3 ms·

Why is nobody talking about the real issue here: the slashed budget of the USPTO that has made them unable to properly vet patent applications for a decade? Th

Why is nobody talking about the real issue here: the slashed budget of the USPTO that has made them unable to properly vet patent applications for a decade? The problem isn't the trolls, it's the fact that these bullshit patent requests are being granted in the first place. The USPTO can't afford to have experts review patent applications, and this is the result.

rick_2047 16y ago
While I totally agree with you (the problem is not actually the trolls) but these kind of patents can be easily identified as trolling.
- redthrowaway 16y ago
  By someone who knows what they're doing, sure. Consider, however, the patent for Linked List that was granted: Any programmer would have balked at it, but the unspecialized clerk reviewing it wouldn't have had a clue.
  - kajecounterhack 16y ago
    Whoa Linked Lists got patented?!?! That's crazy. Do you happen to have a link about it, I'm really curious about this. That's friggin crazy...
    - redthrowaway 16y ago
      Yeah, it's pretty damned discouraging. http://www.google.com/patents?id=Szh4AAAAEBAJ&printsec=abstract#v=onepage&q&f=false http://www.google.com/patents?id=Szh4AAAAEBAJ&printsec=a...
      - alain94040 16y ago
        Sorry, but that patent does not cover linked lists. It covers the merging to two lists using linked lists so that there are multiple ways to loop though a list of items. Fairly obvious? I would say so. But not the sensational claim that "linked lists are patented". By making such bogus claims, you actually hurt the cause of people who want real patent reform. This opinion coming from someone who had to go through the legal system to file and defend bogus patent infringement cases.
        Natsu 16y ago
        Would it make you feel better if he'd said that it covers doubly linked lists? Because those appear to satisfy every element of claim #1 that I can find. IANAL, but I'm pretty sure that's the same as being patented. Specifically, we have a "plurality" of objects in a doubly-linked list because it's a list and we have a primary and auxiliary pointer that lets the computer go through all the items without resorting them. In a doubly-linked list, the primary pointer points forwards, the auxiliary pointer points backwards and it lets a computer program go through the list without sorting. Notably, there's no limitation I can see on claim #1 that forbids the auxiliary pointers from all going backwards and the field of invention might as well just say "computers." Maybe there's something limiting it in the prosecution history, but I'm not going to hold my breath. EDIT: Also note that nothing in that claim's text requires merging two lists. If I'm misreading anything, feel free to cite the exact portion that says otherwise. At first, I honestly thought this was from a decade or more ago. But then I noticed this: Filing date: Sep 26, 2002 Issue date: Apr 11, 2006 Yeah, because nobody was using more than one pointer on a linked list before then. I feel like I should grep some old FOSS code and see how many published examples of this "invention" I can find. I don't know enough LISP to be sure, but I wonder if the cons cell itself can't almost be prior art.
        alain94040 16y ago
        Agreed. I'm all for a wall of shame. In 2002, if you didn't know that double linked lists exist, you deserve to have your name and picture on every billboard in Silicon Valley. Will a generous sponsor (EFF, FSF), spend the $2,000 to get a banner on highway 101?
        Natsu 16y ago
        I think they prefer to get reexaminations to get rid of the patent and have a "Patent Busting Project" to that effect. While a wall of shame is an interesting idea, killing the patent outright is likely a more effective use of their funds. Some of the entities in question aren't really known for their shame and only programmers are likely to appreciate the absurdity of patenting a doubly-linked list in this day and age.
  - nickpinkston 16y ago
    I'd argue: The ability for one man to know what "real innovation", is always <, the ability of a team with every incentive of faking them out to overcome.
    - inerte 16y ago
      Yes. But you can not crowd-review a patent, since it sometimes gives hints about whole business models, and getting others to know your inventions is dangerous (small differences are easily defensible in patent claims). Therefore you can not have (> reviewing applying).
nickpinkston 16y ago
I'd argue that: 1.) There are bad feedback loops between speed / quality / price of patent examination and the amount of ridiculous patents issued. 2.) Increasing the budget to improve screening and throughput will lead to an increase in patents submitted - will this increase at a higher / lower rate than their budget? I.e. will new patent submissions bloat the office even more than now? 3.) We could try to increase patent fees to cover costs, but then you'll "hurt the little guy" and fully benefit the IBMs who can afford to blow money like this. 4.) The boundary is quite gray between trolls (IV and others) and institutions with near-troll behavior (IBM's if-it-sticks patent submission methods). 5.) There are higher ratios of bad patents in certain categories: process patents, software, etc. Conclusion: I bet that reducing the scope of what categories are patentable scales very well in reducing bad patents than any method of trying to make a better filter, because better filters will just lead to more patents to fill the void.
- redthrowaway 16y ago
  I think you really need to re-examine 1 & 2. Increasing the speed and quality of patent review may lead to more patent applications, but would reduce the number of baseless patents granted. As for price, the USPTO was never supposed to be self-sufficient. Doing it properly means the taxpayer picks up the tab. That's just something you have to accept in exchange for living in a highly developed economy. IBM's patents, as well as those of Google and others, are generally defensive. They're meant to provide ammo to protect themselves from patent trolls.
  - nickpinkston 16y ago
    It would reduce the ratio of bad/good patents - not number, this is my argument: increasing ease of submitting (speed, etc.) will probably lead to more patent submissions - the relation of these curves says whether the number would decrease. I wasn't arguing for pricing it at true cost - I'm saying that if you're subsidizing an activity and increasing throughput - you tend to get doubly more of it. Defensive patents also restrict innovation - see: "patent thickets". Having IBM patent worthless things is just as bad - probably worse than smaller ones because of their credible threat.
chc 16y ago
I don't think the causal relationship between budget cuts and bad patents is so clear. If the Patent Office weren't so promiscuous to begin with, the inability to properly vet applications could very well lead to fewer patents being granted.

rick_2047 16y ago

While I totally agree with you (the problem is not actually the trolls) but these kind of patents can be easily identified as trolling.

redthrowaway 16y ago
By someone who knows what they're doing, sure. Consider, however, the patent for Linked List that was granted: Any programmer would have balked at it, but the unspecialized clerk reviewing it wouldn't have had a clue.
- kajecounterhack 16y ago
  Whoa Linked Lists got patented?!?! That's crazy. Do you happen to have a link about it, I'm really curious about this. That's friggin crazy...
  - redthrowaway 16y ago
    Yeah, it's pretty damned discouraging. http://www.google.com/patents?id=Szh4AAAAEBAJ&printsec=abstract#v=onepage&q&f=false http://www.google.com/patents?id=Szh4AAAAEBAJ&printsec=a...
    - alain94040 16y ago
      Sorry, but that patent does not cover linked lists. It covers the merging to two lists using linked lists so that there are multiple ways to loop though a list of items. Fairly obvious? I would say so. But not the sensational claim that "linked lists are patented". By making such bogus claims, you actually hurt the cause of people who want real patent reform. This opinion coming from someone who had to go through the legal system to file and defend bogus patent infringement cases.
      - Natsu 16y ago
        Would it make you feel better if he'd said that it covers doubly linked lists? Because those appear to satisfy every element of claim #1 that I can find. IANAL, but I'm pretty sure that's the same as being patented. Specifically, we have a "plurality" of objects in a doubly-linked list because it's a list and we have a primary and auxiliary pointer that lets the computer go through all the items without resorting them. In a doubly-linked list, the primary pointer points forwards, the auxiliary pointer points backwards and it lets a computer program go through the list without sorting. Notably, there's no limitation I can see on claim #1 that forbids the auxiliary pointers from all going backwards and the field of invention might as well just say "computers." Maybe there's something limiting it in the prosecution history, but I'm not going to hold my breath. EDIT: Also note that nothing in that claim's text requires merging two lists. If I'm misreading anything, feel free to cite the exact portion that says otherwise. At first, I honestly thought this was from a decade or more ago. But then I noticed this: Filing date: Sep 26, 2002 Issue date: Apr 11, 2006 Yeah, because nobody was using more than one pointer on a linked list before then. I feel like I should grep some old FOSS code and see how many published examples of this "invention" I can find. I don't know enough LISP to be sure, but I wonder if the cons cell itself can't almost be prior art.
        alain94040 16y ago
        Agreed. I'm all for a wall of shame. In 2002, if you didn't know that double linked lists exist, you deserve to have your name and picture on every billboard in Silicon Valley. Will a generous sponsor (EFF, FSF), spend the $2,000 to get a banner on highway 101?
        Natsu 16y ago
        I think they prefer to get reexaminations to get rid of the patent and have a "Patent Busting Project" to that effect. While a wall of shame is an interesting idea, killing the patent outright is likely a more effective use of their funds. Some of the entities in question aren't really known for their shame and only programmers are likely to appreciate the absurdity of patenting a doubly-linked list in this day and age.
- nickpinkston 16y ago
  I'd argue: The ability for one man to know what "real innovation", is always <, the ability of a team with every incentive of faking them out to overcome.
  - inerte 16y ago
    Yes. But you can not crowd-review a patent, since it sometimes gives hints about whole business models, and getting others to know your inventions is dangerous (small differences are easily defensible in patent claims). Therefore you can not have (> reviewing applying).

nickpinkston 16y ago

I'd argue that: 1.) There are bad feedback loops between speed / quality / price of patent examination and the amount of ridiculous patents issued. 2.) Increasing the budget to improve screening and throughput will lead to an increase in patents submitted - will this increase at a higher / lower rate than their budget? I.e. will new patent submissions bloat the office even more than now? 3.) We could try to increase patent fees to cover costs, but then you'll "hurt the little guy" and fully benefit the IBMs who can afford to blow money like this. 4.) The boundary is quite gray between trolls (IV and others) and institutions with near-troll behavior (IBM's if-it-sticks patent submission methods). 5.) There are higher ratios of bad patents in certain categories: process patents, software, etc. Conclusion: I bet that reducing the scope of what categories are patentable scales very well in reducing bad patents than any method of trying to make a better filter, because better filters will just lead to more patents to fill the void.

redthrowaway 16y ago
I think you really need to re-examine 1 & 2. Increasing the speed and quality of patent review may lead to more patent applications, but would reduce the number of baseless patents granted. As for price, the USPTO was never supposed to be self-sufficient. Doing it properly means the taxpayer picks up the tab. That's just something you have to accept in exchange for living in a highly developed economy. IBM's patents, as well as those of Google and others, are generally defensive. They're meant to provide ammo to protect themselves from patent trolls.
- nickpinkston 16y ago
  It would reduce the ratio of bad/good patents - not number, this is my argument: increasing ease of submitting (speed, etc.) will probably lead to more patent submissions - the relation of these curves says whether the number would decrease. I wasn't arguing for pricing it at true cost - I'm saying that if you're subsidizing an activity and increasing throughput - you tend to get doubly more of it. Defensive patents also restrict innovation - see: "patent thickets". Having IBM patent worthless things is just as bad - probably worse than smaller ones because of their credible threat.

chc 16y ago

I don't think the causal relationship between budget cuts and bad patents is so clear. If the Patent Office weren't so promiscuous to begin with, the inability to properly vet applications could very well lead to fewer patents being granted.