End Software Patents

by pabs3 • 5 years ago

75 99 5 years ago

warning26• 5 years ago

I agree with this, but my key question is: why stop at software patents?

Many of these arguments also apply to all patents, not just software ones. Is there a compelling argument to keep any patents at all? They seem like a weird throwback to some past society where reclusive inventors spend all their time tinkering and inventing things.

unityByFreedom16• 5 years ago

Patents encourage innovation by rewarding creators. We wouldn't have vaccines without some form of protection for scientists and their backers.

The reason to ditch software patents is copyright already protects software. Copyright allows you to decide how your code is distributed. Software patents restrict innovation in a manner that was never intended to be possible.

charcircuit5• 5 years ago

We wouldn't have vaccines without some form of protection for scientists and their backers.

Do you truly believe this? If we got rid of the patent system 5 years ago. Do you believe no vaccines for COVID would be made?

SavantIdiot3• 5 years ago

The polio vaccine was given away for free.

knowledge-clay• 5 years ago

We wouldn't have vaccines without some form of protection for scientists and their backers.

Pay them with public funds.

https://en.wikipedia.org/wiki/Prizes_as_an_alternative_to_pa...

js83• 5 years ago

I may be wrong, but I thought mRNA vaccine was only developed after key patents for it have expired.

Whether patents help or hinder innovation seems to depend on how general the patented idea is. If it's really broadly applicable, they hinder it IMHO.

I also don't see the patents benefiting researchers anymore, rather than some corporate middlemen.

erik_seaberg• 5 years ago

At 63–75 years, programmable general-purpose computers and networks are so new that it’s still possible to claim the most obvious solution to many problems simply by being first to think about them. USPTO’s bar for what is “obvious” is skewed towards older subjects with a lot more written and indexed prior art for an examiner to cite, and doesn’t weigh likelihood of easy reinvention.

And now there’s enough money at stake to trigger a land rush to fence off entire problems using solutions that just don’t merit protection.

wavesquid• 5 years ago

I agree with this, but my key question is: why stop at software patents?

Because the key argument is that math is not patentable. Software patents are relatively uniquely expressible amoung patents as mathematical conjectures.

johncolanduoni3• 5 years ago

Why are software patents more expressible in mathematics than e.g. mechanical ones? The physical laws that govern a mechanical device are expressed mathematically.

throwawaygh5• 5 years ago

I'm actually more ok with pure math patents than software patents.

If someone constructively proves p=np then they deserve some term-limited profit taking.

Getting a monopoly for being the first pos to mail some legalese to the right government office about clicking a button once to order something for delivery is both not math and not beneficial to society.

The problem is software patents is not that software is math. The problem is software patents is that most patentable software is obvious to anyone without severe brain damage, and no one in the legal establishment has historically had any clue how to assess whether some idea in software is non-obvious.

jandrewrogers4• 5 years ago

It is trivial to prove that all patents are theoretically equivalent. There is nothing special about algorithm or software patents in this regard.

jbay808• 5 years ago

There are still reclusive tinkering inventors today who need to make a living!

It's very hard to otherwise compete with an established company who can copy your brilliant idea, scale, and distribute it way faster and more cheaply than you can.

SavantIdiot3• 5 years ago

In 2013 it cost $33,000 to file a software patent: a startup I worked for was trying to file one. There's no way a starving tinkerer would be able to afford a patent.

But assume s/he did afford one.

Now comes the theft. Amazon steals it, and now our tinkerer needs to cough up millions of dollars to go after a one-trillion dollar company. It lingers in courts for decades. Even pro-bono legal advice would eventually give up. Futile.

Patents exist only for companies to wage war for monopolies.

_vvhw8• 5 years ago

You're assuming that once you have a patent, you're protected. When in fact, you still need capital to defend your patent in litigation. It's turtles all the way down.

diob• 5 years ago

For the most part, even tinkering inventors are simply forced to sell their big idea / invention to a company at less than face value. Not many have the capital to do anything with their inventions.

_vvhw8• 5 years ago

"A weird throwback to some past society" rings true, especially insofar as the world now has a central knowledge repository—the Internet.

Stopping software patents as a first step would set a precedent and provide a beachhead.

CalChris• 5 years ago

The compelling argument, rewarding innovation, is the same as it was in 1789, in the United States Constitution, Article I, Section 8, Clause 8:

  To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries.

burnished2• 5 years ago

Well, the system doesn't seem to function that way, which makes it feel less compelling.

_vvhw8• 5 years ago

Except that today, the promotion of progress of software is most advanced by open-source, and most damaged by patents.

jandrewrogers4• 5 years ago

There is no principled reason to stop at software patents, they are literally theoretically equivalent to any other patent. Industry has already abandoned them in any case.

In software, we’ve already returned to the state that patents were designed to avoid. Everything of value is treated as a trade secret. In several domains of computer science, virtually everything that is state-of-the-art is never published and is instead slathered in NDAs. And for good reason, patents on computer science have been effectively unenforceable for decades. So instead, no one publishes.

This isn’t speculation, it is official policy in organizations I work with. The end state is that public computer science is mediocre compute science. I’m agnostic about it since it doesn’t affect me either way, but I recognize the suboptimality.

unityByFreedom16• 5 years ago

There is no principled reason to stop at software patents, they are literally theoretically equivalent to any other patent.

They're not. Software is already protected by copyright.

koonsolo• 5 years ago

If you need to invest millions of dollars for some pharmaceutical pill, you probably want to have a return on investment for that. The only way to really have that is your exclusive right to sell it, or license it out.

If you can't have that right, you won't invest those millions of dollars.

But software is not like that, because innovation is mainly driven by building on top of small things, not doing a crazy expensive research project.

So industries are different, and I do see the benefit for some.

unityByFreedom16• 5 years ago

I like the initiative. Why tie it to the free software movement?

Software should be free

This is unnecessary. Copyright allows us to decide how to distribute software we create. By including "software should be free", it sounds like they want to abolish copyright as well.

charcircuit5• 5 years ago

The Free Software Foundation is biased.

unityByFreedom16• 5 years ago

The Free Software Foundation is biased.

Ok? I'm not sure what point you are making. I didn't mention FSF.

bo1024• 5 years ago

Actually they like using copyright to enforce software freedom, but I see your point.

unityByFreedom16• 5 years ago

Then it doesn't make sense to include that statement on the home page. That statement never held water for me. I do think we should look at what it would take to ditch software patents. Now that I see this page is from the FSF, it makes sense. I know RMS will never give up on forcibly making all software free.

hackthefender2• 5 years ago

GNU/Linux Free software such as the GNU/Linux and FreeBSD operating systems were developed without software patents. 91% of the top 500 super computers run GNU/Linux.

This is misleading. Linux has all sorts of stuff in it that used to be under patent protection but isn't anymore because the patents expired. The entire argument for patents is that they incentivize both the creation of new things and the public disclosure of them, with the idea that eventually they will become part of the public domain. There is a complicated cost-benefit analysis here that has been studied by academics for years, which the FSF just ignores. This is fine if you view this as an advocacy piece, but not fine if you want to actually learn something about the issue.

hackthefender2• 5 years ago

Just to say more on this quickly, the relevant inquiry is not "does software progress without patents, yes or no?" Obviously it would. But the relevant inquiry is whether we have made more or less progress now than we would have had software patents not existed for the last 70 years.

For what it's worth, I tend to agree with the ultimate conclusion that software patents today do more harm than good. But if you want to convince policy makers, you have to address the issue on a less superficial level. And that means actually acknowledging the other side's arguments and rebutting them with evidence or counterarguments.

jandrewrogers4• 5 years ago

What, precisely, are “software patents”? It is a suitcase term that includes unrelated things. Technically, it tends to conflate business method patents and algorithm patents, which have different legal rationales. Most egregious complaints are against business method patents that are largely unenforceable anyway. Algorithm patents are enforceable globally because they are directly reducible to electronic circuits, and electronic circuit designs are patentable and treated as equivalent.

The concept of IP is built on top of convenient legal fictions. It was never intended to be consistent. Nonetheless, it recognizes the implications of the cost asymmetry of generating IP versus copying it. There are no trivial economic solutions to the optimization of IP generation, it is an extremely difficult theory problem.

aranchelk5• 5 years ago

The “Software is Math” is both poorly reasoned and probably not desirable.

I think the vast majority of us will benefit from software continuing to be classified as creative works so we can use copyright law to either protect commercial interests and/or as the underpinning for open source copyleft.

throwawaygh5• 5 years ago

The “Software is Math” is both poorly reasoned and probably not desirable.

Hey! Don't you know everyone programs in Coq these days and every CSS style sheet is backed by proof? /s

paulmd2• 5 years ago

The concept of "illegal numbers" is indeed quite interesting when you consider the countability of Turing Machines.

Lemma 1: The computation performed by any program or function (algorithm) operating on a Turing Machine can be represented by an initial Turing Machine state and the rules of turing machine computation

Lemma 2: There are countably many turing machines

Lemma 3: Therefore, a unique integer can be assigned to each Turing-machine program that could ever possibly exist.

Lemma 4: All implementations of a function, regardless of language, can therefore be concisely and uniquely represented by a particular integer.

Contradiction: Implementations of a function can be copyrighted, and yet integers cannot.

Contradiction: Abstract algorithms can be patented, and yet integers cannot.

Computer programs and algorithms are very akin to numbers. Most of them are useless, they don't perform any useful computation, many of them are not "correct", and most of them never even halt, most likely. However, there are certain "interesting" numbers that have useful properties due to emergent properties of the mathematical systems in which they exist. This is highly akin to, for example, prime numbers. Can you patent a prime number? Well, I can patent computer program 389355935, and just by giving you that number, I've given you all the information needed to implement that particular Turing Machine.

baggy_trough2• 5 years ago

Isn’t that like arguing you shouldn’t be able to copyright a book because the text could be encoded as a number?

throwawaygh5• 5 years ago

There's a trivial bijection between the natural numbers the set of all strings.

Every patent application is a finite string.

aranchelk5• 5 years ago

I think lemma 4 glosses over the immense variety of ways computation can be expressed in higher level languages.

At one end of the spectrum is source code that’s so poorly written it’s practically impossible to understand or edit. OTOH, well written code with good abstractions can make changes trivial; it can also document relevant information and educate engineers.

There’s a lot of value and creativity embodied in well written applications beyond the derived object code and those qualities are (in my mind) distinct from compiled algorithms.

vasco• 5 years ago

"Information feudalism" is a good book to learn more about the origins of patents and how they affect the world economy and international relations.

That being said, fuck all patents, ideas don't belong to anyone.

chefandy2• 5 years ago

Maybe folks have answers to these concerns, but every time I’ve seen that Stallmanesque messaging over the past 20/25 years or so, I have the same doubts. (And I have a full time job making FOSS at a university all day.)

From a US perspective, I’d more likely agree if we were even sorta good at handling displaced or obsoleted workers. Abolishing software patents without making a bunch of other changes would be a giant hit to the American economy and kill many productive worker’s livelihoods with no follow-up. It’s wrong to do to folks driving truck and it’s wrong to do to commercial software developers and all the other folks who make their living in that industry. Beyond that, lots of technically boring, unsexy software is super important to the world, and the open source alternatives prove that folks just don’t want to work on those problems for free, especially when there are demanding/intricate/fussy interface requirements. The number of instances in which FOSS properly catered to non-technical users with advanced use cases while delivering fantastic usability is pretty low from what I gather. For example, with the art software I work with professionally, Blender’s comparatively recent changes got them there to some extent, but anyone who thinks gimp is a usable replacement for photoshop hasn’t used either enough to make that judgement. People might even get mad you intimating there’s something to fix. Big corps like IBM won’t pay people to write FOSS all day unless there’s a financial incentive. I can picture all enterprise software now being so complicated and convoluted to install that procuring expensive services is the only way to use it. Or maybe requiring custom hardware to perform necessary parts of a process. Maybe even abandoning software altogether and making things mechanical again? Not that computing is perfectly green, but how would that work out for the environment?

I’d love to get pushback on any of this if folks have thoughts. Personally I’d love to see a world where intellectual property just wasn’t a thing… but we don’t live in that world and intellectual property laws are not the biggest barrier to achieving it. It just seems like a super naive demand that assumes software operates in a vacuum and that parity between commercial and open source offerings is even possible in our economic system.

(Downvotes aren’t counterarguments)

zzo38computer2• 5 years ago

prove that folks just don’t want to work on those problems for free

Well, some people do want to work on such problems for free, even if not everyone does.

Maybe even abandoning software altogether and making things mechanical again?

There are some advantages to such a thing, since computers are overused for many things. However, a computer is still useful for many things; mechanical isn't better for everything.

Not that computing is perfectly green, but how would that work out for the environment?

That I don't know, but should probably be considered too. However, often companies put too much extra stuff in computer programs that is used even if you do not want it, which also wastes energy.

Personally I’d love to see a world where intellectual property just wasn’t a thing…

I agree to abolish copyright/patents (although trademarks can be useful, but still there are things that should not be sued for trademark infringement, so it still needs to be fixed due to this). But, there are also ways to go part way toward abolishing copyrights and patents; Question Copyright had some ideas. Another idea would be limiting their scope so that if it has freedom then it can be ignored, or to be ignored for practical works, to limit durations, other changes, etc. Another idea would be that if you declare yourself exempt then you can neither claim copyright or patents on anything but neither can be sued or arrested for it, either. However, I think it would be much better to abolish copyright and patent laws entirely.

chefandy2• 5 years ago

“Well, some people do want to work on such problems for free, even if not everyone does.“

It’s not whether anyone does, it’s whether enough people do.

In FOSS I don’t see evidence of any large project getting significant usability work for free. Firefox, likely the most popular user-facing FOSS application, it’s essentially adware via a nonprofit. Blender’s foundation has significant support from corporations. There are people working on those problems for FOSS but they’re likely paid, and that’s just not feasible for most FOSS. Also, freeness in our society can have unintended consequences— the way news organizations are forced to operate because they can’t easily be directly compensated for their work (because people expect it to be free) is a huge problem. I can’t imagine software being any different in today’s context.

Beyond that, most FOSS is entirely developer-managed and project decisions reflect that bias. To this day, many folks think the only problem with git’s interface is its users unwillingness to read the documentation. Literally zero people who aren’t software developers— including the copywriters, project managers, designers, support people, etc. who end up also having to use it— would agree.

There are certainly usability advantages with many hardware applications over software replacements but being able to use generic logic hardware to control things is a pretty huge advantage. Most folks don’t realize how resource-intensive manufacturing is. I haven’t run the numbers, but I’ll bet a few extra PCBs would equal a whole lot of software bloat. Hardware is also vastly less flexible, and as a result would probably be significantly less efficient to create. I haven’t run any numbers though.

I often advocate for drastically scaling back copyrights— especially in media and software. Their current state is absurd.

Forgive my hyperbole to emphasize the dichotomy: Abolishing software patents, however, seems like fighting America’s imperialist behaviors by demanding an immediate disarmament of the country, entirely. I’d love that if our world was in a state that could facilitate that without immediately being overtaken by China… that wouldn’t be better for anybody.

zzo38computer2• 5 years ago

I would abolish copyright as well as patents (all copyright and patents, not only software patents). Someone might publish something or keep it a secret, but that shouldn't stop anyone else from doing it independently, either by themself or by reverse engineering an existing work.

I had the idea of a license (not yet written) which allows free use like it is in the public domain, and cannot be sued/arrested for copyright violation or for anti-circumvention or other things like that, and no attribution or copyright notice is required, but derivative works etc must be licensed by either the same license or GNU GPL 3 or GNU AGPL 3 or CC-BY-SA 4.0 (only) or later versions of GNU GPL or AGPL. These rights will then apply to patents as well as copyrights (so you can't patent a derivative work either). Someone can claim otherwise but then their claim is invalid, and only in this circumstance can you countersue. It is supposed to be effectively the same if the copyright and patent laws were abolished; if those laws are abolished the only difference is that relicensing them by the other license will not be possible, but the other rights are supposed to be the same. (The reason the limited relicensing is allowed is to be practical.)

Abolishing software patents may be a good start, though. Another alternative would be the law to be made unenforceable against FOSS (although there are probably too many difficulties in doing this properly). However, this would not be as good as just abolishing patents entirely.

When patents are used, they have the effect of making innovation twenty years behind. When used incorrectly, they are even worse.

ordiel• 5 years ago

I think complexity is an aspect that need to be considered here... I personally would feel like instead of spending 10 years working on something that will be copied so someone else takes the benefit I would much rather not to inovate and waste that time.

Similar as to how google did steal the JVM claimimg that just because they "added" "some" memory features they suddenly revolutionized the entire thing... Bullshit specially being open sourced they could have easily added those features so that THE ENTIRETY of the java/jvm ecosystem could benefit from such improvements, even in desktop, ALSO allowing what now we even came to know as "android apps" to run in other platforms since well at the end... It just a jvm. A move which by the way has become even much clearer now that they are changing the tratitional packaging format (casually) now that android apps begin running on other devises. Its clear it's intent was to get a bunch of already well trained coding monkeys by effectively copying the entire API so they could quickly ease adoption of their platform without giving back to the same jvm community that they copied/reused to make billions

Sadly the person choosen to explain this at the SCOTUS was too stupid and not technical enough to put this into words

End Software Patents

Related Stories