New Intel microcode license disallows benchmarking
-
@admiral_p patents are already short enough. The problem is with there being too many patents around, and with patents being so generic that they don't just protect the invention, but also prevent having any competition in your field - even if those competitors already released their products (you totally can sue for patent infringement retroactively).
-
@remi The threshold for patents should be high. Lots of people seem to think any "new idea" should be patentable, and that's terrible. Patents are meant to protect research investments. They should be only for things that actually took expensive work to come up with.
-
@Gąska said in New Intel microcode license disallows benchmarking:
So, in order to stop the bad guys, you become the bad guy yourself?
Hey, I didn't say anything about morality of the system, only about its logic...
-
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
What I'm saying is, basically, that fighting fire with fire is a good strategy if you don't actually care about what is being burnt down in the process.
-
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
So, in order to stop the bad guys, you become the bad guy yourself?
Hey, I didn't say anything about morality of the system, only about its logic...
Their logic is based on morality. "They can walk around our patent with trivial technicality, and that's bad."
-
@anonymous234 I think having to prove that you actually built the thing that you are patenting would also help. At the very least, it would curb down some of the "preventive patenting" that large companies do by patenting any semblance of idea.
Being more strict about not patenting ideas and throwing out all patents about algorithms and the like (i.e. all software patents) would also help a lot. There was a fight some years ago against it, it seems it has been comprehensively lost nowadays and I don't see that it has improved things in any way, apart from feeding more lawyers.
But overall, I wonder if patents have not simply outlived their usefulness. They were very needed a couple of centuries ago, but nowadays... I'm not so sure. We need some form of IP protection but patents might not be the right one.
-
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
It has been proven time and time again that government doesn't care how much it spends. Individual people, on the other hand...
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
What I'm saying is, basically, that fighting fire with fire is a good strategy if you don't actually care about what is being burnt down in the process.
It's not fighting fire with fire, it's fighting reckless public servants with consequences of their decisions.
-
@Gąska said in New Intel microcode license disallows benchmarking:
Their logic is based on morality. "They can walk around our patent with trivial technicality, and that's bad."
You can start with a morally valid generic point and yet stray very quickly into something that is no longer moral when ignoring other moral principles.
Basically, see all garage threads for examples.
-
@remi and that's exactly why I don't think that "what about trivial workarounds" is a valid defense of the practice of filing overly generic patents.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
It has been proven time and time again that government doesn't care how much it spends. Individual people, on the other hand...
Let's back up one step: you admitted yourself that fining individuals had no chance of ever happening. What I am suggesting, however, could potentially happen. So it's probably less efficient, but at least it is realistic.
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
So now you're going to have to evaluate every rejection to see whether it is valid or not? Who will do that? Another patent officer (no vested interest here, move along)? And since it has direct financial consequences for the patent officer, you'll need to throw in an appeal process and go to a judge. For each and every rejected patent (or at least those where the applicant contests the rejection but if you're going to apply, of course you believe your patent should be accepted), and there'll be many more now than before.
Honestly, I don't think that's going to make the system any better. I prefer my solution of letting the system starve itself.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
The idea is that is you just say the tiny little thing that you think is innovative, and only get protection for this, then you might open yourself to trivial workarounds that are not protected.
So, in order to stop the bad guys, you become the bad guy yourself? The people who claim they must make the widest patent possible, are the same people who later claim they must pursue even the slightest, most bullshit infringement case, because "if we don't do that we risk losing our patents by abandonment."
Fuck lawyers.
In order to stop fat, you have to become fat yourself!
Fuck diets.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
It has been proven time and time again that government doesn't care how much it spends. Individual people, on the other hand...
It wouldn't work, because (see below)...
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
...even if the employees accepted such a deal (and they wouldn't, and they'd also have a point), they would have a strong incentive not to approve any application.
What I'm saying is, basically, that fighting fire with fire is a good strategy if you don't actually care about what is being burnt down in the process.
It's not fighting fire with fire, it's fighting reckless public servants with consequences of their decisions.
You really think the largest problem is at the patent office? There may be a problem there, but the truth is that the validity of a patent doesn't depend on the patent itself, but on the courts. Especially nowadays. How can an employee really prevent a patent being talked down, or upheld, in a court? Considering how murky things are in patent lawsuits? A feasible, but probably undesirable way (I haven't thought it through really)would be for patents to be approved only after competitors presenting their case against them, and judging based on that (make it opt-in within a certain time period, and with a streamlined process).
-
@ben_lubar said in New Intel microcode license disallows benchmarking:
Does this mean that all benchmarking and profiling tools are now illegal to use on Intel CPUs? Did Intel's lawyers even talk to their tech people at all?
The topic title isn't quite correct. You can use them - you just can't publish or otherwise share them with anyone.
-
@Gąska And I never said it was. You seem to implicate that I am defending the logic of the system through its morality, whereas I repeatedly said that I don't.
So either you've been saying "I agree with you" but with different words that aren't so clear, or you misunderstood what I said.
-
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
It has been proven time and time again that government doesn't care how much it spends. Individual people, on the other hand...
Let's back up one step: you admitted yourself that fining individuals had no chance of ever happening. What I am suggesting, however, could potentially happen. So it's probably less efficient, but at least it is realistic.
It's not just it would have less efficiency. It would have ZERO efficiency. Or, if you account for added administrative costs of processing such fines, NEGATIVE efficiency. It would fix nothing, and only add new problems.
My idea is only unrealistic because politicians are corrupted to the core, and making public servants personally responsible for their decisions makes corruption harder.
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
So now you're going to have to evaluate every rejection to see whether it is valid or not? Who will do that?
The court, like it already does. All costs being paid by the plaintiff.
if you're going to apply, of course you believe your patent should be accepted),
Or you might hope for the patent office to miss that your obviously bullshit patent is obviously bullshit. Or you might be genuinely unaware of some prior work.
Honestly, I don't think that's going to make the system any better. I prefer my solution of letting the system starve itself.
Just like investment banks in 2008 crisis and retirement pension systems in EU, it's too big to fail - you can't starve a system that the politicians will defend at all cost.
-
@Gąska said in New Intel microcode license disallows benchmarking:
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
The problem is that you can't introduce a risk to a government employee without a potential benefit. Suppose you were offered a job at your normal rate of pay. Except that you would be fined if a good faith decision you made turned out to be wrong. You would receive no bonuses for good decisions. Would you take the job?
I used to have a part-time job at a health-food bakery. The way the employer got around minimum wage laws was to pay the employees minimum wage, but penalize them for burned batches. Since most everyone burned batches occasionally, they were effectively paying less than minimum wage. I refused the condition at the job offer, and they agreed to waive the penalty for me. (I was later fired for not making the Samburgers round enough, but that's within the rules of the game.)
-
@admiral_p said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Benjamin-Hall said in New Intel microcode license disallows benchmarking:
And give them incentives to do their job. How? No clue.
Easy: give them financial penalties every time a patent they approved is declared invalid in court. Really, most problems with government administration can be solved fast and cheap by penalizing employees for making bad decisions. But no government will ever enact anything like that, for obvious reasons.
In the current context of fiscal tightness (more or less everywhere in the world, judging from what makes headlines), I think that even just having the government fine itself would help
But that completely misses the point. We need the employees to personally feel the pain of their bad decisions. Fining "the government" isn't painful to anyone specific.
And yet I'm pretty sure it would work, because "the government" (or each department of it, the patent office in this case) would suddenly have a strong incentive to improve things to avoid getting fined, and then they would find a way to improve their process.
It has been proven time and time again that government doesn't care how much it spends. Individual people, on the other hand...
It wouldn't work, because (see below)...
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
...even if the employees accepted such a deal (and they wouldn't, and they'd also have a point)
Just pay them enough. 2x average salary is enough to convince most people to do just about anything. And as long as they carry out their duties faithfully, they should be fine. (If carrying out your duties faithfully isn't enough to avoid fine, then the system is bad and not at all what I've had in mind.)
they would have a strong incentive not to approve any application.
With or without the fine for wrongful rejection we've just discussed above? Because with this second fine, they'd have very strong incentive not to reject everything.
What I'm saying is, basically, that fighting fire with fire is a good strategy if you don't actually care about what is being burnt down in the process.
It's not fighting fire with fire, it's fighting reckless public servants with consequences of their decisions.
You really think the largest problem is at the patent office?
No, the biggest problem is with having too many patents and being able to file many increasingly narrow in scope patents on a single form. The solution is to reduce number of patents and disallow filing many patents at once.
There may be a problem there, but the truth is that the validity of a patent doesn't depend on the patent itself, but on the courts. Especially nowadays.
As it should be. Though the courts should follow the exact same process as the patent clerk - and if they follow the same process, they should both arrive at the same conclusion every time.
How can an employee really prevent a patent being talked down, or upheld, in a court?
Here's the main problem with our entire civilization: lack of belief in courts handing down just verdicts. If we don't fix that, we can't fix anything that's wrong with any system in any country.
-
@remi said in New Intel microcode license disallows benchmarking:
@Gąska And I never said it was. You seem to implicate that I am defending the logic of the system through its morality, whereas I repeatedly said that I don't.
So either you've been saying "I agree with you" but with different words that aren't so clear, or you misunderstood what I said.
You seem to believe that even though you don't agree with lawyers talking like that, their argument is valid and not based in morality. I believe their argument is completely invalid (ie. doesn't justify their position even if I were to agree with them) and that it's based in morality. In other words, we agree about patents - but we don't agree about lawyers.
-
@tharpa said in New Intel microcode license disallows benchmarking:
Suppose you were offered a job at your normal rate of pay. Except that you would be fined if a good faith decision you made turned out to be wrong.
If it was good faith, then if you carried out your duties with due diligence, you shouldn't be fined.
You would receive no bonuses for good decisions. Would you take the job?
The way you presented it, of course not. But if it wasn't normal rate of pay but some very attractive base salary, why not.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@tharpa said in New Intel microcode license disallows benchmarking:
Except that you would be fined if a good faith decision you made turned out to be wrong.
If it was good faith, then if you carried out your duties with due diligence, you shouldn't be fined.
Take a look at what's happened in Baltimore in the years since the riots. The police have decided not to enforce less serious laws, because the personal legal risk isn't worth it for them.
-
@tharpa because the system is broken - good behavior (pursuing crimes) is punished and bad behavior (letting pass) is not.
-
@remi said in New Intel microcode license disallows benchmarking:
@anonymous234 said in New Intel microcode license disallows benchmarking:
Like all lawyers in all industries, they probably follow the strategy: forbid everything in the EULA, then if it ever reaches court, both sides get to find out what is actually allowed.
Sad that the legal system works like that, but it does.
I was told by a lawyer in my company that this is exactly how patents must be written. You list the "claims" i.e. things that you want to be protected by the patent, starting with the most generic possible thing and progressively adding details, in the hope that when (if?) it gets to a judge, they will obviously strike down the first claims as much too generic, but they might actually grant you one of the claims more generic than what you really wanted to protect.
What the fuck? So in patent law, you have to be worried about partial application thereof?
"oh yeah, I follow just subsection b paragraph 4,the rest of the law is meaningless."
-
@Gąska said in New Intel microcode license disallows benchmarking:
Let's back up one step: you admitted yourself that fining individuals had no chance of ever happening. What I am suggesting, however, could potentially happen. So it's probably less efficient, but at least it is realistic.
It's not just it would have less efficiency. It would have ZERO efficiency.
That's where we disagree. An administration that sees itself starved of resources because of fines would have an extremely strong incentive to do something about it.
It can whine that its budget is large enough (and since fines go back to the government, it's ultimately a zero-sum game so in theory they could just ask for an increase that covers that cost), but this is going to be very bad PR for them, and in times where every department has to justify its budget every year, it's going to be very hard (impossible) to get.
So working within its current budget (more or less), it can shrink down its other expenses to keep paying the fines, i.e. have even less people to approve patents, but either this means approving more bullshit patents (i.e. getting even more fines) or rejecting more patents (all in all, not that bad an outcome as far as I see it); or it can try to find a way to reduce the fines it has to pay, i.e. make it so less bullshit patents get through, so the same thing as above.
Or it can ask for the rules to be changed, of course, but that's always an option so it's not really worth considering in the equation here.
All in all, I don't really see how that could not be positive.
Or, if you account for added administrative costs of processing such fines, NEGATIVE efficiency. It would fix nothing, and only add new problems.
Yes, there are some additional fixed costs. But I'll point out that your idea of fining the civil servants themselves adds even more such costs, because it adds all the fixed costs of my solution, plus on top of that the cost of handling many individual parties (all civil servants) contesting fines etc.
My idea is only unrealistic because politicians are corrupted to the core, and making public servants personally responsible for their decisions makes corruption harder.
No, your idea is also unrealistic because there are (almost) no jobs where you have a basic pay which goes down with fines if you do your job badly (yes, OK, @tharpa gave an example, but that's essentially to show how shitty that is -- and possibly even illegal in some places, btw). The way employment works is that you get a guaranteed salary and on top of that various add-ons (bonus, tips etc.). I'm not saying it's the only way to have a salary system, I'm just saying that this is how the vast majority of jobs work. Your idea requires breaking something that is a fundamental understanding of how employment work.
Have the employee fined if they reject application without good reason. Another problem solved!
So now you're going to have to evaluate every rejection to see whether it is valid or not? Who will do that?
The court, like it already does. All costs being paid by the plaintiff.
Except currently the court only evaluates granted patents. Now your system means it may have to look at all patents applications. That is adding a huge cost to the system (even if paid by the plaintiff, which might also not be legal in some places unless you can prove the law suit was done in bad faith, there is still the matter of having all the judges etc. that go along, so a huge additional complexity).
Honestly, I don't think that's going to make the system any better. I prefer my solution of letting the system starve itself.
Just like investment banks in 2008 crisis and retirement pension systems in EU, it's too big to fail - you can't starve a system that the politicians will defend at all cost.
Well in both ideas we are working on the assumption that there would be enough political will to start reforming the system (and we're comparing different hypothetical ways to do it), so you can't really counter one idea by saying that there is no political will to change the system.
-
@Tsaukpaetra That's not how I understand it. Claims are nested, i.e. claim 2 is always "an apparatus as in claim 1, with the added specific detail of X" and claim 3 is "claim 2, with added detail" and so on.
When you infringe on a patent (well, when you are accused of doing so...), your defence is always going to be "I am doing the same thing as claims 1-5 describe, but these claims are too generic and therefore not valid" (either because they don't describe an actual patentable thing, or because they describe something that isn't innovative (i.e. common knowledge in the field), or some similar reason). Then you are saying "claims 6 onwards describe something that is different from what I do since I don't do those specific details" so these don't apply either. Ergo, you don't infringe.
So this is more along the lines of saying that you did not follow the rest of the law because it is unconstitutional than just because you decided to ignore it. That pure lawyer-logic, of course, but it's not really ignoring the law.
-
@remi said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
Let's back up one step: you admitted yourself that fining individuals had no chance of ever happening. What I am suggesting, however, could potentially happen. So it's probably less efficient, but at least it is realistic.
It's not just it would have less efficiency. It would have ZERO efficiency.
That's where we disagree. An administration that sees itself starved of resources because of fines would have an extremely strong incentive to do something about it.
For example, account for fines in their budgeting decisions. The money is going to end up in bucket B, so let's decrease bucket B and increase bucket A.
It can whine that its budget is large enough (and since fines go back to the government, it's ultimately a zero-sum game so in theory they could just ask for an increase that covers that cost), but this is going to be very bad PR for them, and in times where every department has to justify its budget every year, it's going to be very hard (impossible) to get.
If you think PR is going to keep the government from screwing up, we must be living on different planets. Because here on Earth, people almost universally hate their governments, and have very good reasons to hate them (last year in Poland: trade ban on Sundays, world class horse stud farm being on the verge of bankrupcy, youngest minister's advisor ever (also least qualified minister's advisor ever), retired army priest promoted to general rank, new gas taxes, new retail taxes, new bank taxes) - and yet most of them get re-elected most of the time, and when they don't, they still get hefty pensions and nice jobs at state-owned companies.
Or, if you account for added administrative costs of processing such fines, NEGATIVE efficiency. It would fix nothing, and only add new problems.
Yes, there are some additional fixed costs. But I'll point out that your idea of fining the civil servants themselves adds even more such costs, because it adds all the fixed costs of my solution, plus on top of that the cost of handling many individual parties (all civil servants) contesting fines etc.
That would mostly be covered out of applicant's pocket (they have money for patent, they should have money for court proceedings - and if they're right, the court proceedings costs would be reimbursed). Also, it's different for something that works to have added costs than for something that doesn't work to have added costs.
My idea is only unrealistic because politicians are corrupted to the core, and making public servants personally responsible for their decisions makes corruption harder.
No, your idea is also unrealistic because there are (almost) no jobs where you have a basic pay which goes down with fines if you do your job badly
Because there are very few jobs where it's extra important to not make bad decisions. There are jobs where doing your job badly can PUT YOU IN PRISON - and people are fine with that!
We could make it so the bonus is per good decision. But instead of that, we can just put all the bonuses they'd earn into their base pay.
I'm not saying it's the only way to have a salary system, I'm just saying that this is how the vast majority of jobs work. Your idea requires breaking something that is a fundamental understanding of how employment work.
What other punishment can you think of? Firing? Then you'd run out of employees even faster. Financial fines for someone whose decisions matter so much is perfectly fine in my book.
Have the employee fined if they reject application without good reason. Another problem solved!
So now you're going to have to evaluate every rejection to see whether it is valid or not? Who will do that?
The court, like it already does. All costs being paid by the plaintiff.
Except currently the court only evaluates granted patents. Now your system means it may have to look at all patents applications.
Wait, is it not possible already? Right now, if they reject your application for bullshit reason, that's it?
That is adding a huge cost to the system (even if paid by the plaintiff, which might also not be legal in some places unless you can prove the law suit was done in bad faith
Wut? Seriously that's illegal? What the fuck is wrong with your judical system!?
there is still the matter of having all the judges etc. that go along, so a huge additional complexity).
That point I don't understand. Get along with what?
Honestly, I don't think that's going to make the system any better. I prefer my solution of letting the system starve itself.
Just like investment banks in 2008 crisis and retirement pension systems in EU, it's too big to fail - you can't starve a system that the politicians will defend at all cost.
Well in both ideas we are working on the assumption that there would be enough political will to start reforming the system (and we're comparing different hypothetical ways to do it), so you can't really counter one idea by saying that there is no political will to change the system.
Okay, fair enough. So tell me, what would starve, how exactly would it starve, and what would happen when it starves? Because unlike private companies, public institutions really cannot be shut down - because their existence is literally guaranteed by law.
-
@Gąska said in New Intel microcode license disallows benchmarking:
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
And who's going to work for the patent office if they're fined for every single decision they make?
In that case, not doing any work at all is the only reasonable course of action. Let everything rot on your desk and you might get fired at some point, but at least you don't get sued personally.
-
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
And who's going to work for the patent office if they're fined for every single decision they make?
Who's going to hire an employee whose every single decision is bad?
-
@Gąska said in New Intel microcode license disallows benchmarking:
@tharpa said in New Intel microcode license disallows benchmarking:
Suppose you were offered a job at your normal rate of pay. Except that you would be fined if a good faith decision you made turned out to be wrong.
If it was good faith, then if you carried out your duties with due diligence, you shouldn't be fined.
And how exactly do you define "good faith"?
-
@Gąska said in New Intel microcode license disallows benchmarking:
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
And who's going to work for the patent office if they're fined for every single decision they make?
Who's going to hire an employee whose every single decision is bad?
The patent office, obviously.
-
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
Granted, this way might be to refuse all patents without any consideration, to avoid any risk of being fined, but hey, that wouldn't be that bad an outcome, don't you think?
Have the employee fined if they reject application without good reason. Another problem solved!
And who's going to work for the patent office if they're fined for every single decision they make?
Who's going to hire an employee whose every single decision is bad?
The patent office, obviously.
If that's their enmployees, then patents are forever damned, no matter what rules we come up with.
-
@Rhywden said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@tharpa said in New Intel microcode license disallows benchmarking:
Suppose you were offered a job at your normal rate of pay. Except that you would be fined if a good faith decision you made turned out to be wrong.
If it was good faith, then if you carried out your duties with due diligence, you shouldn't be fined.
And how exactly do you define "good faith"?
Not proven to be in bad faith.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@tharpa said in New Intel microcode license disallows benchmarking:
Suppose you were offered a job at your normal rate of pay. Except that you would be fined if a good faith decision you made turned out to be wrong.
If it was good faith, then if you carried out your duties with due diligence, you shouldn't be fined.
And how exactly do you define "good faith"?
Not proven to be in bad faith.
Great. So, basically a meaningless term.
-
@Rhywden only if "bad faith" is meaningless. And it isn't. We have been using the term "bad faith" in law for centuries, and its meaning is established pretty well.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden only if "bad faith" is meaningless. And it isn't. We have been using the term "bad faith" in law for centuries, and its meaning is established pretty well.
Yes, and we need a full court trial to determine what that term means every fucking time.
That's not a guideline fit to work under.
-
@Rhywden and yet it works everywhere it's applied.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden and yet it works everywhere it's applied.
Might that be because it's only applied to stuff it actually works on? Methinks you're falling prey to confirmation bias here.
"It works on stuff it actually can work on so that proves that it works!"
-
@Rhywden and why it can't work here? What's so special about patent clerks that they can't be expected to work in good faith?
-
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden and why it can't work here? What's so special about patent clerks that they can't be expected to work in good faith?
I'm not sure why you expect this to do wonders. You obviously grandiously misunderstand the actual problems the patent office is facing.
And now you think the magic term "in good faith" to magically solve everything.
"Hmm, the clerks are overworked, understaffed, suffer from bad rulesets and the wrong kind of incentives. How do we fix that? I know! Let's introduce a completely meaningless rule which makes everyone look over their shoulders in addition to that!"
-
@Gąska said in New Intel microcode license disallows benchmarking:
Have the employee fined if they reject application without good reason.
And who decides what is a good reason? Ultimately, that person will just be (supposedly) doing what the patent examiners are supposed to be doing now, i.e. evaluating patents and deciding whether they should be allowed. So why would they do a better job of it than patent examiners currently do?
-
@Rhywden said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden and why it can't work here? What's so special about patent clerks that they can't be expected to work in good faith?
I'm not sure why you expect this to do wonders. You obviously grandiously misunderstand the actual problems the patent office is facing.
Enlighten us then: what are the actual problems the patent office is facing?
And now you think the magic term "in good faith" to magically solve everything.
If you paid attention to the discussion, "good faith" wasn't my addition. All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected. If patents were wrongly approved/rejected for other reasons, there would be no punishment.
"Hmm, the clerks are overworked, understaffed, suffer from bad rulesets and the wrong kind of incentives. How do we fix that? I know! Let's introduce a completely meaningless rule which makes everyone look over their shoulders in addition to that!"
Not in addition - instead of. If something doesn't work, out with it.
-
@Gąska said in New Intel microcode license disallows benchmarking:
If you paid attention to the discussion, "good faith" wasn't my addition. All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected. If patents were wrongly approved/rejected for other reasons, there would be no punishment.
So the assumption is that the majority of all these bad patents were granted in bad faith, and not just by a wrong decision?
-
@Gąska said in New Intel microcode license disallows benchmarking:
@Rhywden and why it can't work here? What's so special about patent clerks that they can't be expected to work in good faith?
It would be very hard to get evidence something was in bad faith to fine them.
IMO, the patent system does more harm than good, and we should get rid of it. Maybe with a few exceptions for things that are really expensive to research, like meds. But definitely not for the stuff it is being used on IT.
-
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
If you paid attention to the discussion, "good faith" wasn't my addition. All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected. If patents were wrongly approved/rejected for other reasons, there would be no punishment.
So the assumption is that the majority of all these bad patents were granted in bad faith, and not just by a wrong decision?
That's the problem with "in bad faith" - that either translates to malice or to corruption.
With all the problems the patent office has - I'm not seeing either of those.
-
@Gąska said in New Intel microcode license disallows benchmarking:
All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected.
A certain amount of that isn't their fault; patent examiners are literally not allowed to do proper due diligence by the procedures they follow. For example, they are explicitly forbidden to use the Internet to search for examples of prior art, which ought to be SOP #1 in any sane modern patent system.
-
@masonwheeler said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected.
A certain amount of that isn't their fault; patent examiners are literally not allowed to do proper due diligence by the procedures they follow. For example, they are explicitly forbidden to use the Internet to search for examples of prior art, which ought to be SOP #1 in any sane modern patent system.
From what I remember, the system is very skewed towards punishing competence - if you're good and you're given a raise due to your experience and diligence, you now have less time to complete your work in.
Which completely nullifies any advantage your experience gives you.
-
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
If you paid attention to the discussion, "good faith" wasn't my addition. All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected. If patents were wrongly approved/rejected for other reasons, there would be no punishment.
So the assumption is that the majority of all these bad patents were granted in bad faith, and not just by a wrong decision?
The assumption is that the majority is due to some fuckup, regardless of employee's faith or ethnicity.
-
@Gąska said in New Intel microcode license disallows benchmarking:
@topspin said in New Intel microcode license disallows benchmarking:
@Gąska said in New Intel microcode license disallows benchmarking:
If you paid attention to the discussion, "good faith" wasn't my addition. All I'm proposing is punishing public servants for their fuckups, defined as not following the procedures and due diligence, resulting in bad patents being approved or good patents being rejected. If patents were wrongly approved/rejected for other reasons, there would be no punishment.
So the assumption is that the majority of all these bad patents were granted in bad faith, and not just by a wrong decision?
The assumption is that the majority is due to some fuckup
If these were just individual fuckups instead of caused by larger systematic problems, we probably wouldn't even be talking about this.
-
@remi said in New Intel microcode license disallows benchmarking:
@anonymous234 said in New Intel microcode license disallows benchmarking:
Like all lawyers in all industries, they probably follow the strategy: forbid everything in the EULA, then if it ever reaches court, both sides get to find out what is actually allowed.
Sad that the legal system works like that, but it does.
I was told by a lawyer in my company that this is exactly how patents must be written. You list the "claims" i.e. things that you want to be protected by the patent, starting with the most generic possible thing and progressively adding details, in the hope that when (if?) it gets to a judge, they will obviously strike down the first claims as much too generic, but they might actually grant you one of the claims more generic than what you really wanted to protect.
(so for example, imagining you were to invent the computer mouse, you'd start by claiming to have invented "any device to interact with another device", then "any device from the previous claim, where the other device is an electronic one" and so on, until you get to really describing how the mouse is different from a lever on a steam engine...)
It's been a while since I've read it, but IIRC, my patent covers (or covered; I think it may have expired) just about any use of buffers to store data, or at least ring buffers, at the highest level of overly broad claims.
-
@topspin there are systematic problems as well as fuckups (mostly caused by systematic problems). In Poland, government employees' recklessness is legendary, and the reason is that they never face the consequences of their decisions. With great power should come great responsibility - it's crazy that there's none.
We should create work environment where patent clerks can thoroughly analyze each and every patent. After that's done, we should require patent clerks to always thoroughly analyze every patent, under the threat of penalty. Public servants should be like robots - they should make every decision according to prewritten algorithms, and have zero freedom in this regard except collecting input information for the algorithm. It should be clearly defined what's needed to deny a patent, and patent clerk should always provide that. It should be clearly defined what steps at minimum the patent clerk must take before approving. And they should be given enough time to do their work properly.
-
@remi said in New Intel microcode license disallows benchmarking:
not that you must show that you have built the thing you patent,
With some exceptions. The one I remember off-hand is that any patent application for a perpetual motion machine must be accompanied by a working implementation. I believe there are similar requirements for a few other commonly invented crackpot devices that defy science, but I don't remember the specifics.
