1) Patent holders only litigate against defendants who have the resources to defend themselves in court.

2) The courts play the role of grown-up and subject patent claims to rigorous and critical review.

The first of the two above conditions seems to hold more often than not, although some notorious early-00′s patent trolls like Pan-IP built up war chests by going after small targets first.

The second condition has not held at all. The CAFC has utterly failed in the role of grown-up, giving birth to a hideously arcane system that primarily benefits the patent bar and stacks the deck in favor of plaintiffs in multiple ways. I see the CAFC, not the USPTO, the patent bar, or even the patent trolls themselves, as the primary villains in the story of modern patent law. It is most satisfying to watch the Supreme Court play the role of REAL grown-up and tear CAFC precedent to pieces in case after case.

As a matter of policy, the USPTO has apparently decided to err on the side of being too liberal in issuing patents, on the basis that it will take litigation to sort out the overlaps in any case. That’s not a completely unreasonable policy, as policies go.

Also, Tim, its hypocritical to accuse Aharonian of lack of argument based on your interpretation that he is insinuating something about Moglen.

I’m going to respectfully disagree with Tim just a bit, or at least suggest a slight change of focus. I don’t think software patents are a burning problem so much as I think BAD patents are a burning problem. I also think the USPTO, as a government bureaucracy full of people who are – to put it nicely – less than the best and the brightest, will ALWAYS be prone to issuing bad patents, and they need a vigilant countervailing outside force to constantly review and correct their work. The Federal court system could concievably serve as such a countervailing force, but not if they are bound by the garbage precedents the CAFC has been spitting out in the last couple of decades. Here’s hoping for more CAFC smackdowns from the SCOTUS!

If I were to delve into the 7,000,000 or so granted patents in the USPTO’s database, how do I know which ones you consider software or combinations of hardware and software? Is a routing algorithm or something like Google’s QoS patent a “software patent”?

As I said, most of the things we do in digital systems are combinations of hardware and software, and it takes the two working together to make something interesting happen.

Anyhow, here’s one patent application I kinda like.

Those who criticize Open Source advocates often accuse them of being relgious ‘zealots’

But what have we here?

A holier than thou tone, accusing FOSS advocates of telling “LIES” and of being “UNETHICAL”

Could we just have solid reasoned argument from anyone at IP Central with out the hate speech???

Just because something that happens in nuts and bolts can be modled by an algorithm in a computer, does not mean that they are the same.

Additionally, the standard of obviousness should apply to most software patents, in that they are often combinations of bits that any skilled programmer could put together.

Because the USPTO/CAFC have “standards.” If you don’t like them, then propose how they can be improved.

Actually that’s very debateable that USPTO has standards. Mt wife’s econd cousin happens to be a Patent Examiner in Munich, and he had occassion several years ago to meet with several IBMers who confided that they could not figure out what the USPTO standards were for software patents. IBM spent a lot of good money trying to do this too.

If IBM with its legions of lawyers and scientists cant figure it out who can.

Think about risk vs uncertainty. There’s a lot of uncertainty out there, created by USPTO.

You still don’t address my points about the statistical insignificance of one patent per week, nor the oddness of basing a patent simply on its claim construction.

Do you not know who Greg is, Tim. He has been very influential in arguing some of the policy positions and reflections on the USPTO you would agree with.

If you’re going to ask Richard to present a patent to you Tim, at least let him know whether you’ll decide on their merit with USPTO/CAFC or your standards. You’ve done 1 patent per week, and I find it interesting to read, but I wish you’d compare the patents’ claims to judicial interpretation and USPTO standards- that would better connect your series to patent policy. Right now, you might as well introduce the series as: “I, Tim, will critique this software patent, and despite it constituting less than .001% of all software patents, please use your imagination of what the other 99.5% of software patents will be like…”

People who know what they’re talking about and know they’re in the right don’t need to resort to name-calling: the facts speak for themselves.

Yes

I’m still not sure I follow your point. The numbers on your thermometer is not software in anything like the same sense that the quicksort algorithm is software. You can’t package the “software” in your example up and sell it as a separate product. So I’m having trouble imagining a patent on a thermometer that we might describe as a “software patent.”

Instead of trading hypotheticals, let’s talk about a real software patent. I’ve now analyzed more two dozen of them. I haven’t found a single one that struck me as meritorious, and so far as I can recall, not a single person has disagreed with my assessment of even one of those patents.

So you’re a smart, technically savvy guy. If you think software patents are worthwhile, give me a real-world example of a software patent you’re willing to defend. Look through my list above, or find another one.

Most of the interesting things that we do in the digital world are combinations of hardware and software, where both are indispensable to the solution. The practice in patents is to attach claims to as many parts of a system that we can identify, mainly because different embodiments of our inventions will assign functions between hardware and software in different ways. You literally can’t draw a bright line between hardware and software in most digital systems patents, nor should you.

Let me give you an example that most of your readers can understand: the thermometer. This is a device which allows us to measure air temperature by measuring the expansion of some liquid in a sealed container. It consists of a piece of hardware – the container with the colored alcohol or mercury inside – and a piece of software, the scale that converts inches of expansion into degrees of temperature.

I maintain that the fundamental element of this invention is the algorithm: converting inches of expansion into degrees of temperature is pretty brilliant. The hardware aspect of the invention is just trial and error with different materials. And newer thermometers act by converting voltages output by analog sensors into degrees, applying a different algorithm but building on the basic insight that the state of a given material is a proxy for the temperature of the air that surrounds it.

So if you invalidate the software aspect of the thermometer, you invalidate the entire invention.

Is that what you mean to do?