LWN: Comments on "Film: "Patent absurdity""

Film: "Patent absurdity"

nye — Mon, 26 Apr 2010 13:31:34 +0000

>Science and math is not patentable, not even in the USA

Penrose tiling?

Film: "Patent absurdity"

daglwn — Sat, 24 Apr 2010 16:09:00 +0000

Every patent is written to cast as wide a net and be as vague and obscure as possible. Software patents are not unique in this regard. I agree that this is a major problem with our patent system.

Intent really doesn't have much to do with this at all. If someone holds a patent and someone else invents the same thing, the second owes royalties to the first regardless of whether it was an independent invention. That's why Bell was so hot to get his patent approved before Gray.

The Bell/Gray story tells us that the patent office has been the source of controversy long before software was even known. If software patents are invalid, we must treat every other patent the same way. I think that's a worthy debate to have. I'm not sure whether patents serve a useful purpose at all today. They are mostly defensive shields now.

Film: "Patent absurdity"

daglwn — Sat, 24 Apr 2010 16:01:52 +0000

Those were invented a few thousand years ago and most certainly fall under prior art.

Film: "Patent absurdity"

anton — Sat, 24 Apr 2010 14:31:40 +0000

If you do not believe that ideas can be rare, then there is no reason to respect or honor anyone who has ever had an idea, because they will just be one of a flood of people with that same idea.
note that this would have to include Einstein, Darwin, Newton and all the 'great' figures in science and math as well.

Science and math is not patentable, not even in the USA. And the research done by these people was done without having an "incentive" from a monopoly (what would that incentive be for science, actually?).

Maybe if you want to have an incentive for inventors, you could look at the incentives there are for scientists; maybe such incentives should be there for inventors, too. But we will never see this happen, because then the big corporations who profit from patents could no longer argue that the patents are needed for the small inventor.

Film: "Patent absurdity"

anton — Sat, 24 Apr 2010 13:56:28 +0000

however without something like patents, what is there to stop a big company from copying something invented by a small company/individual and using their economies of scale and existing distribution to drive the small company/inventor out of business?

Nothing. But the small company can use their invention (hopefully enough to recoup their investment and make a nice profit) before the big company gets going; and that's quite a window of opportunity until the big company has developed or extended their product (if they succeed at all) and ramped up production, marketing etc. There were 700000 copies of VisiCalc sold despite its being cloned soon by several companies, including Microsoft.

OTOH, with something like patents, the big company has a much easier time: Once they notice the small company, they can just send in their lawyers who (probably rightfully) claim a patent infringement and get an injunction, so the small company has hardly any time to benefit from their invention. And the big company does not need to develop or extend, produce, and market anything new.

Film: "Patent absurdity"

nix — Sat, 24 Apr 2010 09:39:22 +0000

It requires creative effort to come up with a new algorithm, sure. But it does not follow that, therefore, anyone *else* who comes up with the same algorithm, using the same creative effort, should be *presumed to have copied the idea from the first guy to patent it* and required to pay. (Encouraging this copying was of course the original intent of the patent system.)

Given that the software world is largely divided into two opposing camps, one pro-patent which *conceals its source code so people cannot copy their algorithms*, and documents them in unreadable patents which nobody ever refers to, and one anti-patent which does none of these things, it seems that the pro-patent guys are trying to have their cake and eat it, demanding payment for something which you were presumed to have copied even though you almost certainly did not copy it *and they know this*.

Film: "Patent absurdity"

patrick_g — Sat, 24 Apr 2010 06:36:36 +0000

>>> Why shouldn't software be patentable? It requires creative effort to invent a new algorithm.

It requires creative efforts to invent division, multiplication, square root, etc.
Do you think these mathematic operations should be patentable ?

Film: "Patent absurdity"

daglwn — Fri, 23 Apr 2010 23:08:26 +0000

Frankly, I don't understand why software gets singled out. Why shouldn't software be patentable? It requires creative effort to invent a new algorithm. The argument that the computing field moves too rapidly doesn't hold water either. Internal combustion engine technology had a similar innovation curve back in the late 19th/ealy 20th centuries to the one we see for electronics today. Just visit the agricultural machinery exhibits at your local state/county/province/etc. fair.

Patents for obvious invention are a problem in every field. The idea of software patents is not the problem. Some of the problems are an overextended patent office, a lack of peer review and all other such things that have been discussed ad nauseum.

There are plenty of entirely reasonable software patents.

Film: "Patent absurdity"

dlang — Wed, 21 Apr 2010 17:35:08 +0000

quote: Inventions are not coming out of nowhere, they are based on past ideas. So it is possible for other people to come up with them.

I think this is the core of the dispute.

if you believe that the idea will come to someone else at about the same time then patents are major harm.

however if you think that there are ideas that will not come to many people around the same time, then patents can be a good thing as they get those ideas to everyone at the expiration of the patent instead of having to wait for someone else to come up with the idea.

If you do not believe that ideas can be rare, then there is no reason to respect or honor anyone who has ever had an idea, because they will just be one of a flood of people with that same idea.

note that this would have to include Einstein, Darwin, Newton and all the 'great' figures in science and math as well.

I believe that there are ideas that are very rare.

There are too many cases where someone has noticed something (frequently the result of an accident of some sort) and pursued it to find something worthwhile.

Film: "Patent absurdity"

ibukanov — Wed, 21 Apr 2010 14:12:25 +0000

> Hyundai developed a new engine design that gave them twice the power and MPG of existing engines for the same cost, weight and size. Don't you think that they should be able to license the design rather than just have it copied by Mercedes?

If it would be possible, then most car companies would create such engine on its own. Inventions are not coming out of nowhere, they are based on past ideas. So it is possible for other people to come up with them. And if two persons would invent the same thing why do we want to award one who invented a month earlier?

The real value of the inventions is when they come into life as a product. Patents just slow down that since after getting a patent there is no insensitive to bring the invention into the world quickly as the company can just sit on the patent while deciding on a marketing plan.

Yup - and this is where it all falls apart...

dlang — Tue, 20 Apr 2010 21:54:45 +0000

no, I was arguing against people who were saying that all patents were evil. I have said several times that I am against software patents, and all the examples I have used have been for physical devices or processes with the people answering me saying that those aren't valid cases either.

Yup - and this is where it all falls apart...

jordanb — Tue, 20 Apr 2010 21:50:22 +0000

So you're arguing against a straw man then?

Yup - and this is where it all falls apart...

dlang — Tue, 20 Apr 2010 21:40:05 +0000

I am not supporting software patents, I am arguing against people who say that all patents should be eliminated.

Yup - and this is where it all falls apart...

khim — Tue, 20 Apr 2010 21:05:19 +0000

The entire idea of the patent system is to make it so that those ideas are available for other people to use freely after the patent expires. If there was not such a system then there would be the risk of such ideas being lost, or locked up through trade secrets (with all the inefficiencies involved with maintaining such secrets, even where they are possible) for much longer time periods.

Yet most patents cover ideas which can not be lost for they are embodies in millions of copies around the world and can be easily reverse-engineered.

Our world today is what it is due to technologies that could have been locked up with trade secrets that are instead now freely available to use. Think of such things as how we make Steel or Aluminum and consider what would have happened if those were trade secrets of one particular company, with every employee being locked up under contract to not reveal the secrets.

Sure. If you can not guess how the product is made even if you have it in your hands - then it's fair justification for monopoly. How many patents describe things like THAT? 0.1%? 1%?

Those are a couple examples where the process could be kept secret while selling the result. Prior to the patent system it would have been likely that such ideas would have died with the inventor, and had very little effect outside the immediate area (remember that if secrecy is your only protection against being flattened by your rivels you don't want to let many people in on the secret, which limits how large you can grow)

Nobody argues agains these patents. But why keep all the software patents around? The "great idea" embedded in software can not be easily lost: even if nobody has the source you can still use the same old implementation via emulators and recompilers. The initial justifications for patents is bogus in 99% cases for physical patents and 100% cases for software.

Film: "Patent absurdity"

dlang — Tue, 20 Apr 2010 08:05:02 +0000

thanks for the link, I will read through it (probably tomorrow)

Colt is yet another good example and Bell is too

dlang — Tue, 20 Apr 2010 06:52:30 +0000

IIRC he invented (or supervised the invention of) many other things, the phonograph and movies to name a couple.

Film: "Patent absurdity"

dlang — Tue, 20 Apr 2010 05:17:30 +0000

no, unfortunantly it's hard to prove if the advances in technology are due to patent law, general society, political policy, or everything combined.

like most things in history we don't get a chance to do controlled experiments.

Colt is yet another good example and Bell is too

Wol — Mon, 19 Apr 2010 22:27:30 +0000

What did Edison invent?

Oh -and if you're talking about the light bulb, it wasn't that! Edison patented a light bulb that DIDN'T WORK. The light bulb he made his money from was a copy of Joseph Swann's, which he came across - in production - in England some two years before he filed his own patents.

(Oh, and Swann's bulb was patented - Edison ripped him off!)

Cheers,
Wol

Film: "Patent absurdity"

dark — Mon, 19 Apr 2010 10:47:36 +0000

Have they proven this? Where, and when?

In their book Against Intellectual Monopoly, Boldrin and Levine examine a series of situations that should have demonstrated the innovation-boosting effects of patents, if there were any. They looked at countries before and after patents were extended to cover a field of business, and they looked at neighboring countries where one had patents and the other didn't.

In no case did they find increased innovation from patents. They did find cases where innovation went down, and some cases where the blooming of an entire industry was delayed until certain critical patents expired.

It seems obvious to me. The drawbacks of patents are easily visible, and the benefits remain speculative. Patents harm innovation by discouraging competition and imposing fees on incremental improvement.

No, they don't

tzafrir — Mon, 19 Apr 2010 10:20:25 +0000

Worked well, indeed:
http://en.wikipedia.org/wiki/Perpetual_motion#Patents

Film: "Patent absurdity"

Los__D — Mon, 19 Apr 2010 09:52:24 +0000

Industrial espionage is illegal in most countries, reverse engineering is not.

Why should we need protection against actions that are already illegal?

Film: "Patent absurdity"

mjthayer — Mon, 19 Apr 2010 08:31:47 +0000

>so you think the only advantage anyone should get for a new invention is whatever head start they can get on the competition, and other than that everyone should be able to freely use any ideas that they can get their hands on

Are you arguing in favour of rewarding the individual (or company) or because of the benefit to society? If the second, is there any reasonable proof, not just based on assumptions and logic, that patents do bring the benefit they are supposed to? You know the thing about premature optimisation :)

No, they don't

khim — Mon, 19 Apr 2010 06:54:37 +0000

I am not saying that patents work well for software (in fact I think they don't work for software), but I do think that they work well in other fields.

Bogus patents are norm, not an exception and they often protect obvious and simple things. System break-down started long ago - it's just with software patents it's more obvious.

patents work in many fields, and have worked for centuries, the fact that they don't work in a new field (software) doesn't change this.

Sadly they start to work poorly even in places where they worked for centuries. Where before we got patents for the process of making colored glass (thing which eluded people for centuries) today we are getting patents about how to arrange flowers in a vase!

Colt is yet another good example and Bell is too

khim — Mon, 19 Apr 2010 06:37:54 +0000

As was pointed out by others Colt is yet about example of how original inventor got no benefits out of the patent. If you take look on the huge "patent fights" over easily reverse-engineerable patents you'll see that rarely original inventor get serious money to continue his work (Edison is rare exception). In most cases the people who buy patent for the cheap and then exploit money benefit - or in some rare cases few guys invent something and only one profits out of the invention.

Patents like this are good for extortion but bad for progress: patents which cover results of expensive research are rarely fought over. It's literally cheaper to buy the license rather then reverse-engineer.

Film: "Patent absurdity"

bojan — Mon, 19 Apr 2010 06:24:24 +0000

Yes, I think I understand the theoretical underpinnings of why all patents should be wonderful things. In practice, however, many of them are used just to squash unwanted competition and do not contribute to innovation one bit. Software patents and patents on business methods are prime examples of this.

Compare the situation to, for instance, general principles embedded in our (western) laws. Most times we err on the side of caution when someone's rights are to be squashed to gain something else. With patent law we do the opposite - we let broad monopolization of ideas to gain a very dubious outcome.

So, I remain sceptical whether we should be giving these things out like lollipops. It seems like suppression on a grand scale.

As for trade secrets, they exist right now and are enforced right around the world. If they were the all powerful tool to keep everyone out, create monopolies and endless profits, companies would just use them instead of patents, now wouldn't they? However, they don't last - people are inventive. They either figure out what you're doing or they come up with a better way.

So, no, I don't think that use of trade secrets would ultimately have a monopolistic effect. And if it would, we could simply adjust the duration of trade secrets via legislation. That would be really simple.

You can't - and that's the point.

khim — Mon, 19 Apr 2010 06:22:57 +0000

as a trivial example, consider the safety pin, how could you sell it and keep it's design a trade secret?

You can't - and that's the point. Patent system was already broken back then - it was just not so obvious. Take look on these ancient fibulae. Do they look similar?

What is non-obvious and what can be patented are exact proportions - if you spent money and time, performed and a lof of investigations and found uber-effective and good proportions for pin then others can buy the rights from you or use other, less effective proportions (few percents from optimum is Ok, I think).

If it's the replacement for trade secrets then why it's not declared so?

bojan — Mon, 19 Apr 2010 06:05:16 +0000

I don't think Colt is a very good example of a suffering inventor or a struggling business, as per http://en.wikipedia.org/wiki/Samuel_Colt:

"At the time of his death, Colt's estate, which he left to his wife and son, was estimated to be valued at around $15,000,000."

This is an interesting quote from the same article:

"Colt never claimed to have invented the revolver, as his design was merely a more practical adaption of Collier's revolving flintlock, which was patented in England and achieved great popularity there."

That's the thing about copying - it is very, very useful :-)

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 06:04:03 +0000

as I see it the problem isn't using other people's ideas, it's giving people enough of an incentive that the develop those ideas to where they are useful.

the entire idea of the patent system is to make it so that those ideas are available for other people to use freely after the patent expires. If there was not such a system then there would be the risk of such ideas being lost, or locked up through trade secrets (with all the inefficiencies involved with maintaining such secrets, even where they are possible) for much longer time periods.

Our world today is what it is due to technologies that could have been locked up with trade secrets that are instead now freely available to use. Think of such things as how we make Steel or Aluminum and consider what would have happened if those were trade secrets of one particular company, with every employee being locked up under contract to not reveal the secrets.

Those are a couple examples where the process could be kept secret while selling the result. Prior to the patent system it would have been likely that such ideas would have died with the inventor, and had very little effect outside the immediate area (remember that if secrecy is your only protection against being flattened by your rivels you don't want to let many people in on the secret, which limits how large you can grow)

by giving out a temporary monopoly the patent system is designed to trade the short term advantage to the inventor for the long term advantage of society by making the inventor document the details of the invention. In the worst case (where the patent owner uses the patent to squash all use) society gets the invention a generation later, in the best case, not only does society get the invention free to everyone a generation later, the inventor is able to make good use of it in the meantime.

Film: "Patent absurdity"

drag — Mon, 19 Apr 2010 05:48:32 +0000

> in the software field this is the norm, in other fields I believe that it is the exception.

Yes. Patents have gotten out of hand in other industries, also.. but with software in particular it's a complete disaster.

Patents themselves have proven that they work and are effective at promoting inventions and inventors in the past, but like everything else in the USA government it has spiraled out of control.

Film: "Patent absurdity"

bojan — Mon, 19 Apr 2010 05:43:24 +0000

> after all, if reverse engineering is acceptable, why not industrial espionage

Because industrial espionage is usually a breach of contract or some kind of law. Reverse engineering is not.

Patents come from a very dubious place - a very old law called Statute of Monopolies, which essentially limited crown given monopolies to new inventions only. I'm sure it was very progressive in 1623. Whether this remains the case in 2010 is a subject of many debates.

I can understand that there could be areas that are helped by giving out monopolies. However, we should do that sparingly. Current practice where every man and his dog can patent practically everything has gone way out of hand.

PS. I'm not sure why there would generally be a problem with using other people's ideas. Humans learn by copying. If we were stripped of that ability, we would not be ourselves any more. I cannot see why it should follow that whenever someone has an original thought, it should not be available to others freely. In fact, in many fields of endeavour this is the norm.

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 04:43:42 +0000

so you think the only advantage anyone should get for a new invention is whatever head start they can get on the competition, and other than that everyone should be able to freely use any ideas that they can get their hands on, no matter how they get them (after all, if reverse engineering is acceptable, why not industrial espionage)

If that's your position we will just need to agree to disagree.

Film: "Patent absurdity"

bojan — Mon, 19 Apr 2010 03:40:34 +0000

> say Hyundai developed a new engine design that gave them twice the power and MPG of existing engines for the same cost, weight and size. Don't you think that they should be able to license the design rather than just have it copied by Mercedes?

I think they should be able to come to the market with it first. Which they could, if they played their cards close to their chest.

The others would not able to use that immediately. They would have to figure out what it was, make prototypes, test, build production to do it etc. This takes time. In the meantime, Hyundai would be taking away their market share, provided they ran their business properly.

And again coming back to Nokia v. Apple, Nokia _is_ trying to do what Apple are doing. They have a music store, they have an application store, they have and control smartphone OS (several, in fact). But the problem is that Apple right now does it _better_, so they win. Same with everything else.

There are many products that have these great ideas in them, but otherwise suck. Most people stay away from them, in my experience.

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 02:38:43 +0000

flip the example around.

If Mercedes had the ability to use all of Hyundai's patents at no charge, how would Hyundai fight back?

say Hyundai developed a new engine design that gave them twice the power and MPG of existing engines for the same cost, weight and size. Don't you think that they should be able to license the design rather than just have it copied by Mercedes?

As for the Nikia v. Apple, I'm not sure what in the iphone should be patentable. it's good engineering, and it has nice polished software in it, but is that really novel enough to qualify for a patent? the engineering isn't patentable, they are assembling components purchased from others and packaging them nicely. As for the software, what is in the software that hasn't been done somewhere else by someone else first? They may have been the first to put multi-touch into a phone, but how is that substantially different than multi-touch interfaces on another computer system?

Not every successful product needs to have a patentable invention behind it.

If it's the replacement for trade secrets then why it's not declared so?

dlang — Mon, 19 Apr 2010 02:27:01 +0000

the example was one that people can easily see can be trivially duplicated and trade secrets cannot protect you.

if you want an example where there was a much larger fight, but similar (if less clear) inability to protect against reverse engineering, take a look at the colt revolver patents. I don't remember the full history (and I am not taking the time to look it up right now), but I seem to remember that the initial inventor attempted to make a go at it before selling out to a larger manufacturer (and if I remember, he sold out because people were infringing on his patent and he didn't have the resources to fight them), but even that larger manufacturer waged patent fights for years.

Film: "Patent absurdity"

bojan — Mon, 19 Apr 2010 02:09:21 +0000

> think auto manufacturers, one may have a new idea, but if it works, every other company will have it in their vehicles the next year

Great example. Now, let's take Hyundai and Mercedes-Benz, for example.

Let's say that Hyundai can take all of Mercedes-Benz patents and use them tomorrow, free of charge. Do you honestly think that they would be making the same cars as Mercedes-Benz, just by having those patents available and for the money you pay today for current Hyundai models? That they could just take over that part of the market by virtue of having the same patent pool? I just don't see that happening in a hurry. It takes a lot more than that.

Coming back to the Nokia v. Apple thing. You know, Nokia have N900, they have X6, they have N97 and probably a half dozen other models. And yet, none of them are iPhone. It just isn't the same, because there are so many things to a great product that has charisma, not just patents.

If it's the replacement for trade secrets then why it's not declared so?

bojan — Mon, 19 Apr 2010 01:53:27 +0000

Yes, that is a dilemma for an inventor to decide upon. Do I go it alone or do I sell my invention a sufficiently large manufacturer?

> take the example of the safety pin, it was very non-obvious to come up with, but very easy to reverse engineer.

Walter Hunt was happy to get $400 for the invention at the time (which would be around $10,000 today). If he were to sell this to a manufacturing company at the time, instead of the man he owed the money to, do you not think the manufacturer would make a sizeable profit on this before others moved in? And, they could also be the ones owning the trademark (if they so wished), making the "original" safety pin. And all that for measly $400, or $10,000 in today's money.

I don't see the significance of this example.

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 01:33:07 +0000

huge patent portfolios are only a problem if the patents are really non-obvious enough that people infringe on them without knowing about it.

in the software field this is the norm, in other fields I believe that it is the exception.

one problem is that due to the fact that the patent is published and is (by definition non-obvious), the legal assumption is that if you use the invention you learned about it from the patent. It would be nice if there was some way to prove independent invention and have that nullify a patent, but proving that you didn't hear about something is a hard thing to do. It may work if you have a large research staff and can show years of work on a subject, but it can't work for many other ideas (again the safety pin example, how could you prove that you thought up the idea independently). This is one of the trade-offs in patents, you occasionally punish independent invention in exchange for getting the idea out into the public for other people to use (instead of them being trade secrets) and getting new ideas developed (thereby encouraging new inventions).

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 01:24:10 +0000

copyright isn't enough to protect software. you don't have to do literal copying, a clean-room re-implementation of the algorithm would clear copyright obstructions while still implementing the invention.

in terms of other fields, it doesn't even need to be david vs goliath, any new idea is a risk.

again, using the safety pin example, even a large company is taking a significant risk in gearing up and pushing a new product, there are always other large companies that could dilute, if not beat you on any one product. in some cases it may be size related (say Vons supermarkets vs Wallmart), but in other cases it may just be that you have enough competitors, each of who could copy your success without having to take any risk of failure (think auto manufacturers, one may have a new idea, but if it works, every other company will have it in their vehicles the next year)

the way to eliminate patents on things like anti-gravity machines is to require a working model be available for examination and testing. the patent office used to have this requirement.

Film: "Patent absurdity"

dlang — Mon, 19 Apr 2010 01:15:14 +0000

I am not saying that patents work well for software (in fact I think they don't work for software), but I do think that they work well in other fields

the problem with patents in software is that there is so much independent invention that is not documented in any way that the patent office could check, and the patent office is not qualified to determine if something is 'obvious to someone ordinarily skilled in the art'

add to this the contoured legalese that the patent discousures are written in and you end up with a situation where even the inventor can have trouble understanding what the patent is describing.

I agree that software patents are a disaster and need to be killed off, but when talking about the problems with software patents you need to be careful to say you are talking about software patents, not patents as a whole (unless you really are intending to talk about patents as a whole)

patents work in many fields, and have worked for centuries, the fact that they don't work in a new field (software) doesn't change this.

Film: "Patent absurdity"

coriordan — Mon, 19 Apr 2010 01:05:31 +0000

The problem is, when IBM has 12,000 patents, and the small inventor has a handful, and the IBM infringes one of the small inventor's patents, IBM can always use their portfolio to force a cross-license. So the small inventor is never protected against competition. (RMS presentation)

And that's the rare case where the small inventor has the desired patent. Usually it's the megacorp that has the desired patent, and the small inventor can't force a cross-license, so he either gets told to flat take his product off the market, or start paying a patent tax.

I also have to disagree with the idea that the number or quality of patent today is the problem. I think the opposite. IBM wants the free software community to think that quality is the problem, so that people will support their ineffective, time-consuming patent projects such as DIY-patent-review.

I think the best example of the real harm caused by patents is MPEG H.264. There are 900 patents, held by 29 companies in 50+ countries. No amount of raising the standards will ever solve the patent problems that mplayer, xine, GNOME, KDE etc. face. Complete abolition is the only reasonable solution. Anything else is just hand waving.