 1326, 1329, 74 USPQ2d 1481, 1483 (Fed. Cir. 2005), the court held that when a “‘whereby’ clause states a condition that is material to patentability, it cannot be ignored in order to change the substance of the invention.” Id. However, the court noted (quoting Minton v. Nat’l Ass’n of Securities Dealers, Inc., 336 F.3d 1373, 1381, 67 USPQ2d 1614, 1620 (Fed. Cir. 2003)) that a “‘whereby clause in a method claim is not given weight when it simply expresses the intended result of a process step positively recited.'” Id.<

  20. “I have had a number of cases recently where the office has state that they will give no weight to “configured to” and the like.”

    I get this in non-software cases, and it’s still applied like a per se rule. This comes out of 2111.04 about whereby clauses and the like, and explains that such language “may raise a question as to the limiting effect of the language” BUT “The determination of whether each of these clauses is a limitation in a claim depends on the specific facts of the case” citing Hoffer. It’s not a per se rule. OMG – it depends on the specific facts of the case! I’d appreciate maybe a half-sentence, something to hold the divining rods over, as evidence of that fact-specific determination.

  21. To finish my thought on that… while it may be from a certain perspective to be dancing the head of a pin, I think it’s important to know what will or will not pass muster with the PTO and courts. Otherwise, an otherwise perfectly good patent might get shut down for form over substance.

  22. “I have had a number of cases recently where the office has state that they will give no weight to “configured to” and the like.”

    Sometimes it might be proper for a rejection under 112/2 to clear up the issue. In that case it usually just takes an affirmative statement in response that the functional portions are positively claim limitations.

    “dancing on the head of a needle”

    Hah. I hadn’t heard that one, so I had to look it up. Funny: “The most notable work was that of Thomas Aquinas who wrote earnestly on such topics as “Whether Christ was not an hermaphrodite [and] whether there are excrements in Paradise.””

    Ultimately, every case is judged on a case-by-case basis, and every examiner is different. Personally, I think the examiner should treat a claim in the same manner that a court would (aside from claim interpretation of course). In this case, the primary issue of 112/6 is a very narrow issue where the claim interpretation really shouldn’t differ much between the PTO and courts. But the stuff regarding the Beauregard and method claims made this one interesting in my opinion.

  23. anon, you are joking, and here is why – there is a non-insignificant chance that software/computer program patents will be dead in six months time.

  24. This and the way it looks like Bilski is going are great news for me. I’m a patent attorney with a BS in Comp. Sci.

    Looks like the firms are going to have to hire someone who actually knows that a constructor isn’t something from a buiding site and a binary tree doesn’t grow in a garden. Otherwise, they’re going to get their patents rejected.

    In all seriousness, these ruling are good. The PTO has been letting way too many “computer programs” through without a sufficient detailed description.

  25. Jules, I have had a number of cases recently where the office has state that they will give no weight to “configured to” and the like. Such statement fly in the face of the quoted MPEP section.

    But if considered, then “computer configured to” is essentially wholly functional, but would be construed “narrowly” in that only one fully enabled emobodiment would have to be disclosed.

    We are dancing on the head of a needle here.

  26. I recall that LSI’s in-house (who has since moved on or been replaced) would tout his marvelous software program that was keeping their prosecution costs so low and how he had picked out an ace team of low-budget solo’s – such as the one who wrote and appealed this matter — as part of his phenomenal money-saving prosecution strategy. But scratch and sniff that app: the scrivener evidently thought it would be “limiting” to provide any real-world examples on what was actually invented and instead stuck to the most hateful and vague abstractions imaginable. So now we get this correspondingly obnoxious precedent as a result. Perhaps the scrivener may have been up against a rather medieval cap but still — that app was a real stinker.

  27. there is a case that says that functional claiming is enabled in the computer arts.

    As a per se matter? Wow. I’d like to see that case.

  28. Ned, the claim element does appear to be be structural, but I don’t believe the function should be ignored. Hierarchy’s post at 3:52 addressed this with MPEP 2173.05(g). I’m not sure why you conclude that applicants will have to use means+function language. To me it appears things are moving in the opposite direction, that is unless software people start including code with their patent apps more often.

  29. Sorry, hierarchy. You don’t get to assume that “software = algorithm” and then shoot down what I said.

    If you merely recite (as Jules noted) that a processor and software may perform a function, you lose. Period. That ain’t sufficient corresponding structure.

    And if you all you disclose is “computer executable instructions” may be used to perform a function, guess what? That ain’t gonna be sufficient corresponding structure either.

  30. Jules, looking at your post just above my question, it appears that the BPAI would hold the element,

    computer configured to preform function A,

    to be structure. But, if the Office ignores the “configured to” language as it now seems to, the claim will in fact read on the prior art. Therefore, logically it would seem that such a claim must be functional.

    All this does for me is to continue to believe that the way this will work out in the end is that applicants will have to use means or step plus function.

  31. Just a question, how would the Board handle this element:

    a computer configured to perform function A?

    Computer is a well known apparatus, but the claim element appears to me to be wholly functional.

  32. Hierarchy,

    “Aristocrat held that an algorithm on a processor is structure for 112/6 … Would you say that processor+algorithm can be 112/6 structure?”

    True and yes. But to be clear, the algorithm itself must be disclosed, not just the result/”function”.

    “Would you say that “software running a processor ain’t corresponding structure” is incorrect, if we assume “algorithm”=”software”?”

    Again, I agree with what you say, but only if by “software” you mean “particular algorithm” and not simply result/”function”.

    I stick to what I said that general processors and general software are not sufficient dislosure for “structure” and “acts” respectively in accordance with 112/6. For example, you can’t just have a form paragraph that says “all means-for limitations can be processor+software”. They have to recite the specific algorithm for each result/”function”. In other words, software apps have to start disclosing specific algorithms or straight up code.

    I don’t know if you noticed, but the MPEP was updated recently to include a slew of rejections for form paragraphs relating to 112/6, mostly under 112/1,2 for insufficient disclosure and not clearly linking the “means for” limitations to “structure, material, or acts” in the spec.

  33. I haven’t seen it referenced here, but I know there is a case that says that functional claiming is enabled in the computer arts.

  34. And note, if the patent owner is not protected by Section 112, p. 6, a broader range of art is applicable. Reexaminations will be filed and litigations stayed to eliminate the more conservative protection of the courts.

  35. Malcolm, from your case citation: “The court’s job is to determine if the applicant intended to invoke 112P6, regardless of any “magic words.”

    That might be the job of a court, but it is not the job of the patent office using BRI. The examiner should reject under Halliburton (112, p.1) and let the applicant amend.

  36. You mean kind of like what beauregard claims lack?

    bingo! if software running on a processor ain’t corresponding structure (and it ain’t according to Aristocrat), reciting “instructions” embodied on a computer readable medium isn’t corresponding structure either.

    How could it be? There is no necessary physical correspondence whatsoever between a computer-implementable “function” and, e.g., the pattern of computer-readable pits on a disk. At least if you describe a “means for crushing a watermelon” you’ve eliminated a couple billion possibilities (a piece of paper, for example).

  37. Jules, the original statement was that “software running on a processor ain’t corresponding structure (and it ain’t according to Aristocrat)”. Aristocrat held that an algorithm on a processor is structure for 112/6.

    Would you say that processor+algorithm can be 112/6 structure? Would you say that “software running a processor ain’t corresponding structure” is incorrect, if we assume “algorithm”=”software”?

  38. Ned It’s “protection” is invoked by specific magical words. If they are not used, their non use must be deemed intentional.

    Ned, you’re wrong about this (but I wish you were right).

    112P6 may be invoked by certain “magic words” (e.g., use of the phrase “means for [insert function here]“; or the use of a similar phrase and a statement in the specification or prosecution history that applicant intends to invoke 112P6 through use of the similar phrase) but there are an infinite number of other ways for 112P6 to be invoked during claim construction. The court’s job is to determine if the applicant intended to invoke 112P6, regardless of any “magic words.”

  39. Mooney, your reference to ex parte Gardiner is a bit off topic for this thread. However, it does go to illustrate why the MOT test is a bad test if we believe inventions such as the one discussed related to graphics engines should be patentable.

  40. “Aristocrat did not hold that software running on a processor is not structure. It held the opposite.”

    You could have been a bit more clear with that statement. Aristocrat held that general processors and general software are not sufficient dislosure for “structure” or “acts” respectively in accordance with 112/6.

  41. In discussing the proper “construction” of the claims, the BPAI said this

    “As the court set forth in LG Electronics:
    ” ‘[A] claim term that does not use ‘means’ will trigger the
    rebuttable presumption that 5 112 ¶ 6 does not apply.’ ”
    This presumption can be rebutted “by showing that the
    claim element recite[s] a function without reciting
    sufficient structure for performing that function.”
    LG Electronics, Znc. v. Bizcorn Electronics, Znc., 453 F.3d 1364, 1372
    (Fed. Cir. 2006) (citations omitted).”

    Note that the case they relied on arose from the district courts. There is a host of case law that the USPTO is not supposed to construe claims the way courts do. Rather they are supposed to adopt the broadest reasonable construction. In re Morris. How then is it consistent with their rules of construction to “save” a functionally expressed element by interpreting it under Section 112, p.6 when the claim itself does not use the magic words?

    If the claim can be saved by using the magic words, it should be he applicant who saves it by amendment!

  42. 10 years from now after the supremes sort out this whole “an algorithm is structure” nonsense we’ll all look back at this thread and lol at the people thinking it is.

  43. anonymous: bzzt. wrong answer.

    Aristocrat did not hold that software running on a processor is not structure. It held the opposite.

    From Rodriguez BAPI decision (subject of this blog column): “Recently in Aristocrat, … the court set forth that for a claim to a programmed computer, a particular algorithm may be the
    corresponding structure under 5 1 12, sixth paragraph … The court in [Harris Corp.] characterized the rule of WMS Gaming as follows: “[Tlhe corresponding structure for a S 112
    p. 6 claim for a computer-implemented function is the algorithm disclosed in the specification.””.

  44. You mean kind of like what beauregard claims lack?

    bingo! if software running on a processor ain’t corresponding structure (and it ain’t according to Aristocrat), reciting “instructions” embodied on a computer readable medium isn’t corresponding structure either.

  45. Am I misreading MPEP 2173.05(g)?

    Bruce, where does the “little weight in claim interpretation” come from?

    “A functional limitation must be evaluated and considered, just like any other limitation of the claim, for what it fairly conveys to a person of ordinary skill in the pertinent art in the context in which it is used. A functional limitation is often used in association with an element, ingredient, or step of a process to define a particular capability or purpose that is served by the recited element, ingredient or step
    … “In the absence of modifiers, general descriptive terms are typically construed as having their full meaning.” ”

    True, an apparatus’s structure must be what distinguishes over the prior art, the structure can be functionally defined; if the function defines and limits the structure (e.g., “a non-ferrous cog”), then that can distinguish over the prior art.

  46. Bruce, I would invite you to read the opinion and not just Dennis C’s summary. The BPAI did not only construe the functional limitations as MPF. They also rejected them under Halliburtion if they were not properly construed as falling under Section 112, p.6.

    As I said earlier, Section 112, p. 6 is permissive. It is not a command to construe all functionally expressed elements per the corresponding structure. It’s “protection” is invoked by specific magical words. If they are not used, their non use must be deemed intentional. If so, I cannot see how Section 112, p.6 can save these claims as being all but PER SE invalid.

  47. “However, a claim with a preamble identifying it as structural cannot stand absent structural language to limit the scope of the claim structurally.”

    You mean kind of like what beauregard claims lack?

    “Functional language in a structual claim is ok as a way of making the claim more understandable, but should be given little weight in claim interpretation. ”

    I don’t know where you came from but I like you already.

    “http://www.linkedin.com/in/brucehouston”

    That you?

    You were looking good until I saw: Bruce Houston’s Specialties:
    Bruce practices in the areas of microprocessors, semiconductor memories, imaging, telecommunications, Internet-based technologies, navigational systems, telemetry, software, and business methods.

    How did you allow software and business methods to creep in there?

  48. Comments by RedMonkey 11/11 at 4:05p and by Interested on 11/12 at 1:38p, among others above, would appear to ignore the fact that claim 1 is an apparatus claim, not a method claim. The degree of detail with which functional language is parsed out does not matter for an apparatus or system claim. Ultimately, STRUCTURE must be found somewhere to support a structural claim. Functional language in a structual claim is ok as a way of making the claim more understandable, but should be given little weight in claim interpretation. However, a claim with a preamble identifying it as structural cannot stand absent structural language to limit the scope of the claim structurally.

  49. Interested, several of us have pointed out that this is not just another MPF case, nor is it a case where functional claim limitations were DEEMED to be MPF. The board rejected both the structure, method and article claims under Section 112, p. 1, independently of Section 112, p.6, requiring them to enable the full scope of the functional language employed, legally relying on Halliburton via their prior board decision Miyazaki.

    While they went through the Wand factors, they also emphasized in discussing the apparatus claims that it is the rare case where a functional claim term is fully enabled, citing the Judge Rich opinion in In re Boller. Indeed, Halliburton had a fully described and enabled embodiment, but the Supremes still condemned it because the broad functional language literally covered the inventions of others that did not employ the disclosed structure.

    The ONLY hope for such claims in the future is IF the Feds hold that such functional language MUST be construed under Section 112, p. 6, regardless — thereby saving the claim if it has ANY fully enabled and described embodiment.

  50. I am getting the sense that the invention wasn’t complete yet when the application was filed. There wasn’t sufficient structure disclosed because the inventors didn’t know what the structure would be. I don’t know why they didn’t file CIPs when the invention was actually enabled.

    lack of corresponding structure does not equate with lack of enablement

    in the past, most software applications omitted particular algorithms for functions that could be readily implemented. Why? because merely reciting the function enabled the function.

    Arisocrat throws a wrench into that practice

  51. This case is essentially rehashing old law (Aristocrat, etc.). WHY is no one discussing the truly disturbing aspect of this case which dealt with rejecting the method claim, which had -nothing- to do with MPF?? As the blog post states, “as the use of MPF claims has dropped, this third portion of the opinion may be the most important.”

    With respect to the MPF issues, I do, however, echo Red Monkey’s comment
    “The question really is in how much detail must an algorithmic step be broken down before it can be regarded as supported under Aristrocrat?”

  52. “I’ll get ya’, get ya’, get ya’, get ya’ ”

    Cripes, Mooney is losing it. And here I thought he had hit bottom years ago.

  53. bob, I am getting the sense that the invention wasn’t complete yet when the application was filed. There wasn’t sufficient structure disclosed because the inventors didn’t know what the structure would be. I don’t know why they didn’t file CIPs when the invention was actually enabled.

    Lionel, if I am right, I don’t agree that the claims were poorly drafted, or not *only* poorly drafted. How can you draft good claims when the invention isn’t done?

  54. “Not so. The applicant faired no better for the claims in which the Board did not invoke 112/6. The claims were found deficient using essentially the same reasoning as the claims that did invoke 112/6.”

    Huh? The board still invoked 112/6. They decided that the words didn’t recite structure so they treated it as MPF. They even went out of their way to say that if they had recited “circuitry” everything would’ve been ok.

    This shows that you get what you pay for. You want a $5k application, this is what you’ll get.

  55. Is this a $5,000 fixed fee application? It appears to be an application written in house.

    Personally, I think this court and CAFC have mucked up MPF doctrine so much that it’s impossible to tell what will be construed as MPF. Therefore, ensure your specification is replete with hardware examples.

  56. RedMonkey said, “So we are back to where we started. Given a limitation “means for doing X”, how can the function “doing X” be be broken down into little steps [x1, x2,...xn] that are not, in themselves, just as functional sounding as the big step X?”

    As I read this opinion, if the specification describes the little steps x1, x2… xn, where x1, x2… and xn are not hand-waving or magic-happens-here steps, then you’re not going to have a problem claimimg “a means for doing x.”

    It is a matter of degree. Let’s use another example: “means for creating a hash value from the message.” There are lots of ways to create a hash value (link to en.wikipedia.org). But if the specification disclosed that “a hash value may be created passing the message into an MD4, MD5, SHA-0 or SHA-1 algorithm,” your going to satisfy 112p6.

    There really is nothing to see here. This case does what WM Gaming and Aristocrat said should happen.

  57. Just thought I’d mention BPAI’s recent Ex parte Gardiner decision again.

    link to des.uspto.gov

    In light of the views expressed by the Justices during the Bilski arguments, perhaps it’s more prescient than I previously thought:

    “As claimed, this transformation is to create a volumetric obscurant, which Appellants argue has uses in simulations such as creating the visual effect of an airplane flying through a cloud. Reply Brief 5. These arguments, however, have not persuaded us that the claim is drawn to transforming an article to a different state or thing. The polygon is discussed on pages 1 and 2 of Appellants’ Specification as defining an area in which to simulate realistic-looking obscurant elements (e.g. clouds) in an image. Thus, neither the polygon nor the obscurant represents a physical item; rather, the polygon is an abstraction to define a region of an image. Thus, (1) the steps of spreading sub-pixels recited in claims 1 and 20, and (2) the step of modulating the density of the sub-pixels recited in claim 1 do not transform any physical object. The image is not a physical object and the data, which is part of a simulation, does not represent a physical object. Accordingly, we conclude that claims 1 and 20 do not transform such an object or data representing such an object.”

    And a tip of the hat to Aaron at the 12:01 Tuesday blog for pointing this decision out.

    link to 1201tuesday.com

  58. so basically, if you act as your own lexicographer in a software application, you’re acting as your own executioner

  59. “such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” requires the “construer” to interpret the claim per the spec. 112(6) does not require the inventor to “limit the scope of the claim to “the corresponding structure, material, or acts” that perform the function, as required by section 112 paragraph 6.”

    The end of 112(6)provides interpretation instructions to the examiner, potential infringer, and judge. The end of 112(6) does not provide drafting instructions to the inventor. This is bad. Very bad.

    You get what you disclose plus a little extra (equivalence). The inventor admitted that “means for generating a random system configuration file” is known in the art via the inventor’s statement “[a]ppropriate software coding can readily be prepared by skilled programmers.”

  60. “First, you have to have to get 112/6 invoked.”

    Not so. The applicant faired no better for the claims in which the Board did not invoke 112/6. The claims were found deficient using essentially the same reasoning as the claims that did invoke 112/6.

  61. “He failed to include even a generic black-box computer diagram. Not good.”

    Including a generic black-box computer diagram would really make a diifference? Not good.

    Not being able to rely on the common general knowledge of the PHOSITA? Doubleplusungood.

  62. So other than arguing their case in light of this decision; the implicit question then; given their spec; is whether there’s any way LSI could word their claims to pass muster.

  63. Also important here … the Beauregard/CRM claim was rejected similarly, and even the method claim rejected under 112/1 enablement thrown in for good measure.

  64. Just one more point to stir the pot here, the board made it somewhat clear that if a wholly functional claim limitation did not receive the benefit of Section 112, p. 6, it would be subject to Halliburton. Halliburton, if you recall, invalidated a functional claim element DESPITE their being an adequately disclosed embodiment.

  65. The only interesting thing in this decision was citation to a Judge Rich decision from 1964. No wonder one of the S.Ct. Justices gave a nod to Rich.

  66. To be clear, the board held that terms such as “system configuration generator” had no art-recognized structure like “detent” in “detent means.” Thus it, plus its appended functional language “configured to” perform a function, were as purely functional as any means plus function.

    They then held the particular claims unpatentalbe under several grounds.

    If construed to be means plus function:

    1) no corresponding structure described – 112, p.2;

    2) lack of enablement under 112, p.1; and,

    if not construed to be means plus function:

    1) lack of enablement for essentially the same reasons under 112, p1.

    I can see these “alternative” rejections entered against such claim in the future until the Feds sort out the proper rejection.

    My own preference would be the latter course, as Section 112, p. 6 is a savings ground for functional claiming that is permissibly invoked by an applicant. If is intentially not so invoked, I think it improper to construe the claim to be means plus function regardless.

  67. I love me some blondie.

    Whares NWPA when you need him?

    Hierarchy seems to have decided to ignore the “third portion” of the decision. Miyazaki up in da house rockin’ ur house to tha ground.

  68. This case illustrates the dangers inherent in $5000 fixed fee applications. You get what you pay for and nothing more. C’est la vie.

    Having said that, this was a poorly draft case even for a low-budget fixed-fee case. At minimum the attorney could have placed the invention in context in a computer system. He failed to include even a generic black-box computer diagram. Not good.

  69. If so, then the claim includes “structure” to perform the function, and therefore 112/6 is not applicable

    Maybe. But I still think you run the risk that you’ve recited a multi-faceted function with no underlying algorithm

    you’ve recited a function of “receiving a set of X” but no algorithm for such receiving

    you’ve recited a function of “computing a hash of each element of X” but no algorithm for such computation

    you’ve recited a function of “storing the hashes in a two-tier cache” but no algorithm for such storing

    I’m not sure if you add a “preamble” (for lack of a better word) that states “for generating … by” will work, but it’s a decent strategy, I suppose.

    how about a generic function like “a module for computing a result by …”? Would that be enough?

  70. anonymous, I don’t think this is new; read the cases a number of things have to occur to get stuck the way LSI did.

    First, you have to have to get 112/6 invoked. For exmaple, you need some element of a claim that has no ordinary understood meaning and which has no structural limitations that perform the function. “module” might or might not be a problem. “a sorting module to sort the list” might not be a problem. “a module to generate a [previously unknown data strucutre]” could be a problem.

    Second, if 112/6 is applied, you have to be lacking some structure (e.g., algorithm+CPU) in the specification. Recall that 112/6 is about, in effect, using the spec to define the claim. If you have “a module to sort the list”, even though there may be hundreds of known algorithms for sorting, the claim is limited to those disclosed in the specification. It might be enough to simply list the names of a bunch of known sorting algorithms. But if you disclose no sorting algorithms, the claim is indefinite.

    In my practice, I always look apparatus claims over carefully and ask if 112/6 is a danger. There are things to avoid the danger, but the first question is whether an apparatus claim is really a good choice.

    Here’s a question. What if the module actually describes the algorithm in the element. In other words, “a module for generating a new-foobar-data-structure by receiving a set of X, computing a hash of each element of X, and storing the hashes in a two-tier cache”. If “structure” is considered to be “algorithm+computer”, isn’t the recitation of an algorithm+computer in the claim element considered to be structure. If so, then the claim includes “structure” to perform the function, and therefore 112/6 is not applicable.

    If this is correct, then another way to avoid the problem in this case is to recite the steps used by the unit/module/component/etc to actually perform the function, because steps/algorithm + computer = structure (according to the case law).

  71. A precedential opinion of the BPAI …. yawn …. wake me with a REAL court decides something.

    Your wakeup call is coming when examiners follow this opinion, you can’t get any of your clients’ software-related apparatus claims allowed, and your clients can’t afford to appeal to a REAL court.

  72. HierarchyOfPontificationBuckets: There’s no new law in this decision. There doesn’t appear to be any new way of applying the law.

    There is a new application of means-plus-function analysis to terms akin to those commonly used in software applications, e.g., “system configuration generator,” “system builder”

    What’s next? “ module”?

  73. Right, and according to Aristrocrat, for an MPF software claim, the “corresponding structure” means an algorithm.

    So we are back to where we started. Given a limitation “means for doing X”, how can the function “doing X” be be broken down into little steps [x1, x2,...xn] that are not, in themselves, just as functional sounding as the big step X?

  74. DC: In many ways, this MPF structure requirement runs parallel to the written description doctrine. It may well be that one of ordinary skill in the art could easily generate a random system configuration file without undue experimentation. That potential fact is irrelevant for MPF definiteness analysis. Rather, the question for definiteness of a MPF claim is whether a corresponding structure is described in the specification.

    And ever so slowly the art unit grows up.

  75. link to youtube.com

    I’ll walk down the mall, stand over by the wall
    Where I can see it all, find out who ya’ call
    Lead you to the supermarket checkout, some specials and rat food
    Get lost in the crowd
    One way or another I’m gonna get ya’
    I’ll get ya’
    I’ll get ya’, get ya’, get ya’, get ya’

  76. This is not a closed case. LSI is free to try to show what skilled artisans in the field knew at the time of invention, what was well known and therefore implicit in the disclosure, and so on. The Board decision was based mostly on the applicant’s description of the specification’s support for the claims (though they also claim to have looked over the entire specification). The applicant probably has more to say about support/enablement. It well could be that each step in the claims is individually well known in the art, and it is the combination of steps that applicant is claiming as new.

  77. I think 6 is right. Trouble is abrewing.

    I don’t think the problem is the absense of pseudocode. The real problem is that the PTOs interpretation goes against previously settled law that dictates how disclosures are written.

    The settled law is that describing the functionality and not the specific algorithm for implementing an element is generally sufficient to enable one of ordinary skill to write the software to implement the function. So there was no need for source code or pseudo code or to specify the algorithm. In many cases, the applicant could have beefed up the spec with at least some non limiting examples, but the law said there was no need to do so.

    It might well be that the description here falls short of describing the required functionality even under FC precedent, and should be rejected. The real problem here is the invoking of 112, 6th to do the heavy lifing.

  78. Even if an algorithm were disclosed, each step in an algorithm is functional, i.e. each step is essentially “doing X”.

    For example, suppose support for “means for generating a random system config file” amounted to a flowchart saying “generate a random number and retrieve config file corresponding to that random number”. Would that satisfy Aristocrat? After all, those steps are just as functional as the single step they purport to support.

    You could of course break these steps down further, but ultimately, it seems like these atomic steps are just as functional as the more complex step that they built.

    The question really is in how much detail must an algorithmic step be broken down before it can be regarded as supported under Aristrocrat?

  79. So if the en banc CAFC decides in Ariad v Lilly that there’s no separate written description requirement, i.e. that written description and enablement are the same thing, how will that affect the decision in Aristocrat and how will affect the BPAI’s reasoning here?

  80. There’s no new law in this decision. There doesn’t appear to be any new way of applying the law. LSI needs to write better patent applications.

    Note that the Rodriguez specification incorporates another application which is not publicly available. Neither applicant nor Board referred to it.

  81. You know, it was funny, when I woke up today it was as if I had a feeling like that old dude in star wars. He was like “I felt a great disturbance in the force, as if millions suddenly cried out in terror and were suddenly silenced”. Except it was for functional language instead of millions of people.

    When was it published deathto?

  82. “. But I think a rigid requirement for actual psuedocode is not going to withstand opposition, especially where a detailed flow diagram and related description of the process are provided in the specification that clearly enable the invention just as well or better than psuedocode.”

    Makes sense, broje.

  83. I noted long ago, probably on this blog, that system claims that recite functional modules are just means for claims in disguise. And the same was true of many apparatus claim elements for electro mecahical systems claims. Then came the case that said that means for claims to software functions were not enabled absent a psuedocode algorithm in the spec. I’ve been dreading this day. But I think a rigid requirement for actual psuedocode is not going to withstand opposition, especially where a detailed flow diagram and related description of the process are provided in the specification that clearly enable the invention just as well or better than psuedocode. The patents that did not laboriously go through the process steps, however, are going to be in trouble.

Comments are closed.