>> The trouble is you and most programmers don’t understand that the claims define the exclusive right. If you layer on 26 specifics in the claims you could make anything patentable pretty much. Then if I want to copy you I make my product with 25 of those specifics, but leave one out. I could never be found to infringe you.

Let’s look at a few underlying problems before diving into this example about the 25 claims:

For starters, it’s easiest to come up with the broadest claims where the fewest details need to be known (and can even be guessed at with little recognition for difficult subtle issues), yet these are the claims that give you the widest scope of coverage! This means the people with less sophistication have been empowered to place barriers and hand-cuffs on the smartest of inventors, and they can do this early in the evolution of the art and science. [This has nothing to do with software, except software has such low capital requirements to manufacture, retool, distribute, etc, that we find many thousands of developers participating rather than tens (if that many). Thus, the opportunity costs of monopolies (and “taxes”) are that much greater for areas like software, meaning the liabilities of these patents are that much greater.]

To encourage the less capable to file the patents fastest and so get them, the patent system awards a staggering **20 years** of potential monopoly and/or discriminate taxing/injunction rights. There is also a fairly low “innovation” bar to be met (discussed further in a moment). This means you have potentially many less than brilliant (or less than highly skilled or even modestly skilled) people all able and trying very hard to grab the broad patents. It’s also true that some of the more anti-competitive will have a much greater tendency to participate over those that tend to value collaboration or who dislike hand-cuffing their peers.

To bias against the vast majority of inventors (and greatly in favor of large corporations with many contracted employees and years in business), the patent system makes it money-wise expensive to come up with a good patent or even with any patent. Certainly, the patent is not free or automatic as is copyright. This means that if you create a large work ahead of others (or alongside others), you might have many thousands of details and larger ideas that could be patentable, but most people will not know which of these if any to try and patent. Most people don’t have the money, naturally, even if the ideas are already implicit in their carefully crafted and finished work. Patents have always been a game of money played by those with high ambitions (eg, needing millions or billions in investment). It is not a game designed to deal with ordinary individuals and small businesses; hence, we see why software patents are such a failure and biased against most developers/engineers/inventors/innovators. The patent holder can block you from the patented ideas, and can use all of your unpatented ideas freely.

With so many ideas in every single body of code (think of fiction.. I’ll explain better in a moment), the patent system is making developers walking liabilities. In some capital intensive industries, for example, companies can protect themselves because there might only be a few other likely competitors. With software, where the bar is extremely low, we find that the many ideas will be patented by someone at some point in time, and the patents will land in the hands of many distinct groups — this is a mathematical probability/statistics reality.

The many ideas that don’t get patented right away and perhaps not until they were fully developed by several independent groups across many years and products, can still get patented and frequently don’t get challenged sufficiently in courts because of the very high costs to do so and difficulty in recovering evidence. [See the Red Hat virtual desktops suit they just won with 3 cases of provable prior art. Red Hat had help from the wider open source community that was helping them find prior art, and still they spent 3 million dollars and 3 years I think it was.] This means that the patents (and associated 20 years of control) aren’t going to those that first or best know the inventions. That’s an awful lot of power to put into the wrong hands (assuming any single person’s hands could even be deemed worthy to block off many thousands of others).

It also means that many people are able to copy ideas already expressed by several groups, get a patent, and then sue many years later when it is very difficult to track down the prior art.

Writing software is like writing fiction: you are limited in concepts you can create only by your imagination and “writing” skills because you are creating virtual worlds. Mother nature never interferes or hides deep secrets that impede or prohibit general software development (thanks to digitalization and to the well-defined mathematical, idealized underpinnings). You build based on models with rules you create.

This also means ideas are particularly cheap for software (and for fiction), while more detailed and precise implementation is the much more difficult part and the part more closely dependent on your skills and understanding.

Obviously, models matter, but that is only the beginning, and frequently the core of the models come from distinct disciplines where patents are not awarded (and this is a main reason why those models got developed). Some software is simply a representation of mathematical realities, for example — the sort of stuff many mathematicians have been developing for years and years, thankfully, without a patent/monopoly framework.

What a low bar:

The bar to getting a patent is essentially that you find something non-obvious to the ordinary practitioner. This is a ridiculously low bar.

[Note that a very high bar will still get in the way of the most advanced implementations of the most advanced models. There is a reason mathematicians don’t patent their research (but, unfortunately, others take these and patent them, eg, as software implementations).]

First, we’ll find that many people will be able to come up with the inventions read by patent claims without much help at all beyond the same things the patent authors had accessible (which is a lot; note that inventions are hardly out of their time because society’s inventors and inventions evolve alongside each other). How do we know this? Just look at the bar: “non-obvious to an ordinary practitioner”! In fact, most things that would pass this bar could be developed “independently” by almost **any** practitioner at some point before 20 years. What might not be obvious right away, might eventually become obvious (without the help of the patent or patent author). That is what non-obvious means, that it might be well within your means but simply not immediately clear and perhaps not until you’ve thought about the problem for a while longer.

I don’t know how patent law stayed with such a low bar matched up with such a long monopoly period for so many years, but perhaps the clue is in that when there are significant hurdles already (eg, like very high costs and/or mother nature fighting you), the patent is the least of your worries, and most won’t apply or can’t/won’t afford to experiment with costly laboratories. On the other hand, the fact anyone can participate writing software (doing math, fiction, music, business, etc) and contributing ideas and suggestions brings the “obvious” folly of this system to the fore.

Einstein needed social context (key experiments and mathematics done by contemporaries, and perhaps many other clues also coming from peers) in order to make his great inventions. He didn’t need to be motivated by patents (and did know about them very well), and in fact would not have been able to come up with his inventions of others had been patenting along the way. Of course, science and mathematics is (or has been) out of the scope of patenting. It’s clear monopolies on ideas stifle greatly if enforced. And not even a real Einstein can come up but with a few of the key ideas ahead of everyone else.

Finally, before addressing the quote at the top specifically, let’s tie together a few more points. Most people already are priced out of the patent game. Even if they wanted to play a little, in the software and fiction worlds, they are overwhelmed with ideas and approaches. They could spend all their time patenting and never get anything else done or cover all their ideas. AND many others would surely beat them to many cases, even if the person was very smart (the world is full of smart people). Also note that knowing how to write software (or at least well) isn’t even a requirement for getting broad “software” patents that might read into many future inventions. You mostly have to be quick and be willing to spend money. You can even get help with the technical details after you have started writing the patent application. Software patents are a “land” grab open to almost anyone and which simultaneously turns many infinite resources into scarcities. Well, the main point to be made in this closing paragraph is actually that, the more you know, the more things are obvious to you and the less you might think is actually patentable and try to patent (eg, if you decide to dedicated a great bunch of hours and dollars to play the game rather than develop the “obvious” ideas into something of utility); however, if the law grants patents to what is non-obvious to the ordinary practitioner, then it follows that it will grant patents to what is obvious or even ridiculously obvious to some of the smarter folks.. or to the not-so-obvious to those that are not as skilled but worked on a particular problem for a few months (perhaps while many of the more capable peers were busy with more sophisticated matters and other problems).

Alright, onto the 25 claims brief discussion:

Patent claims are structured so that many claims have many fewer than 25 properties/features. Many patents exhibit a step-wise approach with respect to required features described by the claims, where a new claim repeats the broader invention claimed ahead of it and then adds a single (or two) new feature(s). And every now and then, several branches are covered.

This means that an implementer avoiding the 26-feature claim doesn’t gain much if s/he then bumps into 10 other broader claims that preceded it.

Finally, you can’t make a better wheel than a round one (meaning trying to patent around “roundness”). You can’t make 2+2 equal anything but 4. You also can’t also cover a distance with a shorter segment than a straight line.

In other words, we can’t patent around things many times unless we accept inferior (longer, etc) solutions. This is the case, even if we can prove that the patented solution is optimal as shown by the mathematics or logic (and quite possibly when the patent author wasn’t able to do such a proof).

To conclude:

Patents are stifling, and this shows up very clearly when we try to apply them to creations that many are capable (and sometimes very willing) to create. This tends to be the case for works of the mind only, where the final product is information (eg, “processes” and “methods” accessible to most people). And the extreme problem cases are magnified thanks to the Internet and other modern inventions. Also, open source has shown that a monopoly is not only not necessary as a motivator but is undesired by many software developers because of its stifling, biased, and other unfair aspects of the system (and of monopolies in general).

P. Quinn, there are still many areas (traditional heavy industries) where patents will not impede as many people (although the Internet and computing are bringing solutions of all sorts to the hands of many collaborators). As an alternative to abolishing all software patents, changing the law to disallow infringement claims against individuals, non profits, and small and medium businesses and/or to lower the duration to a couple of years, would go far in removing the arguments against software patents and opposition to them. With respect to patent duration: The modern world works too fast. The world of the late 1700s was much much slower and less developed in many ways.

The don’t look advice is good advice.

No patent infringement claim on the use of software on a programmable device might be able to avoid First Amendment violations (assuming it gets past the “abstract” test) if the software was developed relatively independently of all other work (esp, of claim descriptions which are incredibly broad).