en.swpat.org is a wiki.   You can edit it.   May contain statements End Software Patents does not endorse.

November 2014: About Microsoft’s patent licence for .NET core

SitemapCountriesWhy abolish?Law proposalsStudiesCase lawPatent office case lawLawsuits

Talk:Software does not make a computer a new machine

From en.swpat.org
Jump to: navigation, search


[edit] Comparison with automobile patents

I don't think that this argument is comparable to automobile patents. The act of driving to a certain street does not involve creating anything new or useful. You are just using already acquired skills (driving, navigating) and goods (automobile, fuel) to achieve a certain goal (reach your targeted street). As opposed, developing a piece of software takes a lot of resources and time to create something distinctly new and potentially useful to a select group of individuals.
That this part of the argument seems to be poorly thought out and not enough to make the entire idea convincing. Creating software more similar to how the music industry uses patented equipment to create content (songs) which serves a specific purpose (entertainment). Music is uniformly regarded non-patentable, and would perhaps be a better argument. Share your thoughts on this. Unauthorized 06:15, 5 October 2009 (EDT)

The analogy should be improved alright. Music, and literature are indeed good arguments. They're discussed on the analogies page. If you think you can help us improve this article or any other, please do. Ciaran 07:44, 5 October 2009 (EDT)

[edit] Weak arguments

I don't think the argument works. So if a hammer and chisel is patented, anything you make from the hammer and chisel can't be patented? The wright brothers used patented saws, screws, wire, and fabric to make an airplane.. Does that mean they shouldn't have received a patent for it?

All inventions combine other, less specific technologies to create a new, specific thing that does something valuable.

The problem people have with software patents is only that they can't touch it. I also think software engineers don't like software patents because they are already familiar with the technology, and can envision themselves coming up with it... So to them, its not that impressive. I'd imagine the germans working on airplanes in 1901 didn't think wright should receive a patent either "Oh, he just combined what I found and what he found" etc

This page certainly needs improvement (feel free to jump in). One difference between software patents and the Wright brothers is that the latter made something that was physically different from what existed before. With software and a computer, the end result is just a computer. As one EPO patent examiner said when rejecting a particular software patent "The computer could already do that; you just programmed it to do so". Ciaran 19:48, 18 September 2010 (EDT)

[edit] Not sure if the house analogy works

I haven't read the whole text, but it seems to be at least ambiguous about whether it's an analogy for hardware or software.

Computers (hardware) are patentable and we've no problem with that.

Ways of ventilating buildings can also be patentable, and we've also no problem with that.

If someone wants to patent a computer-controlled way of curing rubber (Diehr), or operating a washing machine,[1], or Anti-lock braking - that's all fine. Those aren't software innovations, they're curing, washing, and braking innovations.

What we're against is computer-controlled innovations in areas such as storing audio (mp3), video, and interactive displays of information (webpages).

Calculators and piano-players are good examples because they can't be used to do anything patentable. A house with automatic doors is different because there are patentable applications. It could be used as a mouse trap, and mouse traps are patentable. (A silly example, but there are so many, I just picked one.) Ciaran 11:33, 23 November 2010 (EST)

The house is patentable. It is patented once when created. Here ends the patentability -- as concerns the making on machines.
When you add software to "create a new machine", THEN, in contrast to when you built the computing house, you are not creating a new machine.
To explain to someone that doesn't know what software is that in fact you are not adding anything, I state very roughly how a computer works (by focusing on a computing house that uses familiar doors) and in so doing recognize that software is nothing but the opening and closing of doors on this house. If you open and close the doors a particular way, you enable the house to effect the algorithm encoded into those doors (as opened or closed configuration of doors).
This analogy tries to get to the essence of software being information (a configuration of open/close) and not an actual new machine. Any suggestions or rewritings that will improve this are welcomed. Jose X 17:13, 23 November 2010 (EST)
I've now read the first paragraph and a half, and I think there's indeed the above problem. Quite simply, changing the configuration of doors in a house can create a new machine. It can create a mouse trap. It can create a fan. It can create a security tunnel through which people can only flow in one direction.
No. If you change the door structure or capabilities, then perhaps you would be creating a new mousetrap. I stated that we are dealing only with changing the configuration on an existing system into its accepted states. You might want to call this a process patent but it is not the creation of a new machine. If I take scissors and open them to a particular angle to help hold a particular window open or to help cut down a mouse in size, I am using the same "machine" for a new purpose but have not created a new machine unless I change the hardware in some way beyond what it already supports as its intended usage. Jose X 02:55, 24 November 2010 (EST)
The "new machine" problem comes from patent applications which say that by configuring a computer to play MP3s, it's no longer a computer, it's an MP3 playing machine. There's never any talk of changing the structure or capabilities of the computer. This trick would not work with a calculator or a piano player - no judge would accept that either has been turned into a new machine. But this trick would work with your house. It's a house, but if I configure it to shut the doors when a mouse enters, it is no longer a house, it has become a mouse trap and mouse traps are patentable subject matter. The house turns into a construction material of the invention (mouse trap).
No, you don't shut the door on a mouse. The doors only hold information. It's just information processing. The piano playing of the information, perhaps after having been calculated with the calculator, corresponds to a display screen showing the information of pixel values calculated with the computer. Jose X 08:55, 24 November 2010 (EST)
You have to compare the materials before and after. Before, there's a house. After, there's a mouse trap. When you remove the non-patentable stuff (the house), you have to look at whether the stuff that's left (mouse trap) is patentable subject matter. With a calculator, before and after having a sum keyed into it, finished thing minus the pre-existing (non-innovative) thing, leaves a sum. Not patentable. Same for the piano player with new sheet music. We say the same goes for software. But this does not work for your house. Many things can be done with the house to produce a patentable innovation. (The page In re Alappat (1994, USA) is a good read for this topic) Ciaran 08:33, 24 November 2010 (EST)
The sum left by the calculator is the new wav file data left by the computer simulating mp3 calculations.
The house components can be changed in capabilities but that is changing the hardware, not changing the software.
Calling a computer an mp3 player is calling a paper with certain writings on it a contract or a will or a mortgage or an instruction manual. The information on the paper does nothing physically functional. A person can then do something with that information. A robot can then do something with that information. A sophisticated cooking machine can then do something with that information. A computer application can then do something with that information. The human, robot, cooking machine, or computer application do not become a new physical object, they simply take on a different label we create as humans to describe its new use. This page is about new machines not about process patents. Jose X 09:02, 24 November 2010 (EST)
A computer is a machine able to perform mathematics (a glorified calculator or very fast human) on data. Some of the data (instructions) defines what mathematics to perform on the other data. It's nothing but information processing. It's nothing but seeing which doors are open or closed (instructions) and using that to decide how to modify other doors (data).
The other examples are good because they liken computers to other things whose use is never patentable. Use of a calculator, use of a piano player, is never patentable. Likening computers to these things supports our policy. You've likened computers to something that can be used for patentable things.
On a stylistic note, the other examples also have the benefit of being simple, which is good because we're trying to simplify this topic for politicians and judges. Maybe if your text was much shorter and simpler, the problems (or lack there of if I'm just misunderstanding your text) would be easier to spot so the text can be fixed (or removed if it's fundamentally flawed). Ciaran 17:48, 23 November 2010 (EST)
It may be beneficial to separate out that analogy into an article on what software is. Afterwards maybe create a simple analogy making a reference to the longer explanation.
OK, I don't consider your argument about legitimate. Before changing this page, can you give another run at it and see if you still think there is a logical flaw that would warrant scratching it or redoing it significantly? Jose X 02:55, 24 November 2010 (EST)
Maybe putting it on a new page would be a good idea. Template:Brainstorming could be used to encourage people to analyse it and provide comments. The scope of the new page is up to you. Ciaran 08:33, 24 November 2010 (EST)
We're trying to make things simple! Who's going to read all that? The more text you use, the higher the chances that the sceptical reader will find a problem with your analogy. Ciaran 08:48, 24 November 2010 (EST)
Look at the paragraphs at the very beginning.. before the "in more detail". The rest will have to be cleaned up and maybe posted elsewhere. lol Jose X 08:52, 24 November 2010 (EST)

Just updated the house analogy to mention machine vs process. Let me know what you think and remember that the top can be adjusted and then everything from "in more detail" onward removed. Jose X 09:13, 24 November 2010 (EST) Update: I removed the long stuff and now have a house analogy plus an mp3 simulator analogy. Jose X 09:41, 24 November 2010 (EST)

I can't review the text, I'm working on something else. I generally don't read any of your texts completely, they're too long. I read the first paragraph, find problems, and I tell you about them.
I read the first few words of your new "mp3" section and you're talking about your house again. That's the opposite of simplicity. Each example should stand on its own. I should not have to read the house section to understand anything about the mp3 section. The mp3 section should be a separate analogy so that there's an alternative for readers who find the house analogy to be poor.
Removed that reference. I considered removing it earlier but thought that someone that read the house example right above it might gain a little from mentioning house. Jose X 12:40, 24 November 2010 (EST)
I haven't read much of your new medical calculations text but already I can imagine that there will be problems because medical systems (including, surprisingly, dosages) are patentable. If we try to argue that some medical idea is not patentable and that software is similar and therefor software should not be patentable, not only might we be committing an error in our comments about the patentability of the medical procedure, but we're also making life needlessly difficult by invoking opposition from the medical-patent community (who are very active, they submitted briefs for Bilski).
This example does not follow that pattern as it stands so it doesn't belong on that page I suppose. I used the opportunity to contrast software on computer (unpatentable) with hardware system (patentable) when they each compute the same algorithms.
I'll rethink the example or see if it can be used elsewhere. Jose X 12:40, 24 November 2010 (EST)
I read over the top part of the page, and I see a problem. The intro to the analogies states that all the analogies aim to show that using something a particular way is not patentable and this matches the software use scenario. I think patent lawyers, at least in the US, are trying to patent processes/uses. So some of the arguments I have been using are focused on the title of this page, to show that software does not make something a machine, rather than focusing on confidently stating that a "use" of something (like a car or calculator) is definitely not patentable, since some lawyers want it to be and the patent office is handing out some process patents. Jose X 12:53, 24 November 2010 (EST)
Update: I reorganized a bit the introduction to the analogies to enable some of the discussion (the longer analogies) to be more flexible. I also call then longer comparisons rather than longer analogies.
The object of the medical diagnostics and the mp3 sections as they stand now are to show that these pieces of hardware are not the same thing as software running on a computer. I mention how calling something by what it simulates does not make it be that thing physically.
The object of the house section is to give insight into just how intangible software is by looking at a house whose parts we can feel and touch vs trying to imagine electronics. It shows the nature of software and that this nature clearly is not new matter.
I'm worried that when I do finally get around to reading the texts you've added to various pages, I'll have to remove most of them. Ciaran 11:16, 24 November 2010 (EST)
I did a bit of minor rewording and added a brief explanation of the goals of the longer comparisons. If you get further reading them (success) and have more criticisms, shoot. The moving around of a few paragraphs makes the diff kind of large, but you can see the changes here http://en.swpat.org/wiki?title=Software_does_not_make_a_computer_a_new_machine&action=historysubmit&diff=21413&oldid=21394 Jose X 13:39, 24 November 2010 (EST)
BTW a primary theme in the long comparisons is that software does not make a computer a new machine or turn it into the simulated machine. I start off the comparisons by stating something along those lines. A second theme, which generally is made a bit explicit near the conclusion of each analogy, is that software is just information and a computer following the instructions does not create any new tangible object. A computer can process information but it cannot manufacture a physical object. We simulate results, but we don't create new machines. Patents are for when we want to create sufficient incentives to manufacture and disseminate machines. Copyright already exists to handle incentives for producing purely intellectual content. Jose X 15:14, 24 November 2010 (EST)

Ok, I read another paragraph, and again there're multiple problems.

An mp3 player can be made as a new device entirely. This is different than loading software unto a computer.

They are usually not different. An mp3 player is usually a small computer with software on it.

I just verified by rereading your earlier comment above that I misunderstood your position on mp3 players. I threw in that mp3 example under the assumption that (like electronic medical diagnostic tools) mp3 players had other other features that you considered might be patentable. Jose X 21:00, 24 November 2010 (EST)
The mp3 device would use cheaper parts than a full blown computer (plus speaker system).

So what? And how does that help our argument? Our cost-related argument is that software+computer is so cheap it's commodity. What good does it do to argue that something else is even cheaper??

Keeping in mind that I misunderstood what you had in mind by an mp3 player, what this shows is that you can compete on price and quality if you manufacture using a particular and optimized process focused on mp3 players. Meaning that a patent can help you start such a costly manufacturing process and then you would compete in the store front against other such players but not directly against computers simulating as mp3 players as the latter would be priced much higher (and hence out of the market) for someone looking for a typical mp3 player. The patents have to have a reason for existing, and this is potentially a fair reason: it helps you raise capital to manufacture a device without worrying that those that get a pass on the patent infringement will be able to compete directly with you (they won't).
If you read this, I hope you agree it makes sense, even if the mp3 comparison is not one you wanted to make.
It would not be subject to being "erased".

First, so what? Second, there's nothing stopping the mp3 player from having it's software erased.

The computer, meanwhile, is still the same computer as before.

As before what? Before the software was put on? And the mp3 player (small computer) is the same as before it got software put onto it.

The idea I went with here was that the mp3 player might be implemented with something other than a general purpose core. You have taken the position to ignore cases that are not clearly software related, and I thought you were including mp3 players in this category of things that might be patentable. Jose X 21:00, 24 November 2010 (EST)
It is not a new computer or a new mp3 player. The computer does what general purpose computers do. They carry out algorithms, even if those algorithms simulate the logic manipulation of an mp3 player.

You're proving nothing. You're just stating obvious contextual facts and repeating yourself.

The paragraph thereafter then talks about something being "just a new use of an old machine" - but new uses of old machines are patentable! (If the new use is an innovation in patentable subject matter) That's basic patent law. If someone were to use that text in a letter to a politcian or a judge it would discredit their whole argument.

OK, I realized what happened. I misunderstood your position on process patents. I thought you were ignoring process patents or viewed them as not legitimate; hence, it made sense to me to use that as a conclusion.
In the US, a problem we have is that patent applicants are arguing that if something is a process, then it should get a patent.
Reading your comments here (and the introduction to the brief analogies) again more carefully, I now see that you see some processes as clearly not patentable even if someone attaches them to a machine but perhaps see most processes as patentable. It really wasn't clear to me that this was your position.
Note: 'Our goal is to show how computers are the same as other things whose use cannot be patented. When you use a record player, you get music. The music might be technical, innovative, new, etc. but no one will ever get a patent on use of a record player' is part of the reason I was confused and thought you were considering all uses to be unpatentable. Maybe in the EU a process is not patentable but clearly in the US I get the impression many lawyers and the USPTO would argue that a use on a record player that is novel is as patentable as anything else. Jose X 21:00, 24 November 2010 (EST)

I've tried at length to explain what we're trying to build here at swpat.org, but you seem to have a very different vision and our discussions are not leading to convergence.

I do get a bit confused sometimes because I don't know exactly what you think on different issues. I believe there is ample room for me to be in the dark on EU patent law, so I am willing to accept certain interpretations of what I read here.Jose X 21:00, 24 November 2010 (EST)
I keep saying that we need concise texts wherein every sentence is meaningful and correct. You keep adding long texts with errors in patent law and logic errors (or maybe it's just unclear expression of a correct logic but the result is texts which appear illogical or pointless). I don't see any EU/USA misunderstaning. New applications of known ideas are patentable in both regions. Ciaran 07:16, 25 November 2010 (EST)
This is tricky material to word properly. For example, many will find it obvious that a software program can be an application of a known idea.
Ah, yeh. It's that some new applications of known ideas are patentable. You look at the patent application, subtract everything that's already known (state of the art), and then decide if the remaining part (the innovation) is patentable subject matter. A new way of stopping a car is. New music isn't. Ciaran 11:04, 25 November 2010 (EST)
[Update: I'll leave the following (longish) comment for reference, but it can safely be ignored (it was intended as a criticism of you criticism). .. Keep up the good work, and I'll try to stay out of the way.]
How about software?
My point is that you can't just say blah blah blah (limited text) and expect other people to know what Ciaran means (in detail and accuracy) because many lawyers and others disagree over many details. Aren't we having a battle over the patentability of sofware? Did not a lot of people support the Bilski business method patent and aren't many still likely supporting many business method patents? Is not the USPTO supporting software patents in many ways? Aren't there people claiming that new software in not part of the prior art? Aren't many or all of these people also claiming that "some new applications of known ideas are patentable?" ...Why am I repeating so much? Because it seems you answered the question I asked without answering it and simply falling back to a small number of tired expressions that mean something different to everyone (eg, see your last paragraph above). I agree with much of what I read on this website, and similarly I too recognize that there is a communication problem and potential for holes and missing explanations to be exploited by software patent supporters (even if I do not write as clearly as you do most of the time).
OK, so this whole website is one way for Ciaran to bring into focus a definitive answer to the question of whether or not software is/should be patentable. Yes, I will try not to make anything that might be a significant change to the live website without it having been hammered out already (and probably avoid most things too significant anyway, at least not without asking you), and I will try hard to remember to be mindful to keep diffs honest and limited. I'll probably mostly focus on an independent essay here or there that might serve as ideas or be referenced from here. You can hold me to these words if/when I mess up again, but keep in mind I am not inside your head. Jose X 02:53, 28 November 2010 (EST)

I'm sorry, I think this collaboration is failing. Ciaran 18:10, 24 November 2010 (EST)

[a bit angry...] All of this because of the mp3 misunderstanding? Should I assume that the other 2 examples are also problematic? Should I assume you don't have time to go over any/some edits I make? Should I assume the other things I have written on other pages are also incompatible? Should I stay away for a while? What are you suggesting? Jose X 21:00, 24 November 2010 (EST)
It's not all because of MP3s, it's a pattern. You and I obviously have very different ideas of what should be on a documentation site for anti-swpat campaigns. en.swpat.org happens to be a site I bare responsibility for. It's a public wiki, so a certain amount of diversity is welcome. I don't expect everyone to share my vision or style, but I do have to set some limits on what's acceptable (in terms of quality of writing, quality of information, and correctness) in order to meet the goal I set for swpat.org. Everyone's welcome as long as they stay within those limits. I'm not asking you to stay away, but I'm saying that you have to stick to swpat.org's quality standards, and they're based on very concise writing, clear writing, and factual accuracy. This will require a radical change in your writing style. Ciaran 07:16, 25 November 2010 (EST)
Put another way, I should never have to do any more than glance at a diff to see if a change can stay or not. Reviewing and commenting on long texts is currently taking up a lot of my time and the end results are poor. I should be using my time to make my own improvements to the wiki. The wiki needs proofs. If you could add references and links to confirm existing sentences, that would be great. If you want to add something new, try to condense it to a single sentence. That'll seem very strange or challenging to someone with your style, but it's what this site needs. This might mean formulating ideas in a drafting area such as a subpage of your user page. Add the idea to the wiki mainspace when you've crystallised the idea to a single sentence. Ciaran 09:48, 25 November 2010 (EST)
(I've just tried to edit the Why software is different page and found it's a real mess. There's maybe half a day of work for me there to fix - time I don't have right now. Ciaran 10:23, 25 November 2010 (EST))
(REPLACING PREVIOUS NOTE: The poor state of that article isn't due to your edits, it's due to multiple people, but it's still a good example of an unreadable article that I'll have to work on. Ciaran 20:08, 25 November 2010 (EST))
[a bit more calm...] Let me try to be a little useful in cleaning up.
http://en.swpat.org/wiki?title=Software_does_not_make_a_computer_a_new_machine&action=historysubmit&diff=21404&oldid=21403 is the diff showing when I adjusted the section titles a bit.
The medical diagnostic example does not follow your pattern for this page, but it might be useful elsewhere. A point it tries to capture is that progress in fields that don't rely on patents and fields which are valuable to medicine is curtailed if we add patents to the software made by these experts when they use it on a computer (even if its the same algorithms that might be used within a medical diag device).
I know that having others move things around can be upsetting when you are organizing and are trying to keep lots of details in mind. I didn't anticipate this frustration from trying to improve a house analogy argument "live" or I would have slowed down a bit, left it alone for now, or moved it into a more private section from the get-go.
I still want to reread the house example again slowly to see if it makes sense for this page.
BTW, check out the title for this page. I think I was writing arguments that were in line with trying to show software does not make a computer a new device, even if the arguments were not following the pattern you set out in the intro to the analogies section, which happens to have been a pattern established very recently. I'm just explaining my position. Jose X 21:54, 24 November 2010 (EST)

[edit] Removing some analogies

I'm going to move the weaker analogies from the article to here. The current set seems to invite an infinite list of "X plus Y doesn't make a new X" sentences, but I'm not sure that such a list is useful.

Let's at least keep the ones that have been used by judges and in amicus briefs.

I've moved the "Human" and the "finger-fork" analogies to here. I think they're flawed.

For the medical calculations analogy, I think it might be usable, but it has to be shortened.

[edit] Human

A human performing a set of steps does not become a different human when he or she changes to perform a new set of steps. The human is simply following a new configuration in his or her head.

Similarly, a computer system running different software is still the same (already patented) computer system.

[edit] Fingers v. fork

A human using his/her fingers to scoop up food does not become a fork. Similar effects can be achieved through completely different "machines", eg, where one simulates the other.

[edit] Medical Diagnostics Calculations

When 1000 mathematicians and physicians sit down at their computers to collaborate over the Internet to come up with formulas and algorithms that use values given to it beforehand in order to derive a medical diagnosis, they are not creating a new machine, even if the software they create can run on a computer and simulate the information processing taking place inside a proprietary medical diagnostics hospital machine.

The medical diagnostic machine requires that certain hardware exist to take certain measurements (eg, to analyze blood samples or measure temperature), and these all come in a robust package optimized for hospital use. Manufacturing this machine is different than having a world of professionals collaborate to produce the top quality algorithms (as software) that produce the best diagnoses and then installing these algorithms on a computer.

Algorithms to process medical information are within the purview of the mind only. Those creating the algorithms (software) are not creating (building) hardware. Their intellectual creation process should never be hand-cuffed with a patent, but that is exactly what is being done if patents are to be used to prevent them from testing out and optimizing those algorithms on a computer or to prevent them from using the algorithms to provide diagnoses for others.

Copyright is what was designed to cover products of the intellect like software. These needs are different than the needs patents address (to enable the creation of new hardware).