The modular design of microkernels makes for easier design & debugging, and with some designs the freedom to make user space services that can only be in privileged space in monolithic designs, but does one want to pay the overhead for all that message passing?
No, which is why Apple's XNU runs in one address space for the most part (I don't even know whether there are parts which don't), and most message passing has been reduced to plain function calls. They still have the design advantages of something which is conceptually built from different subsystems with clean interfaces though.
Personally I define stealing as implying a restriction of the original posessor's ability to use the thing stolen.
You can introduce a personal definition for everything and then use conventional terms in all sorts of new situations of course. This obviously makes communication more difficult and may lead people to make false assumptions, especially in the context of law and rights, where these terms often have a very rigid meaning (for good reason, you don't want the law to be applied arbitrarily).
Using this definition, if MS copies an idea and then drives the original company out of business, they could very well be said to be stealing an idea.
No, that's called free competition in a free market (as opposed to one nailed shut with all sorts thickets of exclusion rights, or by government mandated monopolies).
All I can say is wow. Considering MS is the biggest stealer of ideas in history
And so you are doing exactly what they want: spread the fallacy that "ideas" or "thoughts" can be "stolen". Even most IPR law scholars agree that "intellectual property" is something entirely different from physical property and that you can't "steal" it.
The natural rights doctrine (I "made" it so it's all mine and mine alone) does not hold in the world of immaterial creations. It is introduced by creating artificial scarcity using laws, which should only apply in cases where they have overall positive effects.
With their "How would you feel if..." oneliners, Microsoft reaches out to the inner desire of many people to be able to get rich simply by being the first to think of something. It can however easily be reversed: "How would you feel if you worked 2 years on a computer program completely on your own and when you tried to sell it, all sorts of people would start asking money from you even though all they did was pay a patent lawyer to file some documents describing ideas they once had?"
How about Shapiro ("slightly" more recently, and "somewhat" less known than Newton):
Today, most basic and applied researchers are effectively standing on top of a huge
pyramid, not just on one set of shoulders. Of course, a pyramid can rise to far greater heights
than could any one person, especially if the foundation is strong and broad. But what happens if,
in order to scale the pyramid and place a new block on the top, a researcher must gain the
permission of each person who previously placed a block in the pyramid, perhaps paying a royalty
or tax to gain such permission? Would this system of intellectual property rights slow down the
construction of the pyramid or limit its height?
"Quartz 2D Extreme" is a term used only in a debugging utility -- it isn't something they've demoed to the public. It's not even that interesting -- it just makes some 2D operations faster. Quartz Extreme (the feature introduced in Jaguar) introduced hardware-accelerated compositing, which enabled new functionality, like Expose. "Q2DE" doesn't enable anything new.
Actually, Q2DE is a lot more interesting that QE imho (although the former requires the latter for optimum efficiency). It allows for translating pretty much the whole GUI interface drawing into OpenGL fragment shader programs. This means a lot more GUI code gets offloaded from the CPU to the GPU, freeing it to do other, useful work.
If you read the blog post of the Microsoft exec handling this, you can read:
Now, before today, Team 99 was secret. I've learned from my messups with Jim Allchin's dinner not to do secret stuff anymore. Make everything transparent. Transparency is good.
I've seen Tiger on a 1.42 Mac Mini with 512MB RAM. The only thing that worried me is that each instance of a widget in Dashboard took 5-10MB of real memory
How much of that is shared?
and about 100MB of virtual memory.
A default command line program under Panther uses 27MB of virtual memory (including an 8MB stack). A simply GUI app uses 200MB. But I think you are confused due to the connotation that "virtual memory" has under Windows and had under Mac OS classic.
Virtual memory does not mean "swapped out" memory. It simply means "allocated memory space". As long as you do not actually use this memory, it's free for other programs to use, will not cause any swapping and does not consume any RAM whatsoever.
use vmmap <pid> to get the memory map of an actual application. At the end, you'll get something like
Ghostscript doesn't support transparency in pdf's. Compare this document when viewed with Ghostscript/xpdf and when viewed Acrobat 5 or higher (or Mac OS X Preview).
The details of this action paper are by their nature sensitive and confidential, and therefore not appropriate for publication.
"Hey look, we have a secret plan to pursue harmonisation in order to protect our own interests! So don't worry, just sign those "free trade" agreements and introduce the same patent system as we have..."
Under the test in Alapatt and other cases, so long as you articulated the program in technical terms, you had patentable subject matter.
That's the way the European Patent Office works today as well. The reason is that they only require something "technical" to appear somewhere in the claims. Rocard proposes that the novel, inventive stuff should be technical, and additionally insists that the term "technical" be defined (because currently, the EPO considers things like "taking into account how a computer works" and "processing image data" as "technical").
Have a look at e.g. this. More via Google. Of course, Australia could have said "no" to it or demanded different conditions, but that's not the easiest thing to do if a 500 pound gorilla wants to have it another way. Trade policy is a very strong weapon between so-called "developed" countries.
The speed bumps aren't all that significant if they truly exist
It depends on whether they are PPC970MP's or not. Going from 2 1.8 GHz single core to 2 2.0 GHz dual core processors is quite a jump. Of course, the price will probably be adjusted accordingly (I doubt that would be another "more power for less money" type of upgrade).
Not if the USPTO does their job by enforcing the obviousness criteria and tracking down prior art, as they are required to do.
Nope, due to the sequential nature of innovation in the software field (or larger, in the field of maths and logic, applied or not), this is inherently the case.
it turns out patents are not used there to protect investments, but only for strategic purposes.
This is patently false. My previous employer had a couple patents that were specifically used to protect their investments in software research.
Let me rephrase that: it turns out patents are generally not used to protect investments, but mainly for strategic purposes. Both studies show this, and people from the industry confirm it.
if there are software patents, you have to play the game, you can't opt out
Wrong. I can use the system honestly.
I never claimed you couldn't. I simply wanted to say that if there are software patents, you have to obtain them for trading purposes, or save money for licenses and litigation, and probably both. There is no way to not get influenced by the system (unless you ignore it and get lucky enough that no one else notices you).
And, I can push for improvements to the system (as IBM is doing today) to correct abuses we've been seeing.
Sure you can. Whether the end result will be good remains a very big question mark though.
I also disagree with Shapiro, in that if you take the protections away, I will stop bringing my blocks to the pyramid (ie. keep my most innovative ideas secret-- do custom software with restrictive licenses).
That's only a problem if
Programmers would actually use the patent database to look up information (I guess this is another thing you want to fix)
The chance of independent discovery of what you develop is extremely low (as well as the chance of independent discovery of alternative equally good or better methods)
Microsoft can throw 50 programmers at reimplementing whatever I do, bundle it with Longhorn, and call it innovation, and I'm screwed. However, even the evil empire does on occasion buy IP from other innovators.
And most of the time they send a nastygram or 50 lawyers. And by the time Longhorn ships, you've had plenty of time to gain back your investments. Copyright indeed gives you a shorter exclusion term in practice (on the "idea"), but that's a feature, not a bug. It gives society the best deal.
Shapiro's argument primarily is built on the supposition that the USPTO will not or cannot enforce the obviousness criteria. I think they can.
No, Shapiro's argument is primarily built upon the fact (not supposition) that most innovation in the software sector is sequential/cumulative.
Copyright as the sole payoff is not enough. If my idea is novel enough that nobody else is likely to think of it, but is relatively easy to implement, then copyright gives me very little protection. Thus, no payoff.
IBM will always have much more patents then you and won't have any problem shaking you down as soon as you ship even one program, since your program will infringe on a lot more patents than theirs.
You yourself have stated that patents have only applied to SW for 7 years. And, we all agree that the current implementation of patents on SW is flawed. So, how could there be any positive empirical indications?
For 7 years is patents on "pure" software, software patents in the hardware industry have been around since a bit longer already (somewhere mid-eighties). And it turns out patents are not used there to protect investments, but only for strategic purposes. That has nothing to do with the quality of the granted patents.
I don't know any economists. Perhaps there are "many" that disagree with me. (How many economists do you know?)
Have a look at the study overview I pointed you to earlier on.
What I know is human nature-- people don't do things generally unless there is some kind of potential reward. This includes the kind of rewards such as a good feeling that you've helped somebody, and money. The idea that I could get a temporary monopoly on implementations of my innovative idea will spur me to publish my ideas, so others can take advantage of them. Without that, I'd just keep it to myself. Additionally, it may spur me to pursue intriguing ideas, investing time and money to see if they work out. Without the payoff, I am not interested.
And with the threat that as soon as you bring a product on the market using your great idea, there may be countless patent parasites and large companies who suddenly come asking you for protection money, you won't be very inclined to bring a product on the market either.
As a small time software guy, the chance that you'll win in the patent lottery is a lot smaller than that you'll win. And the problem is that it's not a free choice you have: if there are software patents, you have to play the game, you can't opt out. And again, this has nothing to do with patent quality, but with the fact that every computer program is built on many ideas and the fact that pretty much all innovation in software is sequential (along with the fact that large companies have more money to obtain more patents, of course).
Today, most basic and applied researchers are effectively standing on top of a huge
pyramid, not just on one set of shoulders. Of course, a pyramid can rise to far greater heights
than could any one person, especially if the foundation is strong and broad. But what happens if,
in order to scale the pyramid and place a new block on the top, a researcher must gain the
permission of each person who previously placed a block in the pyramid, perhaps paying a royalty
or tax to gain such permission? Would this system of intellectual property rights slow down the
construction of the pyramid or limit its height?
I disagree, if with some changes it can be made to work in practice. I argue for making those changes, because some (most?) of us will not seek to innovate unless there is a potential payoff.
Copyright gives you a limited exclusion right to applications of your innovations (by limiting others ability in copying your work, and forcing them to re-implement it from scratch, giving you a lead time advantage). Copyright has over the years proven to be a great stimulant for software innovation. There are no empirical indications that patents overall increase innovation in the software field, and many to the contrary.
I do argue precisely this. The baby is the idea that patenting innovations can stimulate further innovation. The bathwater is the poor implementation.
Many economists disagree with you (in fact, I don't know of a single economists which is convinced that patents are inherently good for software innovation, and that only the implementation causes problems).
Additionally, we aren't introducing a system, (in USA) we've had it for 200+ years.
Software patents were codified in US case law 7 years ago (1998, State Street decision), although there were some prior stepping stones (like Diamond vs Diehr, but that wasn't really about software patents yet).
You are pointing out many flaws in the system. I see none that cannot be corrected without tossing the entire system.
How about first implementing those corrections, check whether they actually work and only then continue expanding the system to new subject matter? There is no inherent economic or other law which states that the patent system per definition stimulates innovation in any field as long as it's "properly" applied.
One way to save their credibility would be to support the fight against software patents in Europe, as opposed to pressuring politicians to support their introduction.
IANALOAK (of any kind), but the scenario you described is an abuse of the patent/judicial system, not the scenario it is designed to protect.
The intention is pretty much irrelevant if it just doesn't work in practice.
(Don't throw out the baby with the bathwater.)
But is it really worth all the trouble? No privacy at all could do wonders for fighting crime and corruption, and if no one abused the information gathered this way for other purposes and it would never leak out, it might be really great. But we know that it won't work like that and that there will be abuses.
Nevertheless, no one argues for "not throwing out the baby with the bathwater" and to introduce such a system, while at the same encouraging people to fight for preventing abuse of such a system. Even though a particular system could have some positive effects in an ideal world, that does not mean that abandoning it in the real (and not ideal) world is a bad decision.
In fact, patents (if done right) should be a benefit to any developer. They should stimulate people to work hard at developing new ideas, so when they share them, there can be some personal reward (or at least a payback to cover the investment of time and effort).
All studies show that the driving force for innovation in the software field is competition (e.g. the FTC study). There is no need to introduce the inherently costly patent system in this sector to encourage innovation. On the contrary, the patent system has resulted in a transfer of R&D money to patenting, because software patents are virtually only used for strategic purposes: to lock out the competition (so there is less competition and there are thus less incentives to innovate).
In summation, the problem isn't with the patent concept itself, but with its implementation-- benefitting the powerful to the expense of the individual.
That is one of the properties of the patent system, and it's seen as a good trade-off in sectors where you can't do anything without a couple of millions backing you up (like in medicin, although even there people are now putting up question marks), and where innovation is mainly revolutionary/discrete as opposed to evolutionary/sequential.
It's not just implementation, there are lots of indications that the principle of the patent system is simply unsuited for a field like software. More here.
Something obviously must be done to fix this problem but I'm not sure how the proposed system would work. As soon as "volunteers" have the ability to submit prior art challenges to patent applications, you'll likely see as many or more prior art challenges than patent applications.
Not only that, but this places yet another extra burden on the economy and the public. Keep in mind that the public and the economy is supposed to profit from the patent system. His proposal is only useful if the combined costs of time and money spent on finding all this prior art by the public and the negative effects of the monopolies of the patents which do get granted, are offset by the positive effects of the patents (possible stimulation of investments in innovation).
If he is seriously suggesting that the public should voluntarily bear these extra examination costs, I hope he has some pretty good evidence that it is actually worth it (as opposed to the public just voluntarily helping IBM to get rid of a few nasty patent parasites, so they can focus better on patent farming other companies).
The reform they are asking for is generally quite unsubstantial. For example, they want to keep the money coming in from patent applications at the US patent office. That's how it goes at the European Patent Office. Consequence? You get a closed economic system trying to optimize itself. Patent Offices should get funds to achieve certain policy goals, not just to do whatever they want.
Prior art submission by third parties during examination is of course nice, though one could wonder how many third parties have the resources (time, money and people) to keep up with the deluge of patent applications that is submitted and published, and to additionally spend time on finding prior art. This is definitely an extra cost of the patent system which should be factored in when evaluating its efficiency.
The "administrative challenge" as permitted in Europe does not really help. In 2001, 5.7% of all granted patents were opposed. I can't find the link currently, but I previously read (also somewhere on the European Patent Office's website) that in about 70% of opposition cases, the patent is maintained. This means only about 1.7% of granted patents is rejected using this procedure. In 2002, the opposition rate even declined to 5.4%
Depending on how the "willful infringement" clause is reformed, it may become less dangerous to search the patent database for information. Then again, this assumes that you can actually decipher those patents to get the useful information out, of course. Most people will still find scholar.google.com more useful, probably.
"Increasing harmonization across international boundaries" probably refers to "get those software patents in the EU going asap". Not Good (tm). Not sure what it has to do with a reform of the US patent system either (unless they mean they want to get rid of software patents in the US, which I somehow doubt).
Why not simply shorten the lifespan of patents to say 2 years?/blockquote>
Even if that were The Solution(tm), it's forbidden by the TRIPs treaty, which states that if you allow patents on something, the conditions and resulting patents must be the same as for everything else.
FWIW, in Europe that would still constitute an increase from 0 to 2 years of the duration of patent monopolies on mathematics and algorithms.
Slashdot Mirror
No, which is why Apple's XNU runs in one address space for the most part (I don't even know whether there are parts which don't), and most message passing has been reduced to plain function calls. They still have the design advantages of something which is conceptually built from different subsystems with clean interfaces though.
And so you are doing exactly what they want: spread the fallacy that "ideas" or "thoughts" can be "stolen". Even most IPR law scholars agree that "intellectual property" is something entirely different from physical property and that you can't "steal" it.
The natural rights doctrine (I "made" it so it's all mine and mine alone) does not hold in the world of immaterial creations. It is introduced by creating artificial scarcity using laws, which should only apply in cases where they have overall positive effects.
With their "How would you feel if ..." oneliners, Microsoft reaches out to the inner desire of many people to be able to get rich simply by being the first to think of something. It can however easily be reversed: "How would you feel if you worked 2 years on a computer program completely on your own and when you tried to sell it, all sorts of people would start asking money from you even though all they did was pay a patent lawyer to file some documents describing ideas they once had?"
There was an interesting discussion on this topic a while ago on Apple's Darwin development list a while ago.
Virtual memory does not mean "swapped out" memory. It simply means "allocated memory space". As long as you do not actually use this memory, it's free for other programs to use, will not cause any swapping and does not consume any RAM whatsoever.
use vmmap <pid> to get the memory map of an actual application. At the end, you'll get something like
All ReadOnly portions can be shared with other applications.Ghostscript doesn't support transparency in pdf's. Compare this document when viewed with Ghostscript/xpdf and when viewed Acrobat 5 or higher (or Mac OS X Preview).
Apart from the software patents.
- Programmers would actually use the patent database to look up information (I guess this is another thing you want to fix)
- The chance of independent discovery of what you develop is extremely low (as well as the chance of independent discovery of alternative equally good or better methods)
And most of the time they send a nastygram or 50 lawyers. And by the time Longhorn ships, you've had plenty of time to gain back your investments. Copyright indeed gives you a shorter exclusion term in practice (on the "idea"), but that's a feature, not a bug. It gives society the best deal. No, Shapiro's argument is primarily built upon the fact (not supposition) that most innovation in the software sector is sequential/cumulative.As a small time software guy, the chance that you'll win in the patent lottery is a lot smaller than that you'll win. And the problem is that it's not a free choice you have: if there are software patents, you have to play the game, you can't opt out. And again, this has nothing to do with patent quality, but with the fact that every computer program is built on many ideas and the fact that pretty much all innovation in software is sequential (along with the fact that large companies have more money to obtain more patents, of course).
As Shapiro said in 2001:
One way to save their credibility would be to support the fight against software patents in Europe, as opposed to pressuring politicians to support their introduction.
Nevertheless, no one argues for "not throwing out the baby with the bathwater" and to introduce such a system, while at the same encouraging people to fight for preventing abuse of such a system. Even though a particular system could have some positive effects in an ideal world, that does not mean that abandoning it in the real (and not ideal) world is a bad decision.
All studies show that the driving force for innovation in the software field is competition (e.g. the FTC study). There is no need to introduce the inherently costly patent system in this sector to encourage innovation. On the contrary, the patent system has resulted in a transfer of R&D money to patenting, because software patents are virtually only used for strategic purposes: to lock out the competition (so there is less competition and there are thus less incentives to innovate). That is one of the properties of the patent system, and it's seen as a good trade-off in sectors where you can't do anything without a couple of millions backing you up (like in medicin, although even there people are now putting up question marks), and where innovation is mainly revolutionary/discrete as opposed to evolutionary/sequential.It's not just implementation, there are lots of indications that the principle of the patent system is simply unsuited for a field like software. More here.
If he is seriously suggesting that the public should voluntarily bear these extra examination costs, I hope he has some pretty good evidence that it is actually worth it (as opposed to the public just voluntarily helping IBM to get rid of a few nasty patent parasites, so they can focus better on patent farming other companies).
Doesn't work with slot-loading CD drives though...
You get that privilege in return for the tax from the law.
Prior art submission by third parties during examination is of course nice, though one could wonder how many third parties have the resources (time, money and people) to keep up with the deluge of patent applications that is submitted and published, and to additionally spend time on finding prior art. This is definitely an extra cost of the patent system which should be factored in when evaluating its efficiency.
The "administrative challenge" as permitted in Europe does not really help. In 2001, 5.7% of all granted patents were opposed. I can't find the link currently, but I previously read (also somewhere on the European Patent Office's website) that in about 70% of opposition cases, the patent is maintained. This means only about 1.7% of granted patents is rejected using this procedure. In 2002, the opposition rate even declined to 5.4%
Depending on how the "willful infringement" clause is reformed, it may become less dangerous to search the patent database for information. Then again, this assumes that you can actually decipher those patents to get the useful information out, of course. Most people will still find scholar.google.com more useful, probably.
"Increasing harmonization across international boundaries" probably refers to "get those software patents in the EU going asap". Not Good (tm). Not sure what it has to do with a reform of the US patent system either (unless they mean they want to get rid of software patents in the US, which I somehow doubt).