|
|
Subscribe / Log in / New account

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Out-Law.com is carrying a critical response to Linus Torvalds on European software patents. "Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions. Allowing for patent protection on software inventions is a requirement of the World Trade Organisation's TRIPS agreement which states that patents must be available in all fields of technology." (Thanks to James Heald).

to post comments

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 15:42 UTC (Wed) by jedidiah (guest, #20319) [Link] (2 responses)

This joker seems to have a fundemental misunderstanding of what the position of Linux is. While I can't speak for Linus, many of us believe that intellectual property should exist to promote creativity and should never be used as a blunt object to suppress creativity.

We don't want to allow for artificial monopolies on software inventions.

We've been doing fine without them.
Don't try to fix what is not actually broken.

The point of patents (or any law really) is to serve some internal policy objective, and not to maintain compliance with some international treaty. Any lawyer should understand this and be willing to acknowledge this.

The justification for patents should be a question of whether or not they benefit society as a whole, not whether or not they are compliant with some treaty.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 20:00 UTC (Wed) by ccchips (subscriber, #3222) [Link]

As a former member of the National Organization for the Reform of Marijuana Laws in the U.S., I have one warning.

Keep your eye on jokers like these.

A really close eye.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 3, 2004 7:32 UTC (Fri) by alexs (guest, #13637) [Link]

lets try to be compliant with some treatys... ;-)

allow patents for drugs, for mechanical machines,
for any sort of apparatus, for an electrical component,
for the software prduct that you can buy in a shopt -
that means patents for just for any prodoct that is
new for an expert of the subject, inventive at a non trivial degree
and teaches you something new of the applied science of physics.

thats what is roughly requested by TRIPS - and you can still
tune the three criteria for an individual country.
lets say the USA possibly has a rather "trivial" level for patents.

BUT ideas, algorithms, boolean logic, math, business practices
and all alikes are not covered by the applied physics.
other than that its hard to proove that some specific idea
in such non materialistic fields was never previousely used.

patenting of such areas is therefore neither requested by TRIPS
nor is there a reasonable way for the patent office to make sure
that there was never any prior art out there. ideas are thougts
and therefore need some sort of object to materilize for proof.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 15:46 UTC (Wed) by hummassa (subscriber, #307) [Link] (16 responses)

Mathematical discoveries are not inventions.

Algorithms = Math.

Thanks $DEITY my country explicity prohibits patenting of "computer programs".

OTOH, the article starts with "patent law firm says..." Oh, yeah, a patent law firm opposes something being non-patentable. Very, very, surprising.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 16:23 UTC (Wed) by josh_stern (guest, #4868) [Link] (15 responses)

I'm in favor of far less patents than we have now, but I don't follow the
logic of why mathematical discoveries are any less inventions than
discoveries in other fields of endeavor. New mechanical or chemical
discoveries do not come about because the laws of physics have changed -
people simply notice the possibilities of new ways of arranging stuff,
and that is also what happens in math. In fact, in math it is easier to
describe new realities, since the only constraint is the internal logical
consistency of the description itself.

The problem with patents is that a patent takes away rights that people
had before the patent was granted and this should only be done for very
strong reasons of societal benefit. The system has become skewed by
people with the false presumption that any patent is somehow good for
business and that people have some natural property right to prevent
others from "using their ideas" whether the other people thought of them
independently or not. This ignores the fundamental difference between
preventing others' usage of a finite resource - material property -
where the usage effects the primary holder and others' usage of an
infinite resource - intellectual property - where the usage has no
direct effect on the original holder but may indirectly provide
commercial competition (generally a good thing).



Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 19:44 UTC (Wed) by pointwood (guest, #2814) [Link] (3 responses)

Why shouldn't patents be allowed?

Well, one good reason is that *all* reports/studies/etc. have concluded that *at best* software patents wont have a negative influence, but most likely they will.

All the pro-software patents people haven't been able to show even *a single study* that says software patents is a good good thing. That's pretty telling don't you think?

The only people that says software patents are a good thing are patent laywers and the other few (mostly huge companies) that will be able to earn money on it.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 22:25 UTC (Wed) by allesfresser (guest, #216) [Link]

s/earn/extort/

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 23:12 UTC (Wed) by josh_stern (guest, #4868) [Link] (1 responses)

We are basically in agreement. Any type of patent, including a
software patent, should only be granted if patenting that type of
thing is defensible by some strong argument of societal benefit, and
you are saying that there is no evidence for benefit from software
patents. The basis for the idea of societal benefit would be the claim
that the invention would not have happened without the promise of
financial award that the patent can help realize.

The only place I might disagree is about the strategy of how we should
practically categorize what type of patent leads to societal benefit. I don't
think the categorization should be based on field of endeavor, but
rather on whether a strong case can be made that a particular discovery
would not have been obtained without a large prior investment of money
(or something inter-translatable with money) directly supporting the
research (maybe the length of the patent should even be proportional
the required investment size). So, for example, I'm mostly sympathetic
to the drug patenters that require large investment to find new successful
drugs and get them approved for safety. But if someone finds the
next miracle penicillin mostly by accident, they shouldn't get a patent
on that just because it is in the medical field. Similarly, if there was
some software discovery that really required enormous investment
to obtain that went beyond regular product development, then it
should be eligible for patenting. Something that could readily
be discovered by a volunteer developer would be excluded
by design.



Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 7:11 UTC (Thu) by pointwood (guest, #2814) [Link]

The idea might seems nice, but handling that would be a huge "mess"...

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 19:47 UTC (Wed) by rriggs (guest, #11598) [Link] (3 responses)

I don't follow thelogic of why mathematical discoveries are any less inventions than discoveries in other fields of endeavor.

There was a point in time where science or "discoveries" (information about the way our world works: math, chemicals, etc) was not patentable. Only engineering works or "inventions" (tools, gadgets, etc) were patentable.

The following example was given to me. Even if you create a chemical that has never been seen in nature, it is still just a chemical and therefore in the realm of science. You cannot patent it. What you can patent is the process used to create the chemical (the engineering bit).

I don't think the patent office thinks this way any more.

My problem with software patents isn't that they are bad in nature. They are only bad in practice. Software patents, to me, do fall in the patentable category described above. But the way software patents are actually granted today in the US, they are just ridiculous. The US patent office has probably done more to discredit software patents than the open source movement ever could hope to.

Chemicals *are* patentable

Posted Dec 1, 2004 21:34 UTC (Wed) by hazelsct (guest, #3659) [Link] (1 responses)

Actually, in the U.S. at least and I believe most other industrialized countries, arrangements or states of matter such as chemicals and metal alloy compositions *are* patentable.

Chemicals *are* patentable

Posted Dec 1, 2004 22:53 UTC (Wed) by glenalec (guest, #26113) [Link]

I think rrigs' point was that chemicals et. al. weren't always patentable,
not that they are not today. (I'm not commenting on the accuracy of this
claim, though, as I have no knowledge in this field.)

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 23:20 UTC (Wed) by josh_stern (guest, #4868) [Link]

The distinction between a process and a static doesn't really do any work
here (even if one buys into the idea that the static is somehow natural
and the process man-made) because a program can also be seen
as a process to create specific configurations of data or machine
states (take your pick).


Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 20:07 UTC (Wed) by jd (guest, #26381) [Link] (5 responses)

An invention creates something that is new, that had no existance before. A discovery transforms something that already exists (eg: gravity) into something else, but distinct (eg: Newton's laws of motion). Technically, the transform is an invention, as that did NOT exist before, but the existing system is concerned with the end result and not with how you get there.

To my way of thinking, the only valid software patent would be a patent on the translation process from an abstract requirements specification - ie: what you want to do - into the software itself - ie: how the computer is going to go about doing whatever it is. That translation is a genuine invention, it is unique, and I see no opposition to patenting that.

The end product - the software - is just a different representation of the requirements. A program to add two numbers is merely a representational transform of the need to perform that operation. It's not a new thing, it is merely a new form. A vector is a vector is a vector, whether it's written as a matrix, drawn as a diagram, stored on a computer, or imagined by a mathematician.

A car is an invention, because although it is just a re-arrangement of existing components (the Watt engine, wheels, etc) it had no pre-existing form. It is a transformation, rather than a translation, of something else. Gravity existed whether people observed it or not, but the car didn't exist at all before it was built.

I would also claim there are different degrees of invention. Not all transformations are equal in magnitude. Some are based on pre-existing information (eg: the mathematics of fixed-wing flight pre-existed the first aircraft), others work largely from scratch. Some are simple transformations, others are highly complex and demand considerable thought. At present, we just lump them all under one label - invention - and one type of protection - the patent.

Patent lawyers who argue that they should have more power, over areas they have no historical right to, are merely arguing they (and their employers) should have a fundamental right to income. Since there is nothing new in an algorithm - the algorithm as-is has always existed - the IP protection can't be about the creation of something, only about the claiming of it. And if you can own something through the mere act of claiming it, then it can only be seen as a fundamental right. This is how copyright works, for example.

Personally, I don't like the idea of "Intellectual Property", but if we have to suffer with it, we need to simplify the existing mess and illuminate what is protected and why by adding more categories whose protections reflect the nature of what is being protected, not complicate and obscure it by stuffing random bits of some into others.

Apologies if people take this as flame-bait - it's hard to talk about such issues without getting into areas that are, well, controversial. If you disagree with any of the above, I'd like to hear your perspective. I can't learn anything new if I only hear my own views from myself and others. Where's the fun in that?

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 21:04 UTC (Wed) by niner (subscriber, #26151) [Link] (2 responses)

Just a little question for understanding your view: are you a programmer? And I think of the heavy stuff, not just some little scripts or small programs that satisfy some use.

No offend ment, just asking.

Coding styles

Posted Dec 2, 2004 1:16 UTC (Thu) by jd (guest, #26381) [Link] (1 responses)

Yeah, I'm a programmer. I do most of my programming in C, but I can also write "cold" in 80x86 assembly, ADA, Java, Pascal, Perl, Python and PHP. If the reference manual's handy, I can also do Prolog, Occam, Fortran, COBOL and SISAL. I'm learning D and PL/1 for fun. I tend to think more in the Software Engineering side of the field - I can read/write Z specifications - and this does influence my perspective.

Z Specifications, EBNF, Jackson Structured Diagrams, Flow Charts and Source Code all show the same data, it's simply which way you rotate the paper and link the elements. In fact, if they didn't, they wouldn't be useful. If you couldn't change the view to one that was suitable for what you wanted to do, without changing the thing you're viewing, it would be absolutely horrible to maintain anything.

Having been in the computer industry, professionally as a programmer, for over a decade, and having been programming in general for 25 years, I know just how bad it gets without proper documentation and source control. However, that doesn't mean you couldn't engineer such documentation from any source you had. Actually, this would make maintenance a whole lot quicker - debugging a spec, and then translating that into the code, is quicker than debugging the code directly.

Thank you

Posted Dec 2, 2004 12:59 UTC (Thu) by niner (subscriber, #26151) [Link]

So you obviously know what you're talking about. I was just asking, because I'm still undecidet about how I should think about your statement. Although most programs are just an implementation of some specification, of some things I could think of as an invention. Maybe some program languages themselves, something like regular expressions. Maybe something like the first compiler.

Certainly nothing I ever did, but I could just imagine. One could argue, that really everything came out of a need (like making programming easier) and nearly everything existed in some way before. But the same is valid for nearly ever patentable "invention" out there. Most are just some way X to reach goal Y more efficient, cheaper, whatever.

I'm certainly not for software patents. I'm not even for any patents, but it's an interesting question just what an "invention" really is.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 23:25 UTC (Wed) by josh_stern (guest, #4868) [Link]

Practically every software program creates a process for creating a modestly
unique spatio-temporal mapping between its inputs and its outputs. Nobody
cares about that, but I am just trying to show that this kind of highly abstract
word play for describing what should be patentable doesn't get us anywhere.

Invention is overrated...

Posted Dec 1, 2004 23:33 UTC (Wed) by grantingram (guest, #18390) [Link]

I'll avoid your comments on what should constitue a patent and what shouldn't as that is way too deep for me. This however sparked my interest:-

... Not all transformations are equal in magnitude. Some are based on pre-existing information (eg: the mathematics of fixed-wing flight pre-existed the first aircraft), others work largely from scratch.

I've never seen an invention that was worked largely from scratch. Every single invention I've seen has been a derivative of something that went before. That's not to diminish the achievements of inventors - it's simply the way the world works.

Everyone stands on the shoulder of giants, some people just don't look down!

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 3, 2004 7:18 UTC (Fri) by alexs (guest, #13637) [Link]

patents do cover inventions in the field of physics.
but since when the science of math is a subsection of the science of physics?

What is a software invention?

Posted Dec 1, 2004 16:02 UTC (Wed) by libra (guest, #2515) [Link] (3 responses)

Can that attorney be kind enough to define what a software invention could possibly be?
I have seen algorithms and mathematical formulas that are worth patenting, but I don't think to have seen a single software invention since I've started to work with computers.
I think such a thing can not exist unless it is defined broadly and in terms that hurt common sense and global knowledge of the IT world.

Beside I would like a law being passed that forces any software retailer to give all the details necessary to build the algorithms required to access any data stored by the softwares he sells. If some patent are involved and prevent others to use those algorithms it is fine, as long as I'm free to implement them for myself in case I want to retrieve my data through that method and not the only one provided by the software maker.

I want to remind software makers that while they own the code of their software, we own the data that we decide to store with it. I think it shall be unlegal to not give me a way to access MY data without using only their software. Software are to be sold to give you tools to work on your data, not to force you keep using them once you start having your data locked inside.

As some software makers are unable to figure that themselves, I think a law shall be passed to give them the clue.

What is a software invention?

Posted Dec 1, 2004 19:19 UTC (Wed) by XERC (guest, #14626) [Link] (1 responses)

...Beside I would like a law being passed that forces any software retailer to give all the details necessary to build the algorithms required to access any data stored by the softwares he sells....

So would I. As we all know, the secrecy of the M$ Office file formats
is the main thing, that keeps the
Office(=> therefore also Windows) license
sales going. Considering the fact, that Office and Windows seems to
be M$ main and almost only revenue(What Money Can't Buy ),
then a really fine and full crack of the file format would be
a very, very, considerable blow to Microsoft, at least, in
an environment, where software patents are not enforceable.

I wish someone had time to build a torpedo
for M$ by cracking the Office file formats.
It certainly would do a big bang.

MS Office file formats

Posted Dec 1, 2004 22:34 UTC (Wed) by danielthaler (guest, #24764) [Link]

It certainly would do a big bang.

I disagree. While they are horribly complex MS Office file formats are fairly well understood; for older Versions there is actually documentation from MS (on wotsit.org).
OpenOffice does a good job of displaying such files; the reason it isn't perfect is that OO simply can't use all the information in other formats.
So that probably isn't the reason people continue to use MS Office

How is a software invention distinct from software code?

Posted Dec 2, 2004 1:16 UTC (Thu) by kirkengaard (guest, #15022) [Link]

This seems to be a distinction that is easy to make if you don't understand software except by analogy. Implementations are patentable, where ideas are not. (yes, oversimplification) All programming starts with a problem, around which you build an algorithm, which algorithm is implemented in a language, creating source code, which is compiled into binary form. It seems to me that problems are universal; there may be a range of possible algorithms usable to solve any given problem, and so an algorithm may be unique, but is itself an idea, and not patentable; the code used to implement the algorithm is the only "patentable" thing in the chain. The code is the invention.

“Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions."

Come again? How, then, is copyright not sufficient? I seem to have missed that one...

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 16:06 UTC (Wed) by kleptog (subscriber, #1183) [Link]

I like this bit:

Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions.

It probably never occurred to them that we don't want patent protection. That copyright is enough.

This, says Dr Collins, will mean that many innovations in protocols in the telecoms industry, for example wireless-to-wireless communication or wireless-to-landline communication, could be copied without infringing the patent.

The problem being? I'm beginning to question the point of patents anyway, businesses don't seem to need "encouragement" these days. I think the market is a good enough driver.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 16:27 UTC (Wed) by QuisUtDeus (guest, #14854) [Link]

And who says that the WTO deserves any attention, other than to see it as the business/economic arm of the international bankers. It is an organization that owes its existence to politicians who are willing to sell out their own countries' sovereignty for a fast buck.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 16:47 UTC (Wed) by sandy_pond (guest, #9734) [Link] (16 responses)

Torvalds and his supporters lack a fundamental understanding of intellectual property rights

Oh contraries. Me thinks that patent lawyers and their supporters lack a fundamental understanding of the software development process. Patents just don't work correctly for software. Any more than patents work for math or language.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 17:12 UTC (Wed) by forthy (guest, #1525) [Link] (15 responses)

I start to feel that patents don't work. Full stop. No matter in what
field. Take pharmacy. A few years ago, a company obtained a patent for NO
(nitrogen monoxide). NO is a gas that's cheap to produce, and that helps
babies with respiracy problems. Now, being patented, the company wanted
$2000 per day, instead of the $70 the therapy did cost before. After some
years, the EPO finally found out that this patent was null and void, and
now the price is back to $70. There are a lot of other examples in
pharmacy, where patents are abused to the same result. People nowadays
even patent genes, things they have read in the code of life. No invention
whatsoever.

The conclusion is: The patent system is now so much broken that we all are
better without it.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 17:44 UTC (Wed) by simlo (guest, #10866) [Link] (2 responses)

I start to feel that patents don't work. Full stop.

I agree. I work for a major wind turbine producer. We can't sell our best product in the US and Canada because of a patent by Enron, now bought by GE, on a certain kind if converter technology in wind turbines. Therefore the same equipement which can produce a power of 2.0 MW, can only produce 1.8 MW in North America. It must be pointed out that the same technology was used here in Denmark long before Enron even thought of producing wind turbines!

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 20:54 UTC (Wed) by emkey (guest, #144) [Link] (1 responses)

Doesn't that mean there was prior art though? Thus the patent should be void...

It would likely cost a lot of money to prove that though.

Patents are regional...

Posted Dec 1, 2004 23:23 UTC (Wed) by grantingram (guest, #18390) [Link]

Well I'm not an expert on engineering patents but I did go to a talk once....

I think it is perfectly possible to have a patent in the UK that doesn't cover the USA for example. The UK Patent office site seems to imply that this is the case.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 18:11 UTC (Wed) by mrshiny (guest, #4266) [Link] (10 responses)

I agree that patents can be misused in pharmacy, but I can assure you that many drugs wouldn't get developed in pharmacy because the cost to design the drug far, far outweighs the cost to manufacture the drug. The entire value of the drug is in its formula; that formula can not be copyrighted and so competing companies can quickly come up with a clone version of the same drug once a drug is produced. So there would be zero incentive to design new drugs if there were no patents to protect them.

Where the patents get abused is when existing chemicals (Drugs) are re-patented for "new indications", that is, using an old drug to treat some illness that was never treated by this drug. I don't know of a concrete case, but my wife (a pharmacist) deals with this all day. A hypothetical example: The drug used to treat male-pattern baldness started out as a blood-pressure drug. If it was patented as a blood-pressure drug, nobody else could make it to treat high blood-pressure. Let's say they discovered that it also treats baldness (which, in fact, they did). Then they can patent the drug as a baldness cure. This means that the pharmacist has to ask the patient why the drug is being prescribed: if the blood-pressure patent has expired, competitors can make a clone drug and doctors can prescribe that. The clone drug will be cheaper but just as effective. However, if the patient wants to use the drug for baldness, they MUST buy the patented, more expensive version: it's the law of patents. And this is garbage because no innovation ocurred, it's just an old drug being used a new way.

(Some would argue that this is indeed innovative, because who would think to use a blood-pressure drug to treat baldness? Anyway, there are plenty of other cases where the same drug is re-patented for almost the exact same indication as before, thus causing lots of confusion and higher costs for patients).

So, I would say that patents are indeed useful in the pharmacy industry; without them only governments would spend the money on things like clinical trials, etc.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 18:56 UTC (Wed) by sandy_pond (guest, #9734) [Link] (2 responses)

"without them only governments would spend the money on things like clinical trials"

Maybe a good thing? We'd certainly end up with less money going into drugs like Viagra and more money going into drugs for less likely but more severe diseases.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 20:08 UTC (Wed) by mrshiny (guest, #4266) [Link] (1 responses)

It's possible that more money would be spent on "serious" drugs but then again, there'd be far less money to spend on it; profitable drugs like Viagra can help fund other drugs. Some drugs are made (For serious diseases) even though no profit will be made. Drug companies aren't all good, but they're not all bad either. We just need to make sure we don't give them too much power or influence... kinda like most corporations, really.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 10:11 UTC (Thu) by XERC (guest, #14626) [Link]

I don't know exactly, what it's like in the
team of biologists, but we, programmers,
develop tools for our corporate biologists and
in our case it's exactly so, that people tend to do exactly
what's more intereasting, more thrilling, exiting and
what they believe, will give as good results as they
can get for biologists---to develop better drugs.

So, I have a reason to believe, that, almost no matter
what the management thinks, there are probably many
people among the biology team members, who also,
tend to sneak to most interesting
tasks, most thrilling tasks. As the biologists
are also humans, the intereasting tasks may very well
be drugs for really serious diseases.

I visited one inter-corporate conference
once, whiches name I leave intentionally unsaid, and
the guys there presented research on such a horrible
NATURAL diseases, that most of us have probably never
even heard of. Yep, and it was a halve-fun project for
them, not something that the management ordered them
to do.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 21:17 UTC (Wed) by niner (subscriber, #26151) [Link] (6 responses)

Some time ago I've read an interesting comment saying the pharmacy industrie spends a multitude of the R&D budget for marketing. Would they really drop a small entry in their budget and losing at least the headstart research gives them when bringing out new drugs?
Or would they maybe even spend more to be the first in many fields instead of just sueing the competition because of patent infrigments?

I think both are possibilities at least worth considering but nothing could be proven unless it is tried.

So I take the old horror stories about medicine would stop leaping forward without patents (interesting, it did develop even before patents where invented) not for granted. It's just a possibility but not likelier than any other.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 2:46 UTC (Thu) by mrshiny (guest, #4266) [Link] (5 responses)

I don't have any data, but my impression is that drug companies DO spend lots of money on advertising. This is, in fact, a problem for doctors and pharmacists, who find themselves constrained when trying to select medications for patients because patients insist on something they've seen on TV. I believe this sort of thing can be handled with legislation targetted at the drug industry. I don't think the same is required for other fields; all cars are roughly equal in saftey (compared to drugs) so the choice of cars isn't a major risk to consumers, therefore let car makers advertise cars. But the choice of drugs is very important, and many people are over-medicated. This is a burden to healthcare systems and is not a benefit to society; therefore an argument can be made that drug companies should be regulated (more). But I certainly believe that there would be fewer drug companies making new drugs if they didn't have patents.

And anyway, patents in the drug industry don't have the same harmful effect as they do in the computer industry. If I need to share data with you, but that data is in a patented format, I'm out of luck unless I pay up for the patent. There are lots of other examples where the software industry is harmed by patents, these do not apply to the drug industry. For one thing, if a new software algorithm is released, someone else might be able to immediately apply it to a novel new idea. However, if it's patented, they have to pay for it. Waiting 20 years for the patent to expire is not feasible. However, with drugs, inventing a new drug, even based on an old one, takes years, then it takes years to approve... in that kind of timeframe a many-year patent is not as big an impediment to innovation.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 4:15 UTC (Thu) by XERC (guest, #14626) [Link]

...then it takes years to approve...

I don't want to say, where I know this, but
I believe that I can assure You that nowadays the
lifecycle of a totally new drug is no more than
6 to 7 years. That is, from the the handfull of smelly chemicals
to the market.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 13:06 UTC (Thu) by niner (subscriber, #26151) [Link] (3 responses)

On the risk of dramatising, but ask some people in the third world countries facing month salary high costs for drugs on the one side and patents prohibiting them from producing these drugs themselves on the other side about how harmful the effects of patents in the pharmacy industrie really are.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 17:09 UTC (Thu) by XERC (guest, #14626) [Link]

I agree, that the third world countries
(India, China, let's not forget Russia,
all of the African countries) need some
reasonable transfer period(20 years?),
but if they enforce pharma-patents immediately,
then they can only blame their own amazingly
stupid or just corrupt or indirectly corrupt
governments and their own culture.

Their own culture? Well, there is a saying that
every country has a government, that it deserves.
Most easily followable examples: U.S.A., Russia.

U.S.A.: With the U.S. don't-care mentality, one gets
exactly the kind of system, where everybody
sues everybody, poorer averagely talented
people can't get higher education(the few brightest
ones can get a free grant, but that doesn't chance
the big picture) and many people are without
any kind of medical aid. And yes, this system
is at least partly with a positive feedback:
number of court cases increases, duller citizens
vote for duller president candidates(and put up
duller candidates to vote for), duller
leaders(presidents) make duller decisions, which in
turn result something bad.


Russia: For a non-russian person with some intelligence,
who has been living in an an area that was one of
the states of the former(and future?) Soviet
Union, it's easy to understand, that there will
be no democracy in Russia, EVER, no matter what
the government or president. I might be mistaken
here, but I guess that at least the last 500 years
of Russian history supports this statement.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 17:43 UTC (Thu) by josh_stern (guest, #4868) [Link] (1 responses)

It's a tough issue. The premise of the drug patent is that the capitalist company would not have invested to create the drug in the first place without the possibility to get a time-limited monopoly from the patent and profit from that. Companies do regularly set different prices for their drugs in different parts of the world, and then they also get criticized for that (and people in the U.S. want to import the lower priced foreign versions). So there is an interesting moral question about what pricing should be used for a life saving patented drug once it has been created, or where should subsidies come from. I think it would make sense for the patent system to give the govt. the option to buy out the patent in certain cases, like a life-saving drug.

Having said all that, the comment above reads like basically a simple denial of the premise: the patent is bad because it prevents a lower price on a drug that would have been created anyway. If that is the problem then we don't have to go to the examples of life and death because every patent is bad if it covers something that would have been created anyway without the special investment.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 4, 2004 17:28 UTC (Sat) by XERC (guest, #14626) [Link]

without the special investment.

The point is, that very often the market is
a sufficient motivation for making the investment.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 14:10 UTC (Thu) by Seegras (guest, #20463) [Link]

Yes, I second that. I get this feeling too, and Thomas Jefferson was sceptic if patents would be "the right thing to do" as well.

And from my extra-strong libertarian point of view, a patent is nothing more than a government-granted monopoly; thus an unnecessary intervention of the free market; ergo evil ;).

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 17:15 UTC (Wed) by jwb (guest, #15467) [Link]

Unfortunately, the TRIPS agreement violates the MOUNDS agreement, which states that every patent holder must send me a delicious chocolate-covered coconut candy bar.

Oh, you're not a party to the MOUNDS agreement? Coincidentally, I'm not a party to the TRIPS agreement. "Treaties" made in smoky backrooms between 2nd-rank corporate executives have never seemed too persuasive.

TRIPS argument is wrong

Posted Dec 1, 2004 17:38 UTC (Wed) by fermigier (guest, #12330) [Link]

http://www.patent.gov.uk/about/ippd/softpat/1000.htm

Speaker: Paul Hartnack, Comptroller General, The Patent Office

Powerful arguments have been advanced in favour of a change in the law which would allow software to be patented. Most of you will be familiar with the Agreement on Trade-Related Aspects of Intellectual Property Rights - commonly known as TRIPS - which was signed in the Uruguay round of the GATT negotiations.

Some have argued that the TRIPS agreement requires us to grant patents for software because it says "patents shall be available for any inventions.....in all field of technology, provided they are.....capable of industrial application". However, it depends on how you interpret these words.

Is a piece of pure software an invention? European law says it isn’t. Is pure software technology? Many would say no. Is it capable of "industrial" application? Again, for much software many would say no.

TRIPS is an argument for wider protection for software. But the decision to do so should be based on sound economic reasons. Would it be in the interests of European industry, and European consumers, to take this step?

See also: http://swpat.ffii.org/analysis/trips/index.en.html

The fundamentals...

Posted Dec 1, 2004 18:18 UTC (Wed) by emkey (guest, #144) [Link]

My stance on this general topic is fairly basic. From what I understand historically the concept of patents was put in place to protect the little guy. Over time it has come to do just the opposite of the original intent.

Even if I grant the need for patents I have to question the ease with which they have been granted in the US. We've essentially had thousands of patents for things as simple as the wheel.

Says attorney

Posted Dec 1, 2004 19:07 UTC (Wed) by dkite (guest, #4577) [Link]

Of course an attorney would say this. This is their bread and butter. Imagine being in such a powerful situation where everything that everybody does in a technology company has to be vetted by you. At whatever price you want. This is what software patents do.

Linus understands perfectly well. FOSS has been successful at taking power away from the marketing people. Now the lawyers want it.

Derek

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 19:32 UTC (Wed) by XERC (guest, #14626) [Link]

The article is 100% demagogy. The lawyer is not that stupid.

Anti-software-patent lobby should pay attention

Posted Dec 1, 2004 21:55 UTC (Wed) by dwalters (guest, #4207) [Link] (2 responses)

This article is interesting because it shows some of the arguments that the pro-software-patent lobby may well be using when trying to persuade European politicians.

So the anti-software-patent lobby should be mindful of Dr Collins' comments when formulating their counter arguments.

The notion that "Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions", is almost certainly wrong. I'm sure most software authors, myself included (and almost certainly Linus too) are quite aware of the fact that patents don't extend to so called "software inventions". Dr Collins' comment seems to assume that protection for "software inventions" is desirable, and completely ignores the fact that such "protection" would only result in a minefield for small and medium-sized software companies and independent software authors in Europe.

Also, the notion that "Few would argue that these applications of software innovation do not qualify as a ‘technical contribution’ (with reference to software making improvements to digital TV and MRI scanners) is also, I believe, wrong if you were to ask most software authors. Most programmers would probably say that it's unfair to monopilise software ideas, but fair to protect the expression of ideas with copyright (okay, I know the FSF doesn't even think copyright should apply to software in an ideal world, but I'm pretty sure that's is a minority view, even among programmers). Patents that protect these so called "technical contributions" in the USA only result in barriers against creation of compatible software (think playing DVDs or MP3s legally using Open Source software - not possible in the USA, or even, until recently, displaying web sites with GIF files).

Bottom line: Patentability of software inventions effectively allows a monopoly on programming ideas. If software patents were legal in Europe, not only would it be possible for a European software author to infringe on a patent without even knowing it, such unintentional infringement would be very likely to apply to a computer program of any reasonable complexity, given the large number of frivolous and broad software patents that have already been (arguably illegally) granted by the European Patent Office. The EU should pass legislation that explicitly outlaws patentability of software inventions, otherwise giant software companies will amass colossal patent portfolios, creating an oppressive, stifling environment for small and medium-sized software companies and independent software authors in Europe.

Anti-software-patent lobby should pay attention

Posted Dec 1, 2004 23:00 UTC (Wed) by danielthaler (guest, #24764) [Link] (1 responses)

... okay, I know the FSF doesn't even think copyright should apply to software in an ideal world,...

That sounds wrong. I'd say the FSF feels all code should be free (as in open/accessible/modifiable). Copyright (or rather copyleft) is a means to achieve this. If they felt there should be no copright there would be no GPL.

GPL vs. No copyright protection

Posted Dec 2, 2004 1:11 UTC (Thu) by zblaxell (subscriber, #26385) [Link]

The differences are subtle.

The GPL uses copyright law to make the right to redistribute the software
conditional: the software must be redistributed with the same GPL
license, the software distribution must not be encumbered by any other
agreement that might effectively prevent any of the GPL rights from being
exercised, and the software must include all source code required to
exercise all GPL rights.

Removing all copyright protection from all software would effectively put
all software under terms similar to a BSD license. This would grant
everyone the right to copy, use, modify and redistribute all software,
unless they had a separate agreement (e.g. a contract) which prohibited
such activity.

The GPL makes redistribution rights conditional on certain behavior on the
part of the people doing the redistribution. Removing all copyright
protection would grant the redistribution rights to everyone without
imposing the obligations of the GPL on anyone. This would mean that
people would be free to make binary-only proprietary Linux distributions
and distribute them only to people who sign an agreement to keep the code
absolutely secret...something the FSF probably doesn't want.

Amusing

Posted Dec 1, 2004 22:48 UTC (Wed) by bojan (subscriber, #14302) [Link] (6 responses)

I like the bit when an attorney explains to a programmer what software is all about:

Moreover, the open source community believes that software can be entirely separated from mechanical and electronic inventions.

Huh, next time I download the kernel, I'll make sure I also get the hardware with it ;-)

How about this one:

Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions.

That's news. The man that's been slaving for over 10 years over a large copyrighted work doesn't know what copyright is. I was always under the impression that Linus is a rather smart fellow. Well, looks like I've been wrong ;-)

The problem with some lawyers is that they think the law itself is the beginning and end of everything. Laws exist (or at least should) to protect the people and the society. If they aren't doing that, then they have to be changed. But I guess that's a bit to hard to understand for this particular lawyer.

Amusing

Posted Dec 2, 2004 16:55 UTC (Thu) by sepreece (guest, #19270) [Link] (5 responses)

I believe the comment about separation was going to the point that many things that most people would consider to be inventions include both mechanical and software parts, that in many cases the software is where the technical innovation occurs, and that there's no obvious reason why there should be a line between them. If you support patents for anything, that's a reasonably strong argument.

I have no idea how much Linus knows about IP law, but there's no reason to assume that working on Linux for a long time (or on writing a book, for that matter) would inevitably cause you to learn anything about copyright. Nor does being smart imply that you know anything about any particular topic. I suspect Linus is reasonably familiar with copyright and patent issues, but your argument that we should assume that is baseless.

Amusing

Posted Dec 3, 2004 1:44 UTC (Fri) by bojan (subscriber, #14302) [Link] (4 responses)

> I believe the comment about separation was going to the point that many things that most people would consider to be inventions include both mechanical and software parts, that in many cases the software is where the technical innovation occurs, and that there's no obvious reason why there should be a line between them. If you support patents for anything, that's a reasonably strong argument.

Of course there is an obvious reason. Maybe it isn't obvious to you, but last time I checked, software was text. And text is nothing but written down ideas. It is not different in that regard to a bunch of mathematical expressions, a piece of music or a book of fiction, all of which are not patentable. What computers do is not different from a mental process of a human - it is simply information processing. Whether that occurs inside some kind of mechanical invention or not is entirely irrelevant. It should not be patentable.

> I have no idea how much Linus knows about IP law, but there's no reason to assume that working on Linux for a long time (or on writing a book, for that matter) would inevitably cause you to learn anything about copyright.

You must be joking. Linux is done in a collaborative fashion with every contributor retaining the copyright in their contribution. A quick glance over the COPYING file will show you that Linus *does* in fact understand what he is talking about. He would have to be a complete idiot to not know the issues involved, while accepting *copyrighted* contributions from hundreds of different people. While writing a book may not be a good example of learning how copyright works, *choosing* a licence for Linux and doing a whole heap of merging does. You are probably the only person that believes otherwise.

Amusing

Posted Dec 3, 2004 14:38 UTC (Fri) by sepreece (guest, #19270) [Link] (3 responses)

(1) The fact that much software is described as text is irrelevant. Patents are text as well. Nor does software need to be described as text - there are visual and by-example programming methods, too, for instance. Software is a written description of a way to cause a machine to do something. Software also defines behavior, in a way that generally could be replaced by a mechanical/electronic device. It's not clear why the invention should be patentable if implemented in hardware but not if implemented in software.

Here's a gedanken experiment for you, if you believe "text representation" is an important issue. Suppose we built a hardware realization of a programming environment - little functional boxes, corresponding to C functions, that we could plug together as physical entities and would, when a button is pushed, execute as a program. Would you then say that because it was a physical process it should be patentable, but that an exactly equivalent text description should not be?

I still tend to believe that if you accept patents at all, there's no obvious reason why software should be less patentable than other technical inventions. I do, however, completely agree that the current system is broken, grants many bogus patents, and generally patents things for too long, all of which should be fixed.

(2) While I do agree with you that Linus has a reasonable grip of copyright, your original arguments as expressed ("long time working on the kernel" implies "knows copyright" and "smart" implies "knows copyright") are completely without logical support. I was not saying anything about Linus, I was only commenting on your original argument.

Amusing

Posted Dec 3, 2004 22:27 UTC (Fri) by hppnq (guest, #14462) [Link]

While I respect the mental effort, don't let the intriguing philosophical aspects of what contitutes an invention cloud your perception of the fundamentally flawed patent system. One of those flaws is that it was conceived in an age when there was nothing like software, which is why we are forced to discuss the mechanical aspects of software now. ;-)

Amusing

Posted Dec 4, 2004 5:02 UTC (Sat) by bojan (subscriber, #14302) [Link] (1 responses)

> (1) [...] It's not clear why the invention should be patentable if implemented in hardware but not if implemented in software.

Maybe it isn't clear to you, but it is to me and some others. By allowing broad patenting, we are going on a really slippery slope. Idiotic things like business methods are already being patented as "inventions". What's next? Legal proceedings? Music? Literature?

A boundary has to be set somewhere. Software is already protected by copyright, just like other literary works. We don't need another monopoly protection for it. And we sure don't need anyone to publish their "software invention" in order to get some progress.

And your whole "it's not necessarily text" thing is just rubbish. Text is written down *ideas*. That's the point.

> (2)

Nitpicking, IMHO. This is LWN. People know what I was referring to.

The point is concrete COPYING, not abstract

Posted Dec 13, 2004 5:19 UTC (Mon) by xoddam (guest, #2322) [Link]

*Actually* the best reasons not to allow software patents have nothing
whatsoever to do with abstract notions of whether software can or can't
embody an invention. (Plainly, it can).

The question is not quite being asked correctly. What we wish to
prevent, by preventing 'software patents', is the addition of artificial
restrictions on distribution and incremental improvement of software.

Patents are (supposed to) encourage innovation by protecting inventors
from better-connected or better-capitalised competitors who might outsell
them (even before the inventor has arranged any production at all) and
deny the inventor any reward or recognition.

Computer programs are a different category not because they can't be
original inventions, but because there is no capital barrier to mass
distribution. Like smiles, you can give them away and everyone ends up
the richer. In fact, the only people who can possibly lose out from
copying software are those with a protected monopoly; there is no natural
profit to be made from mere copying.

The granting of a government-imposed monopoly on *any* invention which
*might* be implemented as a computer program prevents that potential
program from being developed or distributed by anyone but the monopolist.
In principle, such a patent need not explicitly describe a software
invention.

Free software is completely vulnerable to patent minefields in an
environment which encourages patents on everything.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 1, 2004 23:33 UTC (Wed) by prahal (subscriber, #18427) [Link]

as all lawyer this is a man who play tricks on words ...

where did this respectble shark got the idea that software was technology ?
Because it is part of a technological good , the computer ?

Algebra is it a technology just by the fact without it plane would crash ?

Those are just the old jesuits disguised ... WTO treaty told taht all technology are patentable thus software have to be patented ...
SOPHISM !

good luck with those mad cows ... did they achieved to patent genome in USA , believe so because animals are used by factory to make cosmetics, hamburgers believe us they are technology ...

I always though technology was the fruit of man s knowledge, now we should relearn , technology is the source of science ... nature , universe, ideas, mathematics are technology .

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 0:31 UTC (Thu) by verzonnen (guest, #9406) [Link]

If a patent can only be granted on "new" inventions, than I can see an unsurmountable problem with patents, at least for those of us who claim that we are religious. Think about it, if this god is all knowing, than surely that would be a strong case for prior art. And any sugestion otherwise would be tantamount to denouncing god.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 5:41 UTC (Thu) by ccchips (subscriber, #3222) [Link]

I believe the idea of copyrights ant patents (at least as framed by the founders of the U.S.,) was to have a trade between short-term monopoly and long-term, unfettered public access.

If you ask me, the real problem here isn't whether something should be patented, but that we have to be so concerned about it. Why?

Could it be because our leaders can't be trusted to make any sort of commitment? That they come up with these social contracts, only to turn around and break them when it suits?

In a world where these things worked like the philosophers intended, I believe we'd probably put up with this. The problem is that the people in support of coftware patents aren't going to stop if they get their way.

Then they will start attacking the "limited time" part of the idea.

Understand?

Executive summary

Posted Dec 2, 2004 5:59 UTC (Thu) by hingo (guest, #14792) [Link]

For hose of you in a hurry, I've summarized the article in plainspoken Egnlish below:

Linus Torvalds doesn't know squat about software, but I do, lawyer says

Disagreeing with Linus Torvalds, Rasmus Lerdorf and Michael Widenius as well as previously stated opinions by Vint Cerf, Tim Berners Lee and the Apache foundation, a lawyer who's an expert on intellectual property effectively says that the Internet, the World Wide Web, and the software quartet most commonly used to drive websites: Linux, Apache, MySQL, PHP, are definitevly not innovations by his standards.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 9:29 UTC (Thu) by petegn (guest, #847) [Link]

The whole thing boils down quite simply to the fact that Patents SUCK and no matter how he tries to explain a way round the issue they SUCK and need killing off the way things are heading these daye you wont be able to write a letter with out infringing some idiots Patent somewhere (most likley sponsered by that well known perveyor or dodgey Patents M$ Corp)


John Collins attacks interoperability clause

Posted Dec 2, 2004 14:32 UTC (Thu) by pdundas (guest, #15203) [Link] (1 responses)

John Collins, a patent attorney who has written against the restriction of software patents before, says:
many innovations in protocols in the telecoms industry, for example wireless-to-wireless communication or wireless-to-landline communication, could be copied without infringing the patent.
This is another way of saying:
many common tasks in telecommunications could be caried out by writing new software from scratch, without infringing the patent
The lawyer seems to think this would be a bad thing.

Paul

John Collins attacks interoperability clause

Posted Dec 2, 2004 17:08 UTC (Thu) by sepreece (guest, #19270) [Link]

Actually, I don't think the lawyer says anything about whether it's a good or bad thing, he just says that the proposed Directive would allow circumvention in some cases where current law would not.

I'd be interested in learning more about exactly what the intereoperability exemption in the Directive allows.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 15:24 UTC (Thu) by mmarq (guest, #2332) [Link] (1 responses)

" Torvalds and his supporters lack a fundamental understanding of intellectual property rights as they seem to be unaware that copyright can only protect software code, and not software inventions..."

How in the hell could be software inventions without software code ?...

This argument seems flawed from tip to toe. Its obviously in contradition with the spirit of protection that copirights is all about. Software is obviously only a language, not a product, which primary form is the code, and the imense different *mathematical* speaches that you can make with it,... and if you protect software code you are obviously protecting also software inventions in a large sense.

There are several human speaked words that mean the same,... there are several mathematical algoritms, that a computacional binary machine will deliver the same result,... lawyers will never get that!.

Otherwise this lawyer seem to argue that, copyright should be void some how, and then 'there is an open door' to steal(copy) everbody's code as long it dosent steal the invention apparatus, whatever that might be !..

I only wonder how much F/OSS must be connect/derived/influenced in every sort of invented apparatus software!?

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 16:39 UTC (Thu) by sepreece (guest, #19270) [Link]

In US IP law, copyrights protect authors against direct copying of their "expressions", which could be words, code, musical notes, etc. Copyrights also protect against someone producing "derived works", such as translations into another language. However, they do not protect against someone expressing the same idea independently. Thus, I can tell you what a copyrighted book tells you, but I can't use the author's words. [This is an oversimplification, since the structure and purpose may also be important in determining whether a copyright is infringed, but it's the general principle.]

A patent, on the other hand, covers an idea and a method of implementing that idea. This is much broader protection, in most cases, than a copyright, because it covers the method, rather than just one expression of the method.

So, to answer your question, there could not be "software inventions without software code", but the code is just a single expression of the invention, so copyright would protect that code (and its derivatives) but would not protect the invention.

Linus Torvalds gets software patents wrong, says attorney (Out-Law)

Posted Dec 2, 2004 16:21 UTC (Thu) by sepreece (guest, #19270) [Link] (1 responses)

I suppose it's predictable that most of the comments posted on this note so far are really opinions on whether or not software patents should be allowed, rather than on what the lawyer actually says.

It would be interesting if somebody who actually knows something about European IP law could comment on the assertions that:
- Software has been patentable in the UK and Europe for 20 years or more
- the proposed Directive restricts patentability and grants de-compilation-for-interoperability rights that previous patent law did not

Current UK position on Software Patents

Posted Dec 4, 2004 0:51 UTC (Sat) by wookey (guest, #5501) [Link]

OK, I've been following this closely for the last year or so, so I'll see what I can do.

Software has been patentable in the UK and Europe for 20 years or more

This is fairly contentious. I could write a long spiel about the UK patent law the European Patent COnvention and the case law that has gone on since the 1980s progressively allowing more and more 'softwarey' things to be patented, but instead I'll point you to a couple of URLs which give a reasonable summary.

UK software patent history

And this page from the UK government sums up the current position reasonably well: http://www.intellectual-property.gov.uk/std/faq/patents/computer_software.htm

So EPO (european patent office) and UKPO (UK Patent Office) practice are similar but not identical. In general the UK is more restrictive about what can be patented due to the limitation to technical fields (i.e. image procesing algorithms are technical, but language processing algorithms are not). Most programmers think this is a crazy distinction.

An example of what is at the edge of patentability in the UK at the moment is this patent from ARM limited - which covers the use of copying a pointer rather than the data it points to, when doing CPU simulations (for switching the register bank). This patent went through the special process of arbitration for ones they aren't sure about 'ex parte', but it did get granted. Again, any C programmer is pretty surprised to find that you can patent this highly unoriginal use of pointers in 2002. The point of course is that this is 'only' in the field of CPU simulations, but still - some of us might quite like to write a CPU simulation using this technique.

There is a very cogent explanation of how this demonstrates the empiness of the 'technical contribution' test here: http://www.softwarepatents.co.uk/current/technical-contribution.html

the proposed Directive restricts patentability and grants de-compilation-for-interoperability rights that previous patent law did not

The proposed directive would restrict patentability if it said 'Computer progams are not patentable', or 'data processing is not a field of technology in the sense of patent law', but it doesn't say either of those things. It appears to in places by saying things like 'computer programs are not patenable unless <something longwindeed which turns out always to be true>'. There are several instances of this sort of language which seem pretty clearly intended to confuse anyone not intimately familiar with the debate into saying 'well that sounds OK, I could support that'.

Asking what 'the proposed directive' says is more complicated than you might think - there have been several versions, and the degree to which they restrict patentability, or offer defences, varies enormously. The original Commission draft basically allowed complete patentability. The amended EU parliament(that's elected representatives) version essentially added clauses like the above offering genuine restrictions and and interoperability defence. The current EU Council (that's the collection of government representatives for each country) version took out nearly all the parliament amendments and added a few even more egregious clauses (such as claims on the actual media containing the prgram as well as the program when run, making copying and quoting infringements as well as running) so this allows complete patentability.

I must admit to not being sure if the interoperability clause has survived in any form, but I have read plenty suggesting that it will probably be a very weak defence in practice, depenmding on what the court deem 'interoperability' to mean. We think it means 'being able to play my DVDs, and read/write file formats', but 'they' think it means 'can write a plug-in for a patented program'. i.e. a piece of software which is not talking directly to the patented program is not 'interoperating'.

Hope that helps. If you haven't already read http://www.nosoftwarepatents.com/ , then do so, as that is the clearest site on the subject by a long chalk. It doesn't go into the UK-specific questions too much, but it is the european level that really counts. It is safe to say the the UK is at the 'most pro-software patents' end of the discussion at the moment, and few cracks in that government position have opened up so far, although there are signs that they might be beginning to.


Copyright © 2004, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds