Because those games relied on latency hiding techniques that make security impossible, but the cheaters were too stupid to break a remote attestation scheme? Sad really.
Not to mention the fact that money is an idea, equitable servitudes are ideas, usufructs are ideas, loans are ideas, contracts are ideas, and, now this will really blow your mind --
options on options...
You are a fucking dunce. Each one of these is some series of promises to do something in the future. They are not just ideas like "talking shit on Slashdot". They are obligations. What is being debated is whether the government should be able to take an idea like "talking shit on slashdot" and transform it into an obligation upon every member of society not to think that idea without permission?
I have never understood this logic. I am allowed to photocopy a book as many times as I like as long as I do not distribute the resulting copies. Distribution means passing on to another legal entity. A computer is not a legal entity.
You also copy text from a book into your brain in order to read it, and in your brain it gets copied around lots of different areas, derivative works are created, etc etc, in order for you to comprehend the book. There is no distribution involved, because all areas of your brain are considered the same legal entity.
One weirdo area of copyright law is corporations. For some reason they need to buy more than one copy of a bit of software, even though they are a single legal entity. I have no idea why this is, and I can't come up with any sane theory that doesn't destroy the "corporation is a legal person" doctrine.
I hope you realise that your time argument is very very close to a refutation of the need for a creator : cause and effect is a property of the universe, so we have no reason to believe that the universe itself needs a cause. Ask a meaningless question....
If you link to Qt, you are creating a derivative work. This means you need a compatible licence with the copy of Qt you use. Every KDE app must link to Qt, and so must be GPL compatible or the developers must buy a commercial Qt licence.
Copyright law has absolutely no conception of "interfaces".
Well, all I was saying was that it was worth emphasizing that the GC won't do all the work. Your opinion was that it wasn't worth emphasizing.
I repeatedly said that MM is worth knowing about. I did say "was that point worth making", which was uncalled for. My main aim was to give reasons that your initial statements weren't entirely correct.
Agreed; I was just trying to give an example. In an application, the addListener loop might be part of some other member function in some other class buried deep in some extended class. It which case, it might not be obvious what scope is required.
I have no idea what can be said about this; if you are creating objects without any idea of how their lifetime is going to be controlled, nothing can help you.
I have to say, nothing you said there can't be done with Qt Designer. Also, QT Designer doesn't generate code. It generates a.ui XML file which can be loaded at runtime. Sorry.
Yeah, sorry, was just in a bad mood. Note that I never said you didn't need to know anything about MM... I just said that some of your statements were wrong.
So here, I'm assuming that your SomeListener.getInstance is a constructor and l is the only reference. So you have just chosen a very short lived scope to put your listeners in. Choose or create a scope that will have the lifetime that you want your listeners to have.
You really really really need to go and look up weak references. All of your examples are handled fine with weak references, and proper use of scope - this means a minimal dependence on global variable hacks. Sorry.
And if you think that removing an entry from a map is the same as *EXPLICITLY* setting some values to NULL, we may as well just call the entirety of computer science bit twiddling and declare it trivial. Please learn the meaning of the term 'abstraction'. Just because I need to know a particular machine language to implement my compiler does it mean my users do?
You did have quite the faux authoratative tone going there.
So basically, your point is that in order for garbage collection to work, there has to be garbage to collect? Was that point really worth making? Also, in any correctly implemented GC you should not have to explicitly set things to NULL'. Use weakrefs for caches, and control all other lifetimes with scope Yeah, there are very few correctly implemented GCs out there in commercial use.
Memory management is just like anything else, you need to have some knowledge of what you are doing.
There are many different ways to implement mark and sweep scheduling. Some of these approaches can yield hard realtime performance ( trading temporary memory overuse for predictable performance).
The cycle breaking in CPython - I don't know the complexity off hand. Look it up if you care, I don't think it is huge. The big cost of reference counting is increasing and decreasing the references : this must be done with some kind of coherency between multiple processors, ie a memory barrier in most cases. It also means every object going out of scope must have its reference brought into cache. This is why everyone will eventually end up using mark and sweep as this cost rises with the number of concurrent processors, and with increasing ratio of CPU speed to memory bus speed.
Mark and sweep garbage collection systems have no problems with cycles (the cycle will have no references from the roots, so will be collected). Simple reference counting schemes ( eg boost::shared_pointer) do. CPython uses a cycle breaking reference counting scheme, which doesn't suffer from cycles.
Cache entries are dealt with using weak references, ie these references do not count for the purposes of garbage collection and entries are deleted if there are no other references. In python, this is encapsulated in the weak_ref module.
Your lack of knowledge is astounding for one who would expound on these issues.... The real issue most often encountered is people who don't understand weak references (like yourself) and cache things incorrectly.
You won't get sued for implementing a reader of the open document format, and you can have input on improvements to the format rather than it being controlled by a single vendor. That is the difference.
You have actually just set the new low standard for MS-fanboyism. For slashdot, that is an achievement. Is that the smell of astroturf?
By your logic, a reason for not choosing Microsoft can never be valid, because every reason is just an excuse to "bash". Please learn the difference between copyrights and patents before attempting to equate licences based on these rights. Hint: copyright only affects distribution, patents affect use. Also, it is very strange to attempt to use the deleterious effects of Microsofts monopoly ("obscure packages with even less market presence") to justify its continuation.
In short, you are either an idiot or a paid troll.
A fact is something that has been observed repeatedly or can be repeatably inferred from evidence in the physical world received via our senses.(1)
A law is a constant conjugation of facts that have been observed under some set of conditions and under some level of accuracy.
A theory is an explanation of why a fact or law is as it is. To be a scientific theory, it must be falsifiable - it must make hypotheses which can be proven wrong. Sometimes a theory can be modified only slightly to take account of a falsified hypothesis. One falsification does not necessarily mean that *every* part of a theory is wrong.
A hypothesis is a prediction about the results of an experiment, ie a prediction about the facts that will be discovered in the experiment. A hypothesis is usually predicated on a theory.
These things are not the same, they do not "graduate" into each other according to how many people like them, or how many birthdays they have had. You are falling dangerously close to propagandising for the "evolution is just a theory" muppets.
What I hope you meant to say is that string theory is a theory which has produced very few hypotheses that can be tested by experiments performable with todays equipment and expenditure levels. Thus it has not yet been falsified, but it is not inherently unfalsifiable.
(1) This is the dodgiest part of science philosophically; see Kant, Decartes, and other brain-in-a-jar thought experiments. To do science you must make the leap of faith that your senses have some relationship to a reality outside your own consciousness, ie you must at least try to reject solipsism.
Was it really necessary to take three paragraphs to say that Americans are fat?
Anyway, this really is proposed for big cities with big traffic problems. Think about London, New York, Paris - you have to be insane to drive centrally anyway.
It is not going to work out in the sticks, and will be uneconomic for a long time for even mid sized or sprawling (low density) cities like Auckland.
But not if it was fair use... such as a public interest news story. If one of these people were running for office, they would not have a chance of burying their past publications.
I still think you are equating mathematics and arithmetic.
I only addressed your single idiotic point because that was the idiotic point you made. It would take a very long time to address every idiotic statement you might make.
As to why software should not be patentable: it makes no economic sense, it damages innovation, software is sufficiently protected by copyright, any sufficiently limited software patent regime is indistinguishable from copyright, it is economically inviable to employ sufficiently skilled and numerous patent examiners, the EPC Art 52 says that mathematics is unpatentable, literary works are unpatentable, and software "as such" is unpatentable, it has a chilling effect on education, and numerous other reasons.
The reason directly relating to its mathematical basis is that mathematics is considered to be discovered and not invented (covers "new" algorithm "invention"), and the mere use of a discovery (writing a new program) are not held to be inventive either. Of course you can disagree with this, but this is the reason why people want to make clear that software is a purely mathematical field. If you believe that maths should be patentable, I would love to know what you think shouldn't be.
Whether it ends up inside or outside the computer has no bearing on whether the entire process is mathematical : the entire system of machine, program, and input and output can be emulated on a turing machine. There is no difference between I/O and internal storage other than differences external to the machine : one is observable by an external observer, the other is not.
Look at it another way: lets say I have a sealed machine with no I/O. I then introduce an ultra-sensitive bit of equipment, which allows me to read and modify the internal state of the machine. Does the machine now have I/O? Have the processes governing the machine changed?
The point of the original mark was in the context of programming.
And yes, a mathematical operation ( made up of many smaller ones) is performed in the machine when somebody uses a program ( a sequence of numerically encoded instructions for petes sake) to draw a circle. A creative process may also be occuring in the mind of the user. Whether you believe the later is a mathematical process or not I don't know; a good book on that subject is Godel, Escher Bach by Hofstadter.
Slashdot Mirror
Because those games relied on latency hiding techniques that make security impossible, but the cheaters were too stupid to break a remote attestation scheme? Sad really.
Not to mention the fact that money is an idea, equitable servitudes are ideas, usufructs are ideas, loans are ideas, contracts are ideas, and, now this will really blow your mind --
options on options...
You are a fucking dunce. Each one of these is some series of promises to do something in the future. They are not just ideas like "talking shit on Slashdot". They are obligations. What is being debated is whether the government should be able to take an idea like "talking shit on slashdot" and transform it into an obligation upon every member of society not to think that idea without permission?
Seems to work fine.
I have never understood this logic. I am allowed to photocopy a book as many times as I like as long as I do not distribute the resulting copies. Distribution means passing on to another legal entity. A computer is not a legal entity.
You also copy text from a book into your brain in order to read it, and in your brain it gets copied around lots of different areas, derivative works are created, etc etc, in order for you to comprehend the book. There is no distribution involved, because all areas of your brain are considered the same legal entity.
One weirdo area of copyright law is corporations. For some reason they need to buy more than one copy of a bit of software, even though they are a single legal entity. I have no idea why this is, and I can't come up with any sane theory that doesn't destroy the "corporation is a legal person" doctrine.
You don't know what open source means, do you?
I think you meant to say "source available".
I hope you realise that your time argument is very very close to a refutation of the need for a creator : cause and effect is a property of the universe, so we have no reason to believe that the universe itself needs a cause. Ask a meaningless question....
This is utter shit.
If you link to Qt, you are creating a derivative work. This means you need a compatible licence with the copy of Qt you use. Every KDE app must link to Qt, and so must be GPL compatible or the developers must buy a commercial Qt licence.
Copyright law has absolutely no conception of "interfaces".
Well, all I was saying was that it was worth emphasizing that the GC won't do all the work. Your opinion was that it wasn't worth emphasizing.
I repeatedly said that MM is worth knowing about. I did say "was that point worth making", which was uncalled for. My main aim was to give reasons that your initial statements weren't entirely correct.
Agreed; I was just trying to give an example. In an application, the addListener loop might be part of some other member function in some other class buried deep in some extended class. It which case, it might not be obvious what scope is required.
I have no idea what can be said about this; if you are creating objects without any idea of how their lifetime is going to be controlled, nothing can help you.
I have to say, nothing you said there can't be done with Qt Designer. .ui XML file which can be loaded at runtime.
Also, QT Designer doesn't generate code. It generates a
Sorry.
Yeah, sorry, was just in a bad mood. Note that I never said you didn't need to know anything about MM... I just said that some of your statements were wrong.
So here, I'm assuming that your SomeListener.getInstance is a constructor and l is the only reference.
So you have just chosen a very short lived scope to put your listeners in. Choose or create a scope that will have the lifetime that you want your listeners to have.
You really really really need to go and look up weak references. All of your examples are handled fine with weak references, and proper use of scope - this means a minimal dependence on global variable hacks. Sorry.
And if you think that removing an entry from a map is the same as *EXPLICITLY* setting some values to NULL, we may as well just call the entirety of computer science bit twiddling and declare it trivial. Please learn the meaning of the term 'abstraction'. Just because I need to know a particular machine language to implement my compiler does it mean my users do?
Ah well, what can you expect from a Perl muppet.
You did have quite the faux authoratative tone going there.
So basically, your point is that in order for garbage collection to work, there has to be garbage to collect? Was that point really worth making? Also, in any correctly implemented GC you should not have to explicitly set things to NULL'. Use weakrefs for caches, and control all other lifetimes with scope Yeah, there are very few correctly implemented GCs out there in commercial use.
Memory management is just like anything else, you need to have some knowledge of what you are doing.
There are many different ways to implement mark and sweep scheduling. Some of these approaches can yield hard realtime performance ( trading temporary memory overuse for predictable performance).
The cycle breaking in CPython - I don't know the complexity off hand. Look it up if you care, I don't think it is huge. The big cost of reference counting is increasing and decreasing the references : this must be done with some kind of coherency between multiple processors, ie a memory barrier in most cases. It also means every object going out of scope must have its reference brought into cache. This is why everyone will eventually end up using mark and sweep as this cost rises with the number of concurrent processors, and with increasing ratio of CPU speed to memory bus speed.
Wow, you are a dunce.
Mark and sweep garbage collection systems have no problems with cycles (the cycle will have no references from the roots, so will be collected). Simple reference counting schemes ( eg boost::shared_pointer) do. CPython uses a cycle breaking reference counting scheme, which doesn't suffer from cycles.
Cache entries are dealt with using weak references, ie these references do not count for the purposes of garbage collection and entries are deleted if there are no other references. In python, this is encapsulated in the weak_ref module.
Your lack of knowledge is astounding for one who would expound on these issues....
The real issue most often encountered is people who don't understand weak references (like yourself) and cache things incorrectly.
You won't get sued for implementing a reader of the open document format, and you can have input on improvements to the format rather than it being controlled by a single vendor. That is the difference.
You have actually just set the new low standard for MS-fanboyism. For slashdot, that is an achievement. Is that the smell of astroturf?
By your logic, a reason for not choosing Microsoft can never be valid, because every reason is just an excuse to "bash". Please learn the difference between copyrights and patents before attempting to equate licences based on these rights. Hint: copyright only affects distribution, patents affect use. Also, it is very strange to attempt to use the deleterious effects of Microsofts monopoly ("obscure packages with even less market presence") to justify its continuation.
In short, you are either an idiot or a paid troll.
This is completely wrong.
A fact is something that has been observed repeatedly or can be repeatably inferred from evidence in the physical world received via our senses.(1)
A law is a constant conjugation of facts that have been observed under some set of conditions and under some level of accuracy.
A theory is an explanation of why a fact or law is as it is. To be a scientific theory, it must be falsifiable - it must make hypotheses which can be proven wrong. Sometimes a theory can be modified only slightly to take account of a falsified hypothesis. One falsification does not necessarily mean that *every* part of a theory is wrong.
A hypothesis is a prediction about the results of an experiment, ie a prediction about the facts that will be discovered in the experiment. A hypothesis is usually predicated on a theory.
These things are not the same, they do not "graduate" into each other according to how many people like them, or how many birthdays they have had. You are falling dangerously close to propagandising for the "evolution is just a theory" muppets.
What I hope you meant to say is that string theory is a theory which has produced very few hypotheses that can be tested by experiments performable with todays equipment and expenditure levels. Thus it has not yet been falsified, but it is not inherently unfalsifiable.
(1) This is the dodgiest part of science philosophically; see Kant, Decartes, and other brain-in-a-jar thought experiments. To do science you must make the leap of faith that your senses have some relationship to a reality outside your own consciousness, ie you must at least try to reject solipsism.
Where on earth did you get that population figure? It is insanely wrong. Maybe it is the EU + Russian Federation + former soviet states...
The EU is about 450m people.
Was it really necessary to take three paragraphs to say that Americans are fat?
Anyway, this really is proposed for big cities with big traffic problems. Think about London, New York, Paris - you have to be insane to drive centrally anyway.
It is not going to work out in the sticks, and will be uneconomic for a long time for even mid sized or sprawling (low density) cities like Auckland.
But not if it was fair use ... such as a public interest news story. If one of these people were running for office, they would not have a chance of burying their past publications.
You've never been to prison, have you?
I still think you are equating mathematics and arithmetic.
I only addressed your single idiotic point because that was the idiotic point you made. It would take a very long time to address every idiotic statement you might make.
As to why software should not be patentable: it makes no economic sense, it damages innovation, software is sufficiently protected by copyright, any sufficiently limited software patent regime is indistinguishable from copyright, it is economically inviable to employ sufficiently skilled and numerous patent examiners, the EPC Art 52 says that mathematics is unpatentable, literary works are unpatentable, and software "as such" is unpatentable, it has a chilling effect on education, and numerous other reasons.
The reason directly relating to its mathematical basis is that mathematics is considered to be discovered and not invented (covers "new" algorithm "invention"), and the mere use of a discovery (writing a new program) are not held to be inventive either. Of course you can disagree with this, but this is the reason why people want to make clear that software is a purely mathematical field. If you believe that maths should be patentable, I would love to know what you think shouldn't be.
Whether it ends up inside or outside the computer has no bearing on whether the entire process is mathematical : the entire system of machine, program, and input and output can be emulated on a turing machine. There is no difference between I/O and internal storage other than differences external to the machine : one is observable by an external observer, the other is not.
Look at it another way: lets say I have a sealed machine with no I/O. I then introduce an ultra-sensitive bit of equipment, which allows me to read and modify the internal state of the machine. Does the machine now have I/O? Have the processes governing the machine changed?
Not at all. But governments don't like those kind of statistics.
The point of the original mark was in the context of programming.
And yes, a mathematical operation ( made up of many smaller ones) is performed in the machine when somebody uses a program ( a sequence of numerically encoded instructions for petes sake) to draw a circle. A creative process may also be occuring in the mind of the user. Whether you believe the later is a mathematical process or not I don't know; a good book on that subject is Godel, Escher Bach by Hofstadter.
Explain how your knowledge of the digestive process comes into play when you are shitting. Or when you are talking shit, like now.
Whether you know how something works has no bearing on whether it does or does not work in a particular way.