Asking patent law experts to define the concept of computer-implemented invention?
I think the anti-software-patent movement has somehow failed to analyze the problem properly, and the MEPs are confused. While there are some real differences between the field of software and most other fields, these differences do not make it obvious that software patents is wrong.
The matter with the software field is that patents do more harm in this field, but they do largely the same kind of harm as in other fields. The arguments the movement has put forward have failed to convince the MEPs and have sent them reaching for some kind of compromise or Salomonic solution. All the thousands of explanations of why software patents is a bad idea end up being very confusing, because they normally leave the MEPs wondering why those arguments apply only to software and not to other fields.
We pretend that software is absolutely different, but all we can show are fairly relative differences.
The first true problem is that American courts have invented a test of non-obviousness that is very convenient for the courts and the patent offices. Convenient in the sense that you can determine objectively if that test has been met. Unfortunately the test does not test non-obviousness at all. In reallity it only tests a certain form of non-novelty.
The test consists in requiring the combination of two known principles to be "suggested" or "motivated" in some prior art text, before deeming this combination to be "obvious". If no such prior art text can be found, that combination is deemed "non-obvious". In this way almost everything is "non-obvious".
This obviousness test has not been tried in the Supreme Court yet, some twenty years after its introduction. However, there is a case, KSR International v. Teleflex, that may become such a trial in a relatively near future, if the Supreme Court decides to look at it.
Of course, in the name of harmonization, the European patent offices tend to adopt similar criteria; 260 million Americans cannot be wrong for twenty years.
We should formulate this problem clearly. I guess this will convince the MEPs much more easily. For some formulations and arguments, just have a look at http://www.pff.org/issues-pubs/other/ksramicus.pdf . Then we should work out some rather strong formulations of what it means to test for obviousness.
After that the rest is a lot easier. Suppose a method of desalination of seawater required you to add a certain acid three times in half-hour intervals. Suppose this method is patentable according to all the usual considerations. Does that mean that whoever patents it, patents the number three? No. Everyone agrees to that. The number three is not by itself patented. What if you use a computer to help you count to three? Is the process no longer patentable? What ought to be patened is the chemical process. Wether you use a computer to assist you in carrying out the steps is irrelevant. Only the steps themselves are relevant.
In the case of the computer-controlled breaking mechanism, the computer seems to be an essential part, because of its speed. If you were to read the velocity of the wheel's rotation on a digital display, calculate by hand the difference from the previous reading in order to determine if the wheels are about to block, you would probably relax the breaking force a little bit too late. However, the inventors of the automatic breaking system did not invent the computer. The advent of the computer made a large number of obvious applications possible, that had not been possible previously. To realize that, is not a patentable invention, it is something obvious.
Only if the process of the break regulation has a truly novel and non-obvious step, it should be patentable irrespective of whether the invention is computer-implemented. In this way, almost all the fuzz about computer-implemented inv
I once almosst fell victim to a phishing attack by foolisly clicking on a link in a mail. Fortunately I discovered that the url in the browser did not point to ebay.com.
I forwarded the email to ebay and got an automated response giving some advice. The advice was neither acurate nor as good as it could be.
The said they would never request a user password in an email. That is probably right, but is does not address emails linking to web sites. EBay's web site does ask for a password, and so do the bogus sites.
It also gave the advice to always open a new browser, and type http://ebay.com/ in the url field. This is not bad, but by failing to tell why, you can be sure that a large fraction of the users will not understand the importance and meaning of this advice. They will click on email links and believe they have arrived at the same site the advice would take them, and continue from there.
There is a link at the bottom of the eBay home page titled "Security center". At the bottom of that page there is a link "Deterring identity theft", and that page repeats more or less the advices in the message I got. It says prominently "Never reply to emails that ask for personal information." It says nowhere "Never click a link in an email," which would be far more appropriate. They don't mention fraudulent web sites with a word on that page.
I think eBay is doing a very lousy effort to educate their users. Likely some marketing people have told them that if they place warnings prominently on the first page, more people will get the message that eBay trading is dangerous than the occasional reports in the press about people who has been victimized. It seems to me that eBay gives zero wheight to the suffering of the victims and all importance to their profit.
Slashdot Mirror
I think the anti-software-patent movement has somehow failed to analyze the problem properly, and the MEPs are confused. While there are some real differences between the field of software and most other fields, these differences do not make it obvious that software patents is wrong.
The matter with the software field is that patents do more harm in this field, but they do largely the same kind of harm as in other fields. The arguments the movement has put forward have failed to convince the MEPs and have sent them reaching for some kind of compromise or Salomonic solution. All the thousands of explanations of why software patents is a bad idea end up being very confusing, because they normally leave the MEPs wondering why those arguments apply only to software and not to other fields.
We pretend that software is absolutely different, but all we can show are fairly relative differences.
The first true problem is that American courts have invented a test of non-obviousness that is very convenient for the courts and the patent offices. Convenient in the sense that you can determine objectively if that test has been met. Unfortunately the test does not test non-obviousness at all. In reallity it only tests a certain form of non-novelty.
The test consists in requiring the combination of two known principles to be "suggested" or "motivated" in some prior art text, before deeming this combination to be "obvious". If no such prior art text can be found, that combination is deemed "non-obvious". In this way almost everything is "non-obvious".
This obviousness test has not been tried in the Supreme Court yet, some twenty years after its introduction. However, there is a case, KSR International v. Teleflex, that may become such a trial in a relatively near future, if the Supreme Court decides to look at it.
Of course, in the name of harmonization, the European patent offices tend to adopt similar criteria; 260 million Americans cannot be wrong for twenty years.
We should formulate this problem clearly. I guess this will convince the MEPs much more easily. For some formulations and arguments, just have a look at http://www.pff.org/issues-pubs/other/ksramicus.pdf . Then we should work out some rather strong formulations of what it means to test for obviousness.
After that the rest is a lot easier. Suppose a method of desalination of seawater required you to add a certain acid three times in half-hour intervals. Suppose this method is patentable according to all the usual considerations. Does that mean that whoever patents it, patents the number three? No. Everyone agrees to that. The number three is not by itself patented. What if you use a computer to help you count to three? Is the process no longer patentable? What ought to be patened is the chemical process. Wether you use a computer to assist you in carrying out the steps is irrelevant. Only the steps themselves are relevant.
In the case of the computer-controlled breaking mechanism, the computer seems to be an essential part, because of its speed. If you were to read the velocity of the wheel's rotation on a digital display, calculate by hand the difference from the previous reading in order to determine if the wheels are about to block, you would probably relax the breaking force a little bit too late. However, the inventors of the automatic breaking system did not invent the computer. The advent of the computer made a large number of obvious applications possible, that had not been possible previously. To realize that, is not a patentable invention, it is something obvious.
Only if the process of the break regulation has a truly novel and non-obvious step, it should be patentable irrespective of whether the invention is computer-implemented. In this way, almost all the fuzz about computer-implemented inv
I forwarded the email to ebay and got an automated response giving some advice. The advice was neither acurate nor as good as it could be.
The said they would never request a user password in an email. That is probably right, but is does not address emails linking to web sites. EBay's web site does ask for a password, and so do the bogus sites.
It also gave the advice to always open a new browser, and type http://ebay.com/ in the url field. This is not bad, but by failing to tell why, you can be sure that a large fraction of the users will not understand the importance and meaning of this advice. They will click on email links and believe they have arrived at the same site the advice would take them, and continue from there.
There is a link at the bottom of the eBay home page titled "Security center". At the bottom of that page there is a link "Deterring identity theft", and that page repeats more or less the advices in the message I got. It says prominently "Never reply to emails that ask for personal information." It says nowhere "Never click a link in an email," which would be far more appropriate. They don't mention fraudulent web sites with a word on that page.
I think eBay is doing a very lousy effort to educate their users. Likely some marketing people have told them that if they place warnings prominently on the first page, more people will get the message that eBay trading is dangerous than the occasional reports in the press about people who has been victimized. It seems to me that eBay gives zero wheight to the suffering of the victims and all importance to their profit.
Regards.