Actually, it's a lot of sources beyond sponsors. The TV networks pay bigtime for it, the host government pays pleanty (and gets big economic returns), and the visitors pay bigtime as well. It's a shame that this still isn't enough and it requires sponsorship and advertisement.
The system's implementors would loose credability...
By that argument, any system should be implemented immediately and bugs be damned. Any testing of systems takes time and any terrorists acts during that time would lead to this same problem. It's not an excuse to bypass proper testing.
The point is that the system didn't work like you "presume". Yes, it worked eventually. But "working" or "not working" is not a binary choice here. The speed at which it works is critical. It took 3 weeks for a senator with connections to get his name cleared. How long would it take a normal citizen.
A better way to implement this would have been to have a test period where the travellers are notified that they're flagged but not stop them from flying. Each flag is then investigate to see if it is a false positive, and how long it took to clear false positives. Only after this trial period, and fixing the bugs, should they actually stop people from flying.
I'm not clear on how the difference can be detected. If spammer S sends out spam advertising site X, how is anybody going to know if X had hired S (legitimate spam) or competitor Y hired S (a joe job). Basically, you have to know who hired the person doing the spam. Will this come down to an interrogation of spammer S, review of their financial and phone records to find some connection to X or Y? Even that is fairly easy to trick by having S get X to phone them for some reason, and/or have some legitimate financial transaction between them. Perhaps there's something I'm missing that can show Y is really behind the spam.
I'm confused at why this is news, and the article states "now it can be told". The CAIB report long ago determined that the method of applying foam, including by hand, was part of the problem that caused it to sometimes come apart. Chapter 3, page 52 of the CAIB report states:
The way the foam was produced and applied, particularly in the bipod region, also contributed to its variability. Foam consists of two chemical components that must be mixed in an exact ratio and is then sprayed according to strict specifications. Foam is applied to the bipod fitting by hand to make the foam ramp, and this process may be the primary source of foam variability. Board-directed dissection of foam ramps has revealed that defects (voids, pockets, and debris) are likely due to a lack of control of various combinations
of parameters in spray-by-hand applications, which
is exacerbated by the complexity of the underlying hardware configuration. These defects often occur along ?knit lines,? the boundaries between each layer that are formed by the repeated application of thin layers ? a detail of the spray-by-hand process that contributes to foam variability, suggesting that while foam is sprayed according to approved procedures,
these procedures may be questionable if the people who devised them did not have a sufficient understanding of the properties of the foam.
On page 53 it also concludes
The precise reasons why the left bipod foam ramp was lost from the External Tank during STS-107 may never be known. The specific initiating event may likewise remain a mystery. However, it is evident that a combination of variable and pre-existing factors, such as insufficient testing and analysis in the early design stages, resulted in a highly variable and complex foam material, defects induced by an imperfect and variable application, and the results of that imperfect process, as well as severe load, thermal, pressure, vibration, acoustic, and structural launch and ascent conditions.
The news report is wrong when it says the CAIB "left the matter open". All this new work seems to be related to test and certify a new process.
The Russians still have the best technology in space.
That's a matter of perspective. They are very good at getting things to work, but it usually very low-tech. It's like saying a 1970 Ford Pickup that is still running fine is "better technology" than a 2003 Honda Prius Hybrid that keeps breaking down. (Not that this is true of those vehicles, but it demonstrates the apples and oranges comparison.)
The Russians are quite good at building reliable vehicles. But they do have their own accidents and poor administrative decisions, such as those that caused the MIR accident. Even before the accident, MIR astronauts had to deal with "...fire, power blackouts, chemical leaks, docking failures, nail-biting spacewalks, and constant mechanical breakdowns". For anyone interested, and who believe the Russians have superior technolgy or a superior program, I suggest you give Dragonfly a read. That isn't to say NASA is any better, but in some areas NASA is light years ahead and in others they need a lot of work.
I'm well aware of what you're arguing for - unlimited patentability of abstract algorithms,
Please look up the meaning of strawman argument. I believe this is about the third time you have claimed this and yet I repeatedly correct you on this. I have never argued for unlimited anything. In fact, my argument is based on limits. Yours, on the other hand, is limiting everything. You are taking an extreme and seem to believe I'm taking the opposite extreme. I am not, I am taking middle ground where everybody wins.
Existing rules are relevant because they have not appeared out of nowhere and for no reason.
It is clear you have no idea how to have a debate. This is called "appeal to authority" and is a logical falsehood. Existing rules are irrelevant to what the rules should be. I am coming from the point of view if no rules or laws exist, how would we design patent laws to work. The process would follow something along the lines of:
Do we need patents? Yes, they promote innovation and progress by allow people to exploit their ideas alone for a time and in turn forcing sharing of the idea.
What types of ideas deserve patenting? Ones that: (1) are not inevitable, such as physical laws or obvious ones, (2) would never be created without the incentive of monopolistic exploitation, and (3) are useful for progress of society. Some, but not all, devices fall in this category, as to some, but not all, algorithms.
All patents are inherently bad
If you believe that, why aren't arguing against patents in general instead of algorithms specifically? Patents are a tradeoff and a balance. The bad: they limit freedom for a limited time. The good: everyone gets to see how the idea works, and without the patent the idea might never be created. They are no more inherently bad than any other tradeoff. When you purchase something you lose money (bad) but gain the thing you purchased (good).
Pure fantasy! and not unexpected given that your 'solution' is contingent on an imaginary and idealised patent regime.
See, now you are switching arguments again. This is not an argument against my point, but against the way the patent system currently works. Again, if you missed it the umpteen times I've said it, I completely agree that system of awarding of patents is broken and in fact this is the problem. But you make the assumption that it can never be fixed or work properly. Making a system that works is not pure fantasy.
...but not stated a single one of them and surely you cannot really mean that they consist only of "basically the same rules for physcial objects (obviousness, prior art, physical laws, etc.)"
Aren't you at all uncomfortable with proving yourself wrong in the same sentence. How can you say not have stated a single criterion and then list several I did state?
... because those are exactly the criteria used to grant patents on algorithms you claim to wish to exclude.
You make no sense here either. How can rules against obvious patents be used to grand obvious patents? These rules have been improperly applied or often not at all. That is part of the broken system.
Has it never occurred to you that even invalid patents can cause harm, as can spurious accusations of infringement?
Yes, that has been my point all along. How is it that you can't understand what I am arguing when I repeat it so many times. Invalid patents are the major problem. They are the ones causing the most harm. And they are invalid, by definition. The system of awarding them is broken, and that is the problem. The rules (in North America) allowing algorithm patents is not the problem, it is the system of awarding them.
It is not up to me to demonstrate harm but for you to demonstrate the necessity of extending patentability to programme claims.
I apologize for the misunderstanding, I did not realize you were referring to specific "software" patents that violate European rules. I understood you to be referring to 30,000 harmful "software" patents. What I said is appropriate for that context. There seems to be this ongoing problem that I'm arguing about what should and shouldn't be patentable, and you keep throwing in references to specific European rules. It seems to me that existing rules are irrelevant to what should and shouldn't be patentable so I'm not sure why you keep throwing these things in.
What might suit your industry will damage or destroy others, damage individual civil liberties, damage free software, damage freedom of communication of ideas, promote monopolisation by enabling denial of interoperability etc. etc.
But again, as I've pointed out over and over and you've even recognized is a valid point, patenting algorithms is not inherently bad. The criteria I've suggested for patentable algorithms, which is basically the same rules for physcial objects (obviousness, prior art, physical laws, etc.) weed out all of the examples of any harmful ones you (or anyone else) have provided. In my solution, everybody wins. In your solution, a lot of people lose.
You seem to think that pure software idea patents are valid now but they are not
Actually, in North America they are, and as you point out they are also in some European countries. I'm not sure about the rest of the world. However, I will again point out that I am discussing what should and shouldn't be patentable. Existing rules are not relevent to that discussion.
It is from that starting point that I and perhaps others would be willing to consider amendments for special and exceptional cases such as yours.
See, here's the problem. You see them as special cases. I see them as the norm for what an algorithm is. The ones you are talking about, the ones harmful to open source and such, are the exception and shouldn't even be awarded by simple criteria that already exists for patents in general. And as I've already pointed out, algorithms are not the same thing as software. Any patent that says "Method" is by nature an algorithm.
You insist that any argument against patentability of abstract algorithms must not also apply to physical devices. But by definition it would not:
Um, no. The reason I insist that is becuase if the arguument also applies to devices, it's not an argument against software patents, it's an argument against all patents. But you say (in the above quote) that by definition any argument against algorithm patents would not apply to devices. That's easy to disprove by making one myself. A problem with algorithm patents is that a lot of "obvious" ones have been awarded. That argument also applies to devices, since it is true. There, I broke your "by definition" rule.
Again, I asked specifically to show how such algorithms (as my example) are harmful. You haven't demonstrated it and apparently you can't. Therefore any argument that algorithms are inherently harmful as patents is false. You have to give a specific reason why all algorithm patents are harmful in order to be convincing that algorithms in general should not be patentable.
Indeed much of the argument against software patentability depends on the difference between abstract entities and physical entities.
Which is exactly why you fail to have a convincing argument. I have given concrete examples of where algorithm patents are not harmful. If it is a property of algorithm patents that they are inherently harmful, I should not be able to do that. It is your contention that all are bad that is a logical falsehood. It is not supportable.
"However, it is not the physical device that is patented. It is the design, or rather the abstract conce
Now this is very interesting and to me at least it seems like a strong argument and I'll consider it your central point and return to it later.
Well, thank you for recognizing that there are valid arguments why not all algorithm patents are harmful. And yes, it is core to my opinions on this, but I have other core points that I think are strong arguments.
"You make the assumption that allowing patents in any algorithm means all "general software ideas" are patentable."
Yes, because that is the way it has happened in the U.S. and it is the reason there are > 30,000 such patents
Right, but we seem to disagree on what the root problem and solution are. The cause for the >30,000 "bad" software patents isn't because algorithms are patentable, it's because the criteria for allowing them has either been insufficiently applied (e.g., prior art, obviousness) or insufficiently defined. My solution is to fix the criteria and process by which the patents are approved. Your solution (and you are certainly not alone) is to ditch the entire concept of patenting algorithms. As I hope I've pointed out, that's throwing out the good with the bad and there are better alternatives.
The gist of your arguments seem to be based on the quotes below:
...you and I consider them obvious and unpatentable but others, especially patent examiners, may not.
What is concerning is that in all these areas, fundamental algorithms and methods are being patented - the consequent algorithmic expressions of elementary results in particular fields. To me that is entirely unacceptable.
What I may consider to be purely mathematical, non-novel, overly broad or obvious and therefore unpatentable seems never to have stopped the average patent examiner from granting it a patent.
But again, that's a problem with the process of awarding patents, not with the concept of patenting algorithms. The exact same thing is true with any form of patent. My favorite example is "Method of exercising a cat" (playing with a cat using a laser pointer). How a patent agent missed the obviousness I'll never know. The complaint you make about "general software" and "obvious" patents for algorithms applies to general patents. It isn't specific to algorithms. Getting rid of algorithm patents will not get rid of the problem in general. It's basically saying the system is broken so lets get rid of the system as it pertains to my field, but leave everyone else screwed with a still broken system. (And in the process screw over some people in my field who have "legitimately" patentable algorithms.) I think it's a lousy and inappropriate solution.
As for patenting fundamental mathematical "truths", that is analagous to patenting physical laws. They are "discoveries" not inventions. Yes, for the layperson the difference can be difficult to see, but that's true whether it's math or physics. For example, many people (including me) object to the idea of patenting genes. Again, this is an issue with respect to the criteria for awarding patents, not the whole filed (e.g., algorithms).
"Please explain to me how a patent on a very specify object recognition algorithm is harmful to software development or innovation. Keep in mind that your argument can't apply to other types of patents (mechanical, electronic, etc.)"
The condition of your second sentence is easy to meet: As I have said before, no-one I know of has any problem with the patenting of a physical device that implements some algorithm, even if that is all that it does. The problem is with your insistence that the algorithm itself should be patentable, thereby rendering it's implementation in a general purpose computer subject to legal sanction by the patentee.
There are several problems with this argument; you state t
I do not consider it naive to properly distinguish between instruction and implementation.
It's naive because to the end user there is no difference between a box that performs a function completely electronically versus one that does the exact same thing through instructions running on a computer. It is figuring out how to do it that is the hard part. Soldering a circuit board isn't much harder than writing it in code. The intent of patents, to motivate innovation and progress by exchanging a limited monopoly for forced publication of the details, applies to algorithms as much as anything else. They take just as much time, effort, money, imagination, etc., to develop as a mechanical device. In terms of innovation and progress, the "what it does" is important to the user. The "how it works" is what is patentable.
it is far better that society maybe loses the benefits of disclosure in a few narrow fields than suffers the widespread damaging effects of general software idea patentability
See, here's where you, and a lot of people, fail in your argument. You make the assumption that allowing patents in any algorithm means all "general software ideas" are patentable. Why not change "software" in your sentence to "mechanical" or "electrical" to see how ridiculous that is.
I have yet to see anyone make a complaint about algorithms in object recognition, tracking and measuring the pose of an object, reconstructing surfaces from images, etc., etc. These things take years of research and development and require as much expertise, ingenuity, expense, creativity, and time as any other type of invention. They are not obvious or trivial, and in no sense could patenting them ever cause any of the problems people are complaining about. Nobody could ever "accidently" code them and later find out they were patented. Nobody could see them performing their functions and then run off and do the same thing, not realizing they're patented. (This would make them "obvious" by definition.) The only way to duplicate these types of algorithms is by finding a publication on how they work. I have yet to hear any sort of explanation of how these types of patents would hurt anyone. The only people that it would stop are those who are intentionally trying to copy the work. That's the same thing that "normal" patents stop, regardless of the medium the patent is in. There is no difference in reasoning between "algorithms" and other types of patents for this type of development.
A patent on a method for object recognition does not mean all object recognition is patented. The "what it does" is not the same thing as "how it works". All the complaints are about plug-ins, one-click shopping, and so on. These are not "how it works" and shouldn't be patentable. In fact, most people point out the obviousness and prior art violations of most of these patents. Then there's the "obvious" ones, such as the famous XOR patent.
It is not true that I and others are complaining only about simple or general software patents.
Yes. You are arguing about "the validity of software patentability", but you are complaining about simple and general software solutions. All of the examples I've seen so far fall into this category. The problem is, the "solution" people are suggesting doesn't fit the problem. It's throwing out the baby with the bathwater.
But feel free to change this record. Please explain to me how a patent on a very specify object recognition algorithm is harmful to software development or innovation. Keep in mind that your argument can't apply to other types of patents (mechanical, electronic, etc.) otherwise you are arguing against the whole concept of patents, not just "algorithm" patents.
The granting of monopoly rights is the means by which the purpose is fulfilled, not the purpose itself.
Always? No. Nobody's perfect. But a less-than-perfectly implemented humanitarian effort to feed starving people is a hell of a lot better than going in with guns blazing for manufactured reasons with the intent of sorting it all out later.
What I find naive is the idea that instructions in software is not an "implementation". If I have a box that does someting, say given some input it produces some output, then it doesn't really matter to the end user whether it's done mechanically, electronically, or through some software code. Yes, the box isn't itself patentable, the "how" it works is what is patentable, but it doesn't change the fact of what patents are for. They are incentives, in the form of a limited monopoly, to invest in and produce new ways of doing things. The tradeoff is that in exchange for this limited time monopoly, you must publish the details of how this new thing works so that others can use it or build from it once your patent has expired.
This incentive is true whether the new thing is mechanical, electrical, or some algorithm or method for doing something. As a concrete example, I'm one of the primary developers of an algorithm for automatic target recognition. Our company invested much time and money into the development. If we couldn't patent it, our competition could implement the exact same thing and we'd be screwed. Since they have more money and resources, they could develop it a lot faster than we could and they are already in the right circles to keep us out of the industry. In the fields of defense, security, and aerospace, the reputation, history, and contacts of your company are quite important and there are others who are much deeper in than us.
If we couldn't patent the algorithms, what's our incentive for developing them or investing in developing them? On of our competitors could just steal it and use it instead. (If they write the algorithm in their own code from scratch, it doesn't violate copyright.) Our only other alternative would be to try to keep it as secret as possible, which means never publishing how it is done, and gambling on nobody ever reverse engineering it. In either case, either not developing it or keeping it a secret, the public in general would never benefit from knowledge of how it worked. In other words, the benefit side of a patent, progress, would never happen.
The general point I'm trying to make is that the development of complex (non-obvious) algorithms has the exact same justification for patents as any other type of solution such as mechanical or electrical.
What people have been complaining about is simple or general "software" patents (e.g., one-click shopping). The problem with these is that they are too obvious, have prior art, or are too general. (For example, we couldn't patent "automatic object recognition" since it isn't a method, but we could patent one method for doing it.) Patenting mathematical solutions also doesn't fall into this category. There are an infinite number of ways to do automatic target recogntion, some good and some bad. Mathematical solutions are like physical laws, they are unique solutions.
It [making money] may be your purpose, but it is not the purpose (or intention) of the law
Yes, in fact it is, at least indirectly. Referring again to the U.S. Constistution description of patents, it is for the purpose of promoting progress. As I mentioned above, the way it does this is by providing a limited monopoly in exchange for publishing how it works. The point of the limited monopoly is to allow you to make back your investment and some profit before it is released. It is true that this monopoly doesn't have to be used for profit, just the monopoly control itself is the incentive. But generally, the point is to make back your investment plus a little more.
Again, discussing semantics and wording of specific laws is generally irrelevant. The real argument to me is how does the develop of algorithms, as in my personal example above, fit in with the intent of patents for other items, that is, the tradeoff of a limited monopoly as an incentive for publishing the work. It's not only an ideological question to me, I'm living it. Without the ability to patent, my company never would have invested in the work we've done and developed these things. That's real, and a real result of what you are arguing.
What if our situation is unique or we are the first? You can't build any assumptions from a sample size of one.
There are two problems with this. First, yes, you indeed can build assumptions from a sample size of one, or less. That's the nature of assumptions, they are made without solid data to back them up. If the sample size was much larger, it wouldn't be an assumption, it'd be based on statistics.
Second, you seem to be doing just that. We have one sample, and it tells us that 100% of the solar systems (or galaxies) that we currently have the capability of investigating for life indeed do have life. Since it is a binary question (is there life elsewhere or not), it is less credible to make an assumption against the statistical evidence even if it is a sample size of one.
Put another way, without any other knowledge, the binary question of "Is there life out there or not?" is a 50/50 chance either way. Add to this a sample of one that supports one way (there is life). Even if this is infinitessimal in weight, it still tilts the odds slighly towards it being more likely than not that there is other life out there.
I find this a bit naive. Basically, if I solve a problem and I can do it through software or a mechanical or electronic implementation, what you are suggesting means that if I'm interested in making money on it (which is the purpose of patents), I should do the mechanical or electrical implementation. That way, I can patent it and make money off of it. If I did it through software I couldn't stop somebody from duplicating it (even by re-writing it to avoid copyright violation). So, even though the software approach would be cheaper and more reliable for the consumer, I'd end up making more money by doing it some other way. At least, that's what the world would be like if these things weren't patentable as software -- we'd just avoid doing it in software whenever possible. That's bad, and counter to progress which is the point of patents.
"Patents and copyrights are identical in intent"
No they are not:
Yes, they are. From the U.S. Constitution:
"The Congress shall have power . . . To promote the progress of science and useful arts by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries."
They are intented to promote progress by securing exclusive rights for limited time. If the time was zero (no protection) then people would not have monetary incentive to release their creations. If the time was forever, nobody could build from it so progress is stopped. This argument is true even if it is a software development.
At least in Europe, patents on electronics (in the sense of chip designs) are *not* patentable.
...
ext, in case of a mechanical construction, you do not get a patent on "X implemented in a mechanical way"
This isn't quite what I meant. Perhaps an example will do. How about the flyball governor. In case you don't know, this is that little spinning device with two balls and linkages that you see in movies and such to control steam engines. It was designed to control the speed of a steam train. It's a fairly ingeneous device for the time. It would certainly fall in the category of patentable by any stretch of the imagination.
Now, anyone who knows control theory knows that this is just a simple proportional controller. The exact same control system could be patented, under a different pattent, if it was done electronically. That is, if the control was done by putting a speed sensor on the train feeding into an electronics board which controlled the valve system through a proportional gain which can be set by, say, a potentiometer in the control box. This would produce the exact same result as the flyball governor because it is the exact same control system, but done in electronics instead of mechanics.
Now, if it was done in software, i.e., ValveSetting = Gain*(TrainSpeed - SetTrainSpeed), the result would also be the same, just the "medium" of implementation would be different. So, why would the mechanical and electrical implementations be patentable but not the software implementation?
There are many examples of where an implementation can be done mechanically, electronically, or in software, and yet it's the same underlying algorithm. How about edge detection in a camera? There are patented edge detection techniques done in software. You could also do the same thing in hardware directly without any code. What about encryption? That could be done using a mechanical or electrical machine without any code or done in code. It's the same algorithm. But a "device" that does it is certainly patentable. Why not if done in software?
Again, this bring me to the point that software isn't patentable. People aren't complaining about actual patents on software, they're complaining about patents on algorithms and methods that are implemented in software. If there's reason why it should be patentable in mechanical or electrical systems, I see no reason it shouldn't be patentable in software, as long as it meets the rules for non-obviousness and no prior art.
Finally (and most importantly), there is no inherent "inventor's right". This is in stark contrast with copyright, where there is an inherent moral right of the creator on his creation.
I completely disagree here. Copyright is not an inherent right. It exists solely as a means to promote creation and publication of works by giving the creators limited control over it for a limited time. In the U.S. (though I'm not from there), this is even written into the constitution. Patents and copyrights are identical in intent, they just differ in what they apply to. Patents are for "tangible" things (devices, constructs, etc.) limited by physical laws. Copyrights are for "expressions", i.e., creative works that are only limited by imagination.
So, again, I don't see the inherent problem. It's the poor approval (and objection) process for patents that seems to be problem.
I agree that software patents aren't the problem. In fact, I'm unaware of any actual patents on software. There are patents on methods or algorithms that are implemented in software, but that's justifiable. For example, if I made a control systems algorithm that I implemented mechanically or electronically it would certainly be patentable, so why shouldn't it be patentable if I implement it in software?
What seems to be the problem is the defacto removal of patent restrictions regarding obviousness and prior art, along with the patent duration. Very obvious things are getting patented in software. It's like patenting a plastic doll because it now has a new hat.
But my biggest concern with this Munich thing is why the concern over Linux specifically? Patent violations are just as, or more, likely in proprietary software. In fact, all litigation I'm aware of for so-called "software" patents have been between proprietary software companies and there have been no lawsuits over Linux violations yet. Plus it is quite obvious that Linux developers would be quick to make new "non-violating" implementations very quickly if anything was found. In other words, the reasoning behind slowing Linux implementation in Munich makes no sense to me.
Ah. It looks like we're in general agreement then. I also agree that satire should be protected as fair use. Hell, how many political cartoons have made use of copyrighted material? But what I was pointing out is what the state of the law currently appears to be, according to the Cat in the Hat case and opinions of legal experts (including Lawrence Lessig).
I would be quite interested in seeing the Cat in the Hat case overturned, and I agree with your points about what the first amendment and fair use should protect. I guess it's just a matter of waiting to see what happens, though I not sure this case would clarify it since, as I originally said, there's a good argument for parody in addition to the satire.
A dancing picture of Bush in bra and panties singing "Oops, I Lied to the Public Again" would be protected fair use.
I suspect you didn't read the article. There's a difference between satire and parody. A parody makes fun of the copyrighted work and/or creator/artist. So this would be a parody if it was making fun of Britney Spears or the message of her song. But if it's making fun of Bush it has nothing to do with Britney's song and therefore it is a satire, which is not protected by fair use. (The case in precedence here is the use of Cat in the Hat to make fun of the OJ Simpson verdict, which was ruled as infringement because it wasn't making a comment about the copyrighted work or Dr. Seuss but rather about OJ Simpson.)
While this has become more common, especially in games and DVDs, that is not the history of easter eggs. A lot of them have had nothing to do with the software, were unknown by the majority of users (and hence useless for marketing), and were "signatures" or jokes by the programmers.
The intent is to make a political message about the government not to parady the song.
Not so fast. The primary intent was to make political satire. However, this song was chosen specifically to contrast the political status with the message of the song. As well, the song often used to promote patriotism by those running, and running for, government despite the fact that it is very anti-government.
If they had just picked a random song, you'd probably be right. But because of the specifics of the contrast between the original song message, current political status, and typical use of the song by government, it seems to be very much a parody. IANAL, but I think there's a solid argument there for parody. AFAIK, the parody doesn't have to be the primary intent of the song to make it fair use.
Slashdot Mirror
Actually, it's a lot of sources beyond sponsors. The TV networks pay bigtime for it, the host government pays pleanty (and gets big economic returns), and the visitors pay bigtime as well. It's a shame that this still isn't enough and it requires sponsorship and advertisement.
Looks like NASA finally ran out of acronyms. SVS is also used for their Space Vision System.
By that argument, any system should be implemented immediately and bugs be damned. Any testing of systems takes time and any terrorists acts during that time would lead to this same problem. It's not an excuse to bypass proper testing.
Um, they can do that now. It doesn't depend on whether they actually get to fly or not.
A better way to implement this would have been to have a test period where the travellers are notified that they're flagged but not stop them from flying. Each flag is then investigate to see if it is a false positive, and how long it took to clear false positives. Only after this trial period, and fixing the bugs, should they actually stop people from flying.
I'm not clear on how the difference can be detected. If spammer S sends out spam advertising site X, how is anybody going to know if X had hired S (legitimate spam) or competitor Y hired S (a joe job). Basically, you have to know who hired the person doing the spam. Will this come down to an interrogation of spammer S, review of their financial and phone records to find some connection to X or Y? Even that is fairly easy to trick by having S get X to phone them for some reason, and/or have some legitimate financial transaction between them. Perhaps there's something I'm missing that can show Y is really behind the spam.
Thunder ... Fire ... Sun .. all hot, loud, destructive things. I guess there's no cuddly hippie people at Mozilla.
Yeah, can you imagine a Beowulf cluster of these things?
What do you mean if. BTW, looks like you had a little too much spicy food last night, huh? And next time, don't forget to wash your hands afterward.
On page 53 it also concludes
The news report is wrong when it says the CAIB "left the matter open". All this new work seems to be related to test and certify a new process.
That's a matter of perspective. They are very good at getting things to work, but it usually very low-tech. It's like saying a 1970 Ford Pickup that is still running fine is "better technology" than a 2003 Honda Prius Hybrid that keeps breaking down. (Not that this is true of those vehicles, but it demonstrates the apples and oranges comparison.)
The Russians are quite good at building reliable vehicles. But they do have their own accidents and poor administrative decisions, such as those that caused the MIR accident. Even before the accident, MIR astronauts had to deal with "...fire, power blackouts, chemical leaks, docking failures, nail-biting spacewalks, and constant mechanical breakdowns". For anyone interested, and who believe the Russians have superior technolgy or a superior program, I suggest you give Dragonfly a read. That isn't to say NASA is any better, but in some areas NASA is light years ahead and in others they need a lot of work.
Please look up the meaning of strawman argument. I believe this is about the third time you have claimed this and yet I repeatedly correct you on this. I have never argued for unlimited anything. In fact, my argument is based on limits. Yours, on the other hand, is limiting everything. You are taking an extreme and seem to believe I'm taking the opposite extreme. I am not, I am taking middle ground where everybody wins.
Existing rules are relevant because they have not appeared out of nowhere and for no reason.
It is clear you have no idea how to have a debate. This is called "appeal to authority" and is a logical falsehood. Existing rules are irrelevant to what the rules should be. I am coming from the point of view if no rules or laws exist, how would we design patent laws to work. The process would follow something along the lines of:
Do we need patents? Yes, they promote innovation and progress by allow people to exploit their ideas alone for a time and in turn forcing sharing of the idea.
What types of ideas deserve patenting? Ones that: (1) are not inevitable, such as physical laws or obvious ones, (2) would never be created without the incentive of monopolistic exploitation, and (3) are useful for progress of society. Some, but not all, devices fall in this category, as to some, but not all, algorithms.
All patents are inherently bad
If you believe that, why aren't arguing against patents in general instead of algorithms specifically? Patents are a tradeoff and a balance. The bad: they limit freedom for a limited time. The good: everyone gets to see how the idea works, and without the patent the idea might never be created. They are no more inherently bad than any other tradeoff. When you purchase something you lose money (bad) but gain the thing you purchased (good).
Pure fantasy! and not unexpected given that your 'solution' is contingent on an imaginary and idealised patent regime.
See, now you are switching arguments again. This is not an argument against my point, but against the way the patent system currently works. Again, if you missed it the umpteen times I've said it, I completely agree that system of awarding of patents is broken and in fact this is the problem. But you make the assumption that it can never be fixed or work properly. Making a system that works is not pure fantasy.
Aren't you at all uncomfortable with proving yourself wrong in the same sentence. How can you say not have stated a single criterion and then list several I did state?
You make no sense here either. How can rules against obvious patents be used to grand obvious patents? These rules have been improperly applied or often not at all. That is part of the broken system.
Has it never occurred to you that even invalid patents can cause harm, as can spurious accusations of infringement?
Yes, that has been my point all along. How is it that you can't understand what I am arguing when I repeat it so many times. Invalid patents are the major problem. They are the ones causing the most harm. And they are invalid, by definition. The system of awarding them is broken, and that is the problem. The rules (in North America) allowing algorithm patents is not the problem, it is the system of awarding them.
It is not up to me to demonstrate harm but for you to demonstrate the necessity of extending patentability to programme claims.
(1) I have already done so with yo
I apologize for the misunderstanding, I did not realize you were referring to specific "software" patents that violate European rules. I understood you to be referring to 30,000 harmful "software" patents. What I said is appropriate for that context. There seems to be this ongoing problem that I'm arguing about what should and shouldn't be patentable, and you keep throwing in references to specific European rules. It seems to me that existing rules are irrelevant to what should and shouldn't be patentable so I'm not sure why you keep throwing these things in.
What might suit your industry will damage or destroy others, damage individual civil liberties, damage free software, damage freedom of communication of ideas, promote monopolisation by enabling denial of interoperability etc. etc.
But again, as I've pointed out over and over and you've even recognized is a valid point, patenting algorithms is not inherently bad. The criteria I've suggested for patentable algorithms, which is basically the same rules for physcial objects (obviousness, prior art, physical laws, etc.) weed out all of the examples of any harmful ones you (or anyone else) have provided. In my solution, everybody wins. In your solution, a lot of people lose.
You seem to think that pure software idea patents are valid now but they are not
Actually, in North America they are, and as you point out they are also in some European countries. I'm not sure about the rest of the world. However, I will again point out that I am discussing what should and shouldn't be patentable. Existing rules are not relevent to that discussion.
It is from that starting point that I and perhaps others would be willing to consider amendments for special and exceptional cases such as yours.
See, here's the problem. You see them as special cases. I see them as the norm for what an algorithm is. The ones you are talking about, the ones harmful to open source and such, are the exception and shouldn't even be awarded by simple criteria that already exists for patents in general. And as I've already pointed out, algorithms are not the same thing as software. Any patent that says "Method" is by nature an algorithm.
You insist that any argument against patentability of abstract algorithms must not also apply to physical devices. But by definition it would not:
Um, no. The reason I insist that is becuase if the arguument also applies to devices, it's not an argument against software patents, it's an argument against all patents. But you say (in the above quote) that by definition any argument against algorithm patents would not apply to devices. That's easy to disprove by making one myself. A problem with algorithm patents is that a lot of "obvious" ones have been awarded. That argument also applies to devices, since it is true. There, I broke your "by definition" rule.
Again, I asked specifically to show how such algorithms (as my example) are harmful. You haven't demonstrated it and apparently you can't. Therefore any argument that algorithms are inherently harmful as patents is false. You have to give a specific reason why all algorithm patents are harmful in order to be convincing that algorithms in general should not be patentable.
Indeed much of the argument against software patentability depends on the difference between abstract entities and physical entities.
Which is exactly why you fail to have a convincing argument. I have given concrete examples of where algorithm patents are not harmful. If it is a property of algorithm patents that they are inherently harmful, I should not be able to do that. It is your contention that all are bad that is a logical falsehood. It is not supportable.
"However, it is not the physical device that is patented. It is the design, or rather the abstract conce
Well, thank you for recognizing that there are valid arguments why not all algorithm patents are harmful. And yes, it is core to my opinions on this, but I have other core points that I think are strong arguments.
Right, but we seem to disagree on what the root problem and solution are. The cause for the >30,000 "bad" software patents isn't because algorithms are patentable, it's because the criteria for allowing them has either been insufficiently applied (e.g., prior art, obviousness) or insufficiently defined. My solution is to fix the criteria and process by which the patents are approved. Your solution (and you are certainly not alone) is to ditch the entire concept of patenting algorithms. As I hope I've pointed out, that's throwing out the good with the bad and there are better alternatives.
The gist of your arguments seem to be based on the quotes below:
What is concerning is that in all these areas, fundamental algorithms and methods are being patented - the consequent algorithmic expressions of elementary results in particular fields. To me that is entirely unacceptable.
What I may consider to be purely mathematical, non-novel, overly broad or obvious and therefore unpatentable seems never to have stopped the average patent examiner from granting it a patent.
But again, that's a problem with the process of awarding patents, not with the concept of patenting algorithms. The exact same thing is true with any form of patent. My favorite example is "Method of exercising a cat" (playing with a cat using a laser pointer). How a patent agent missed the obviousness I'll never know. The complaint you make about "general software" and "obvious" patents for algorithms applies to general patents. It isn't specific to algorithms. Getting rid of algorithm patents will not get rid of the problem in general. It's basically saying the system is broken so lets get rid of the system as it pertains to my field, but leave everyone else screwed with a still broken system. (And in the process screw over some people in my field who have "legitimately" patentable algorithms.) I think it's a lousy and inappropriate solution.
As for patenting fundamental mathematical "truths", that is analagous to patenting physical laws. They are "discoveries" not inventions. Yes, for the layperson the difference can be difficult to see, but that's true whether it's math or physics. For example, many people (including me) object to the idea of patenting genes. Again, this is an issue with respect to the criteria for awarding patents, not the whole filed (e.g., algorithms).
There are several problems with this argument; you state t
It's naive because to the end user there is no difference between a box that performs a function completely electronically versus one that does the exact same thing through instructions running on a computer. It is figuring out how to do it that is the hard part. Soldering a circuit board isn't much harder than writing it in code. The intent of patents, to motivate innovation and progress by exchanging a limited monopoly for forced publication of the details, applies to algorithms as much as anything else. They take just as much time, effort, money, imagination, etc., to develop as a mechanical device. In terms of innovation and progress, the "what it does" is important to the user. The "how it works" is what is patentable.
it is far better that society maybe loses the benefits of disclosure in a few narrow fields than suffers the widespread damaging effects of general software idea patentability
See, here's where you, and a lot of people, fail in your argument. You make the assumption that allowing patents in any algorithm means all "general software ideas" are patentable. Why not change "software" in your sentence to "mechanical" or "electrical" to see how ridiculous that is.
I have yet to see anyone make a complaint about algorithms in object recognition, tracking and measuring the pose of an object, reconstructing surfaces from images, etc., etc. These things take years of research and development and require as much expertise, ingenuity, expense, creativity, and time as any other type of invention. They are not obvious or trivial, and in no sense could patenting them ever cause any of the problems people are complaining about. Nobody could ever "accidently" code them and later find out they were patented. Nobody could see them performing their functions and then run off and do the same thing, not realizing they're patented. (This would make them "obvious" by definition.) The only way to duplicate these types of algorithms is by finding a publication on how they work. I have yet to hear any sort of explanation of how these types of patents would hurt anyone. The only people that it would stop are those who are intentionally trying to copy the work. That's the same thing that "normal" patents stop, regardless of the medium the patent is in. There is no difference in reasoning between "algorithms" and other types of patents for this type of development.
A patent on a method for object recognition does not mean all object recognition is patented. The "what it does" is not the same thing as "how it works". All the complaints are about plug-ins, one-click shopping, and so on. These are not "how it works" and shouldn't be patentable. In fact, most people point out the obviousness and prior art violations of most of these patents. Then there's the "obvious" ones, such as the famous XOR patent.
It is not true that I and others are complaining only about simple or general software patents.
Yes. You are arguing about "the validity of software patentability", but you are complaining about simple and general software solutions. All of the examples I've seen so far fall into this category. The problem is, the "solution" people are suggesting doesn't fit the problem. It's throwing out the baby with the bathwater.
But feel free to change this record. Please explain to me how a patent on a very specify object recognition algorithm is harmful to software development or innovation. Keep in mind that your argument can't apply to other types of patents (mechanical, electronic, etc.) otherwise you are arguing against the whole concept of patents, not just "algorithm" patents.
The granting of monopoly rights is the means by which the purpose is fulfilled, not the purpose itself.
Which exactly fits what I sa
Always? No. Nobody's perfect. But a less-than-perfectly implemented humanitarian effort to feed starving people is a hell of a lot better than going in with guns blazing for manufactured reasons with the intent of sorting it all out later.
What I find naive is the idea that instructions in software is not an "implementation". If I have a box that does someting, say given some input it produces some output, then it doesn't really matter to the end user whether it's done mechanically, electronically, or through some software code. Yes, the box isn't itself patentable, the "how" it works is what is patentable, but it doesn't change the fact of what patents are for. They are incentives, in the form of a limited monopoly, to invest in and produce new ways of doing things. The tradeoff is that in exchange for this limited time monopoly, you must publish the details of how this new thing works so that others can use it or build from it once your patent has expired.
This incentive is true whether the new thing is mechanical, electrical, or some algorithm or method for doing something. As a concrete example, I'm one of the primary developers of an algorithm for automatic target recognition. Our company invested much time and money into the development. If we couldn't patent it, our competition could implement the exact same thing and we'd be screwed. Since they have more money and resources, they could develop it a lot faster than we could and they are already in the right circles to keep us out of the industry. In the fields of defense, security, and aerospace, the reputation, history, and contacts of your company are quite important and there are others who are much deeper in than us.
If we couldn't patent the algorithms, what's our incentive for developing them or investing in developing them? On of our competitors could just steal it and use it instead. (If they write the algorithm in their own code from scratch, it doesn't violate copyright.) Our only other alternative would be to try to keep it as secret as possible, which means never publishing how it is done, and gambling on nobody ever reverse engineering it. In either case, either not developing it or keeping it a secret, the public in general would never benefit from knowledge of how it worked. In other words, the benefit side of a patent, progress, would never happen.
The general point I'm trying to make is that the development of complex (non-obvious) algorithms has the exact same justification for patents as any other type of solution such as mechanical or electrical.
What people have been complaining about is simple or general "software" patents (e.g., one-click shopping). The problem with these is that they are too obvious, have prior art, or are too general. (For example, we couldn't patent "automatic object recognition" since it isn't a method, but we could patent one method for doing it.) Patenting mathematical solutions also doesn't fall into this category. There are an infinite number of ways to do automatic target recogntion, some good and some bad. Mathematical solutions are like physical laws, they are unique solutions.
It [making money] may be your purpose, but it is not the purpose (or intention) of the law
Yes, in fact it is, at least indirectly. Referring again to the U.S. Constistution description of patents, it is for the purpose of promoting progress. As I mentioned above, the way it does this is by providing a limited monopoly in exchange for publishing how it works. The point of the limited monopoly is to allow you to make back your investment and some profit before it is released. It is true that this monopoly doesn't have to be used for profit, just the monopoly control itself is the incentive. But generally, the point is to make back your investment plus a little more.
Again, discussing semantics and wording of specific laws is generally irrelevant. The real argument to me is how does the develop of algorithms, as in my personal example above, fit in with the intent of patents for other items, that is, the tradeoff of a limited monopoly as an incentive for publishing the work. It's not only an ideological question to me, I'm living it. Without the ability to patent, my company never would have invested in the work we've done and developed these things. That's real, and a real result of what you are arguing.
There are two problems with this. First, yes, you indeed can build assumptions from a sample size of one, or less. That's the nature of assumptions, they are made without solid data to back them up. If the sample size was much larger, it wouldn't be an assumption, it'd be based on statistics.
Second, you seem to be doing just that. We have one sample, and it tells us that 100% of the solar systems (or galaxies) that we currently have the capability of investigating for life indeed do have life. Since it is a binary question (is there life elsewhere or not), it is less credible to make an assumption against the statistical evidence even if it is a sample size of one.
Put another way, without any other knowledge, the binary question of "Is there life out there or not?" is a 50/50 chance either way. Add to this a sample of one that supports one way (there is life). Even if this is infinitessimal in weight, it still tilts the odds slighly towards it being more likely than not that there is other life out there.
I find this a bit naive. Basically, if I solve a problem and I can do it through software or a mechanical or electronic implementation, what you are suggesting means that if I'm interested in making money on it (which is the purpose of patents), I should do the mechanical or electrical implementation. That way, I can patent it and make money off of it. If I did it through software I couldn't stop somebody from duplicating it (even by re-writing it to avoid copyright violation). So, even though the software approach would be cheaper and more reliable for the consumer, I'd end up making more money by doing it some other way. At least, that's what the world would be like if these things weren't patentable as software -- we'd just avoid doing it in software whenever possible. That's bad, and counter to progress which is the point of patents.
"Patents and copyrights are identical in intent"
No they are not:
Yes, they are. From the U.S. Constitution:
They are intented to promote progress by securing exclusive rights for limited time. If the time was zero (no protection) then people would not have monetary incentive to release their creations. If the time was forever, nobody could build from it so progress is stopped. This argument is true even if it is a software development.
...
ext, in case of a mechanical construction, you do not get a patent on "X implemented in a mechanical way"
This isn't quite what I meant. Perhaps an example will do. How about the flyball governor. In case you don't know, this is that little spinning device with two balls and linkages that you see in movies and such to control steam engines. It was designed to control the speed of a steam train. It's a fairly ingeneous device for the time. It would certainly fall in the category of patentable by any stretch of the imagination.
Now, anyone who knows control theory knows that this is just a simple proportional controller. The exact same control system could be patented, under a different pattent, if it was done electronically. That is, if the control was done by putting a speed sensor on the train feeding into an electronics board which controlled the valve system through a proportional gain which can be set by, say, a potentiometer in the control box. This would produce the exact same result as the flyball governor because it is the exact same control system, but done in electronics instead of mechanics.
Now, if it was done in software, i.e., ValveSetting = Gain*(TrainSpeed - SetTrainSpeed), the result would also be the same, just the "medium" of implementation would be different. So, why would the mechanical and electrical implementations be patentable but not the software implementation?
There are many examples of where an implementation can be done mechanically, electronically, or in software, and yet it's the same underlying algorithm. How about edge detection in a camera? There are patented edge detection techniques done in software. You could also do the same thing in hardware directly without any code. What about encryption? That could be done using a mechanical or electrical machine without any code or done in code. It's the same algorithm. But a "device" that does it is certainly patentable. Why not if done in software?
Again, this bring me to the point that software isn't patentable. People aren't complaining about actual patents on software, they're complaining about patents on algorithms and methods that are implemented in software. If there's reason why it should be patentable in mechanical or electrical systems, I see no reason it shouldn't be patentable in software, as long as it meets the rules for non-obviousness and no prior art.
Finally (and most importantly), there is no inherent "inventor's right". This is in stark contrast with copyright, where there is an inherent moral right of the creator on his creation.
I completely disagree here. Copyright is not an inherent right. It exists solely as a means to promote creation and publication of works by giving the creators limited control over it for a limited time. In the U.S. (though I'm not from there), this is even written into the constitution. Patents and copyrights are identical in intent, they just differ in what they apply to. Patents are for "tangible" things (devices, constructs, etc.) limited by physical laws. Copyrights are for "expressions", i.e., creative works that are only limited by imagination.
So, again, I don't see the inherent problem. It's the poor approval (and objection) process for patents that seems to be problem.
I agree that software patents aren't the problem. In fact, I'm unaware of any actual patents on software. There are patents on methods or algorithms that are implemented in software, but that's justifiable. For example, if I made a control systems algorithm that I implemented mechanically or electronically it would certainly be patentable, so why shouldn't it be patentable if I implement it in software?
What seems to be the problem is the defacto removal of patent restrictions regarding obviousness and prior art, along with the patent duration. Very obvious things are getting patented in software. It's like patenting a plastic doll because it now has a new hat.
But my biggest concern with this Munich thing is why the concern over Linux specifically? Patent violations are just as, or more, likely in proprietary software. In fact, all litigation I'm aware of for so-called "software" patents have been between proprietary software companies and there have been no lawsuits over Linux violations yet. Plus it is quite obvious that Linux developers would be quick to make new "non-violating" implementations very quickly if anything was found. In other words, the reasoning behind slowing Linux implementation in Munich makes no sense to me.
I would be quite interested in seeing the Cat in the Hat case overturned, and I agree with your points about what the first amendment and fair use should protect. I guess it's just a matter of waiting to see what happens, though I not sure this case would clarify it since, as I originally said, there's a good argument for parody in addition to the satire.
I suspect you didn't read the article. There's a difference between satire and parody. A parody makes fun of the copyrighted work and/or creator/artist. So this would be a parody if it was making fun of Britney Spears or the message of her song. But if it's making fun of Bush it has nothing to do with Britney's song and therefore it is a satire, which is not protected by fair use. (The case in precedence here is the use of Cat in the Hat to make fun of the OJ Simpson verdict, which was ruled as infringement because it wasn't making a comment about the copyrighted work or Dr. Seuss but rather about OJ Simpson.)
While this has become more common, especially in games and DVDs, that is not the history of easter eggs. A lot of them have had nothing to do with the software, were unknown by the majority of users (and hence useless for marketing), and were "signatures" or jokes by the programmers.
Not so fast. The primary intent was to make political satire. However, this song was chosen specifically to contrast the political status with the message of the song. As well, the song often used to promote patriotism by those running, and running for, government despite the fact that it is very anti-government.
If they had just picked a random song, you'd probably be right. But because of the specifics of the contrast between the original song message, current political status, and typical use of the song by government, it seems to be very much a parody. IANAL, but I think there's a solid argument there for parody. AFAIK, the parody doesn't have to be the primary intent of the song to make it fair use.