- Site menu:
- Latest Discussions
- Latest Newswire
- Archives
- Submit Story
- Story HOWTO
- Syndicate
- E-mail Alerts
Google open video codec faces second challenger
|
|
|
Comments (116)
96 earlier comments in the full thread.
> Real problem may be the kind and number of software patents being granted.
Actually, I would agree with that. Sharply limited software patents, of limited duration, for truly innovative creations, would probably serve the public good. But that's not what our current system gives us. And you're the one arguing that the current patent system is applicable. So tell me how to fix it in a way that will get implemented within our current system. Then maybe I'll listen. Arguing for an obviously broken system isn't going to convince me, regardless of how internally consistent the argument may be. Especially since the argument of software being too close to the underlying mathematics to be patentable is also internally consistent.
Actually, I would agree with that. Sharply limited software patents, of limited duration, for truly innovative creations, would probably serve the public good. But that's not what our current system gives us. And you're the one arguing that the current patent system is applicable. So tell me how to fix it in a way that will get implemented within our current system. Then maybe I'll listen. Arguing for an obviously broken system isn't going to convince me, regardless of how internally consistent the argument may be. Especially since the argument of software being too close to the underlying mathematics to be patentable is also internally consistent.
@jdixon --
Not as bad as you think, but there is (or, at least, used to be -- might be better now) a major chink. Patent attorneys -- who actually do a bunch of the heavy lifting -- are expected to have an engineering background. When I was in law school, I considered patent law, buy my software background was not sufficiently "engineery" for entry to the patent bar. I'd have had to study up (in addition to patent law) and take an exam certifying me as an engineer.
My guess is that there are plenty of technically proficient people int the patent process, but in the wrong technologies. How else can you explain their inability to understand the incredible creativity of software people?
In my vision of the world, if you have patents, you **can** have software patents, but probably **don't** have very many for the simple reason that the ordinary practitioner of the art (which is the standard of obviousness) is such a creative little devil.
Not as bad as you think, but there is (or, at least, used to be -- might be better now) a major chink. Patent attorneys -- who actually do a bunch of the heavy lifting -- are expected to have an engineering background. When I was in law school, I considered patent law, buy my software background was not sufficiently "engineery" for entry to the patent bar. I'd have had to study up (in addition to patent law) and take an exam certifying me as an engineer.
My guess is that there are plenty of technically proficient people int the patent process, but in the wrong technologies. How else can you explain their inability to understand the incredible creativity of software people?
In my vision of the world, if you have patents, you **can** have software patents, but probably **don't** have very many for the simple reason that the ordinary practitioner of the art (which is the standard of obviousness) is such a creative little devil.
Quoting:Most of the non-trivial software systems (and some of the trivial ones too) cannot be modeled as pure mathematicsOnce the input to the software is taken into account (Turing's oracle), the pure mathematics are all that remain. Or has the Church-Turing Thesis been disproven?
Quoting: The machine may embody the invention, but it isn't the same thing. Patents would make no sense if the invention were the machine.
You could patent the machine, but what would that get you? Somebody could make another machine exactly like that one because is it is the machine that is patented, not the creative contribution embodied in the machine.
I follow your point in the first paragraph above. You are saying that the patent is on the idea (the eureka moment) which is embodied in the machine. But, the second is a non-sequitur. If you have the protection of a patent on a thing (or process), then anyone copying it (without permission) is violating your rights. I'd argue that without some tangible result (or the potential for some tangible result of a process), you do not have something that can be patented.
In the case of (general use) computer software, it's already all been done. There is already; machine produces image, machine delivers sound, machine operates some other machine, machine manipulates data, machine takes X form of input, machine gives status of itself, machine delivers status of X event, etc, etc, ad nauseum. Unless a piece of software delivers some new, spectacular thing that nobody has conceived of (& what THAT could be escapes me), most of what is the possible function of computer software has already been done (& done to death).
The major problems are trivial, non-unique, already been done (prior art) patent filings. With the burden of proof (to a clear & convincing standard) resting on those trying to invalidate the patent, the system can be gamed with what should be invalid patents. A prime example is M$'s FAT32 patent. It was invalidated in one challenge, only to be reinstated in a closed re-hearing. Now it's M$'s front line offensive weapon against open source. But, it had been done before. M$ just put a little tweak to the thing, & "presto," an IP weapon was born.
I know you want IP protection for programmers' hard work. But it should be done by extending copyright protection.
Quoting:Unless a piece of software delivers some new, spectacular thing that nobody has conceived of (& what THAT could be escapes me)Doing it in XML, over the Internet. /sarcasm
JaseP -
On the first, you are close but not quite there. You can't patent an idea, and so the idea is not the invention. The invention is reducing the idea to something useful. It may seem like a picky point, but lots of people had the idea of an electromagnetic telegraph. It took Samuel Morse to reduce the idea to something useful.
As to the second, yes -- you need a tangible result. Couldn't argue for a moment, but there is no reason why that tangible result has to be steel or silicon or what-have-you. The very fact that you are able to type your response indicates that software is, in fact, a tangible result. The browser you are using works in prescribed ways to give the result you expect. It is not some thought experiment or some ethereal "something" for which we can find no evidence in the real world. Fire up your computer, start up the app, and there it is.
On the first, you are close but not quite there. You can't patent an idea, and so the idea is not the invention. The invention is reducing the idea to something useful. It may seem like a picky point, but lots of people had the idea of an electromagnetic telegraph. It took Samuel Morse to reduce the idea to something useful.
As to the second, yes -- you need a tangible result. Couldn't argue for a moment, but there is no reason why that tangible result has to be steel or silicon or what-have-you. The very fact that you are able to type your response indicates that software is, in fact, a tangible result. The browser you are using works in prescribed ways to give the result you expect. It is not some thought experiment or some ethereal "something" for which we can find no evidence in the real world. Fire up your computer, start up the app, and there it is.
@dino:
The letters on the screen are not tangible. The pixels are; I can touch them (once I remove the LCD matrix from the display frame), but I cannot touch their candela output states which combine to make the letter "A". One can no more touch a light bulb's wattage.
Or would you say that the bit combination "01000001" is also tangible, anywhere it occurs, because it represents the letter "A" in ASCII and UTF-8?
The letters on the screen are not tangible. The pixels are; I can touch them (once I remove the LCD matrix from the display frame), but I cannot touch their candela output states which combine to make the letter "A". One can no more touch a light bulb's wattage.
Or would you say that the bit combination "01000001" is also tangible, anywhere it occurs, because it represents the letter "A" in ASCII and UTF-8?
@gus3 -
I would say that your conception of the world is so foreign to me that I cannot conceive of how to continue this conversation.
Cheers.
I would say that your conception of the world is so foreign to me that I cannot conceive of how to continue this conversation.
Cheers.
> In my vision of the world, if you have patents, you **can** have software patents, but probably **don't** have very many for the simple reason that the ordinary practitioner of the art (which is the standard of obviousness) is such a creative little devil.
Even though I disagree with your reasoning, I could live with that. So, how do we get there from here?
Even though I disagree with your reasoning, I could live with that. So, how do we get there from here?
> I would say that your conception of the world is so foreign to me that I cannot conceive of how to continue this conversation.
I would say that's largely been the problem with this entire conversation. We see things so differently that we're talking past each other.
I would say that's largely been the problem with this entire conversation. We see things so differently that we're talking past each other.
@gus3
> Once the input to the software is taken into account (Turing's oracle), the pure mathematics are all that remain. Or has the Church-Turing Thesis been disproven?
Turing machine, lambda calculus are only for defining computability; oracle machine is a thought machine to study decision problems (halting problems etc.) and has got to do nothing with inputs. As soon as I/O facilities of modern computing systems come into play, it cannot be modeled using these precisely because they are designed for dealing with computability. Heck some even hypothesize that the whole universe is a Turing machine, but again that only applies to the computability and not the other aspects of universe.
@jdixon
> And you're the one arguing that the current patent system is applicable. So tell me how to fix it in a way that will get implemented within our current system.
I only said that if patents are applicable to other fields then they are equally to software. And I also said that there obviously look to be problems on how patents (particularly for software) are currently been given. Can't say I know much about how the patent system works internally so suggestions from my side for fixing make little sense. For that matter your position of doing away with software patents is more unlikely to "get implemented within our current system" than any other possible suggestion to fix things.
> Arguing for an obviously broken system isn't going to convince me
straw man
> Once the input to the software is taken into account (Turing's oracle), the pure mathematics are all that remain. Or has the Church-Turing Thesis been disproven?
Turing machine, lambda calculus are only for defining computability; oracle machine is a thought machine to study decision problems (halting problems etc.) and has got to do nothing with inputs. As soon as I/O facilities of modern computing systems come into play, it cannot be modeled using these precisely because they are designed for dealing with computability. Heck some even hypothesize that the whole universe is a Turing machine, but again that only applies to the computability and not the other aspects of universe.
@jdixon
> And you're the one arguing that the current patent system is applicable. So tell me how to fix it in a way that will get implemented within our current system.
I only said that if patents are applicable to other fields then they are equally to software. And I also said that there obviously look to be problems on how patents (particularly for software) are currently been given. Can't say I know much about how the patent system works internally so suggestions from my side for fixing make little sense. For that matter your position of doing away with software patents is more unlikely to "get implemented within our current system" than any other possible suggestion to fix things.
> Arguing for an obviously broken system isn't going to convince me
straw man
krisum:
Tried to explain the 'problem' with pointers: A pointer is not the same thing as what's at the memory address it points to. Or - a street + address is not the same as 'the house at that address'. In the same way "Ceci n'est pas une pipe (R. Margritte)".
So the h.264 scheme is math/information and points to / refers math/information. A civil engineering drawing (if that is what it's called?) is math/information and points to matter.
Information and matter should not be dealt with in the same way. Because information can be (and lots of times is) 'intangible', even while the 'embodyment' of the information may be tangible. But information is independent of the medium. Information can be copied with almost no additional cost. Information simply isn't and cannot be scarce, as Bob R / Daniel O tought me, and I agree. You can't 'protect' information in the way you can protect matter. Everybody on the world can use the _same_ info without stealing.
Matter - on the other hand - is dependent on the medium. It is scarce. It is hard to copy (though of course I follow the great developments of the RepRap Mendel). It's _not_ combinations of atoms which are patented in medicines, as you probably know, because then it's too easy to sidestep by means of adding or changing some unimportant 'groups' of atoms. It's the process of creation, how to deal with the matter which is patented in 'chemical sciences' as far as I learned.
Even though I appreciate the idea of treating matter and information in the same way, it's so full of problems I'd say it's not workable - except for bullying small companies and poor programmers. Current software patents clearly reflect these problems.
But if I/O of computing system cannot be modeled by "Turing", then the part which cannot be modeled by Turing clearly isn't 'information only', probably deals with matter and would be "open to patent". Such a solution is workable, even though I doubt if patents "all together" are a good thing from an economic point of view. Applying the Turing test though (only patent the part which cannot be modeled by Turing) is workable though.
Of course I think transforming / optimizing information should be rewarded (Doh! It's my job), but I think patents are not an efficient way to do so, because they suffer from too many adverse side effects.
Tried to explain the 'problem' with pointers: A pointer is not the same thing as what's at the memory address it points to. Or - a street + address is not the same as 'the house at that address'. In the same way "Ceci n'est pas une pipe (R. Margritte)".
So the h.264 scheme is math/information and points to / refers math/information. A civil engineering drawing (if that is what it's called?) is math/information and points to matter.
Information and matter should not be dealt with in the same way. Because information can be (and lots of times is) 'intangible', even while the 'embodyment' of the information may be tangible. But information is independent of the medium. Information can be copied with almost no additional cost. Information simply isn't and cannot be scarce, as Bob R / Daniel O tought me, and I agree. You can't 'protect' information in the way you can protect matter. Everybody on the world can use the _same_ info without stealing.
Matter - on the other hand - is dependent on the medium. It is scarce. It is hard to copy (though of course I follow the great developments of the RepRap Mendel). It's _not_ combinations of atoms which are patented in medicines, as you probably know, because then it's too easy to sidestep by means of adding or changing some unimportant 'groups' of atoms. It's the process of creation, how to deal with the matter which is patented in 'chemical sciences' as far as I learned.
Even though I appreciate the idea of treating matter and information in the same way, it's so full of problems I'd say it's not workable - except for bullying small companies and poor programmers. Current software patents clearly reflect these problems.
But if I/O of computing system cannot be modeled by "Turing", then the part which cannot be modeled by Turing clearly isn't 'information only', probably deals with matter and would be "open to patent". Such a solution is workable, even though I doubt if patents "all together" are a good thing from an economic point of view. Applying the Turing test though (only patent the part which cannot be modeled by Turing) is workable though.
Of course I think transforming / optimizing information should be rewarded (Doh! It's my job), but I think patents are not an efficient way to do so, because they suffer from too many adverse side effects.
krisum, I was talking to Dino, not you.
And a simple statement of fact concerning my position can never be a straw man, since it has nothing to do with the other persons argument as such.
And a simple statement of fact concerning my position can never be a straw man, since it has nothing to do with the other persons argument as such.
I researched the patent system a little. While I haven't committed the process to memory, I have a basic understanding of the current problem...
Originally, abstract math & things found in nature could not be patented. That excluded algorithms, and things like DNA. Programs were originally considered not eligible for patent protection, as an algorithm. However, several Supreme Court decisions eroded that original standard. What was left provided little guidance to the USPTO. On top of that, USPTO reviewers, while generally having engineering backgrounds, do not necessarily have backgrounds in the fields they are assigned cases for. To compound the problem, the reviewers have deadlines to keep & the USPTO looses out on (renewal) fees if patents are not granted. Hence, as you can imagine, reviewers are "encouraged" to approve patents by their management. You can imagine the effect on granting questionable or borderline software patents.
How can that failure of the system be fixed? The fix could take 3 forms; (1) Congress could remove software patents from eligibility [VERY unlikely, given lobby $$$], (2) The USPTO can be forced [Congress? SCOTUS? Again, unlikely] to reform procedures, or (3) SCOTUS can provide guidance on standards, or weaken the standard of proof in litigation [Given that this is what M$ is asking for in the I4I appeal, this is most likely].
A weakened standard of proof would make patent holders less likely to defend & challengers less likely to settle by paying royalties. In general, this will make cross patent deals more likely as well as reduce the cost of litigation for the "little guy." If you employ a kind of Occam's Razor approach to picking the best solution, a lesser standard of proof is it.
Originally, abstract math & things found in nature could not be patented. That excluded algorithms, and things like DNA. Programs were originally considered not eligible for patent protection, as an algorithm. However, several Supreme Court decisions eroded that original standard. What was left provided little guidance to the USPTO. On top of that, USPTO reviewers, while generally having engineering backgrounds, do not necessarily have backgrounds in the fields they are assigned cases for. To compound the problem, the reviewers have deadlines to keep & the USPTO looses out on (renewal) fees if patents are not granted. Hence, as you can imagine, reviewers are "encouraged" to approve patents by their management. You can imagine the effect on granting questionable or borderline software patents.
How can that failure of the system be fixed? The fix could take 3 forms; (1) Congress could remove software patents from eligibility [VERY unlikely, given lobby $$$], (2) The USPTO can be forced [Congress? SCOTUS? Again, unlikely] to reform procedures, or (3) SCOTUS can provide guidance on standards, or weaken the standard of proof in litigation [Given that this is what M$ is asking for in the I4I appeal, this is most likely].
A weakened standard of proof would make patent holders less likely to defend & challengers less likely to settle by paying royalties. In general, this will make cross patent deals more likely as well as reduce the cost of litigation for the "little guy." If you employ a kind of Occam's Razor approach to picking the best solution, a lesser standard of proof is it.
@JaseP -
You have hit, I think, on a couple of important things.
First, the SCOTUS has already begun the process of providing guidance for tighter standards, and I expect that to continue.
The standard of proof problem is a bit trickier, because the current standard of proof is fine if patents are granted properly and not fine if they aren't. The answer may be some kind of ruling or order that segregates software-only patents granted up to the time of the ruling and removes their presumptive validity with regard to obviousness. IE, if you, as a patent holder, can establish the non-obviousness of your invention, you don't need to jump through any additional hoops to retain the presumption of validity.
That would be, ahem, unprecedented, but something really needs to be done. The patent situation WRT software is a big fat mess.
You have hit, I think, on a couple of important things.
First, the SCOTUS has already begun the process of providing guidance for tighter standards, and I expect that to continue.
The standard of proof problem is a bit trickier, because the current standard of proof is fine if patents are granted properly and not fine if they aren't. The answer may be some kind of ruling or order that segregates software-only patents granted up to the time of the ruling and removes their presumptive validity with regard to obviousness. IE, if you, as a patent holder, can establish the non-obviousness of your invention, you don't need to jump through any additional hoops to retain the presumption of validity.
That would be, ahem, unprecedented, but something really needs to be done. The patent situation WRT software is a big fat mess.
@ Dino,
Since gov't workers tend to be a "lazy" bunch (at the expense of generalizing and being overly judgmental), I put little hope in reforming the USPTO. That leaves us with Court intervention.
They have made "mistakes" before, that either Congress (Search for the Wiki on the "Piano Roll Blues") or they themselves (Brown v Board of Ed, setting the stage for affirmative action) have fixed. But I hold LITTLE hope for a good decision, since both the Conservative & the Liberals on the Court seem to "kau-tau" to big business interests.
I sometimes fear that we are headed for a future like that depicted in "Aliens." I think that if trends are allowed to continue, we'll have mega-corporations controlling us, with the tacit approval of a puppet gov't.
Since gov't workers tend to be a "lazy" bunch (at the expense of generalizing and being overly judgmental), I put little hope in reforming the USPTO. That leaves us with Court intervention.
They have made "mistakes" before, that either Congress (Search for the Wiki on the "Piano Roll Blues") or they themselves (Brown v Board of Ed, setting the stage for affirmative action) have fixed. But I hold LITTLE hope for a good decision, since both the Conservative & the Liberals on the Court seem to "kau-tau" to big business interests.
I sometimes fear that we are headed for a future like that depicted in "Aliens." I think that if trends are allowed to continue, we'll have mega-corporations controlling us, with the tacit approval of a puppet gov't.
@jdixon
> krisum, I was talking to Dino, not you.
> And a simple statement of fact concerning my position can never be a straw man, since it has nothing to do with the other persons argument as such.
Strange, your reply quoted my comment, not Dino's. And "Arguing for an obviously broken system" doesn't look like a position to me rather an expression of that taken by the other person...
@hwkint
> Matter - on the other hand - is dependent on the medium. It is scarce.
Which is besides the point, since matter is not patentable either.
> Everybody on the world can use the _same_ info without stealing.
The point you are repeatedly missing is what is patented. Patent is not about protecting the building from being stolen at all. A building is not what is patented, or more precisely the matter in the building is not -- its the innovation in its design etc that is. It does not matter if that can be expressed in pure mathematics or is only "information".
As Dino has been trying to explain, its the invention that can be patented, not the maths or idea or information ... Pure mathematics is not patentable rather its application to solve a real world problem is (of course, provided it passes the criteria of original innovation, no prior art etc). Whether that be in a software system or other engineering field does not matter.
> krisum, I was talking to Dino, not you.
> And a simple statement of fact concerning my position can never be a straw man, since it has nothing to do with the other persons argument as such.
Strange, your reply quoted my comment, not Dino's. And "Arguing for an obviously broken system" doesn't look like a position to me rather an expression of that taken by the other person...
@hwkint
> Matter - on the other hand - is dependent on the medium. It is scarce.
Which is besides the point, since matter is not patentable either.
> Everybody on the world can use the _same_ info without stealing.
The point you are repeatedly missing is what is patented. Patent is not about protecting the building from being stolen at all. A building is not what is patented, or more precisely the matter in the building is not -- its the innovation in its design etc that is. It does not matter if that can be expressed in pure mathematics or is only "information".
As Dino has been trying to explain, its the invention that can be patented, not the maths or idea or information ... Pure mathematics is not patentable rather its application to solve a real world problem is (of course, provided it passes the criteria of original innovation, no prior art etc). Whether that be in a software system or other engineering field does not matter.
But - like said - there is no applied maths - or innovation. Because the applied maths / innovation partially is pure math, and partially not. And it's easy to find the border.
Innovation in design doesn't refer to pure mathematics, innovation in 'algorithms' does.
Innovation in design doesn't refer to pure mathematics, innovation in 'algorithms' does.
@kirisum -
At some point, the only honorable action is to bail out of The Twilight Zone. Good luck to you for as long as you can manage bringing The Good Fight to SerlingVille.
At some point, the only honorable action is to bail out of The Twilight Zone. Good luck to you for as long as you can manage bringing The Good Fight to SerlingVille.
Post a comment
A free account is needed to post. Write your comment here first: your text is kept in this browser while you sign up or log in, and is waiting for you when you come back.
Your message:Markdown: **bold**, *italic*, [text](https://link), - lists, > quotes, $x^2$ for math. Formatting help
Sign up and post (it is free) or Log in and post
Any software that uses I/O facilities, for example, is not pure math. Or as soon as memory comes into play. But that is not the point...
Dino has tried it multiple times, let me try once more. A civil engineering design, for example, is not bereft of invention because it used only "pure mathematics" for the modeling. Or designing a new drug nowadays typically involves modeling using software for most of the molecule design, i.e mostly "pure mathematics".
Which is a different issue. As I said the nuts and bolts of programming should obviously not be patentable.
@jdixon A hypothesis, and one that can be argued for all kinds of patents. Real problem may be the kind and number of software patents being granted.