Where does the President live? "The White House" or "the Red, Orange, Yellow, Green, Blue, Indigo, and Violet House"?
The white house is accurate because "white" means "reflects all colors of visible light."
At night, do you call it the "dark gray" house? During a flaming sunset, do you call it the "red house"? No, it's still the white house.
Similarly, this veil isn't really black, because it does reflect certain visible frequencies -- these frequencies just aren't usually present in ambient light.
The next time you're pulling your hair out because your new videogame takes 20 minutes to load every level, ask yourself again if 0.176 seconds per texture "seems quick enough" (and double-check to make sure your game isn't shipping with an Erlang interpreter...)
Or hell, just buy a shiny new 3Ghz system -- how dare you make the programmer's lives more difficult by expecting their code to run decently on anything less?
Your own citation of dictionary.com on your "complete rant" disagrees with you. "The low-level software which handles the interface to peripheral hardware, schedules tasks, allocates storage, and presents a default interface to the user when no application program is running
/ The foundation software of a machine; that which schedules tasks, allocates storage, and presents a default interface to the user between applications"
Default interface to the user? Sounds like a shell to me.
So there you go -- operating system = kernel + shell. (One could argue for a bootloader as well)
Windows wasn't able to "carry out multiple tasks in parallel" (non-preemptive multitasking) until 1995, 10 years after the Amiga did it.
1) You're confusing the implementation of parallel tasks with the interface. To the user, there's no difference at all between cooperative and preemptive multitasking, except that the latter is a bit slower and the former locks up the whole system more often. They both involve parallel tasks running at the same time.
2) Even at the implementation level, pre-emptive multitasking doesn't "carry out multiple tasks in parallel" any more than cooperative does. Both types of multitasking only have one program's code running at any one time. The difference is, in pre-emptive multitasking, the operating system can jump in and cause a task switch without the program requesting one (and this usually occurs in response to a timer interrupt, so there's still only one task running at any time).
Perhaps you're thinking of multiprocessor machines (or, to a lesser extent, hyperthreading)?
Think of it in terms of cryptography. XORing *any* stream of bits with a purely random (unbiased) stream will result in a purely random unbiased stream for output, no matter how biased the input was. And XOR with random bits is a very good parallel to "the ground flipping the coin 50% of the time."
I sometimes think probability and statistics are the *most* misunderstood of all the maths.
a right is only one so long as it does not infringe upon the rights of another
which I agree with.
You also say:
Last I checked, both cannibalism and necrophilia were sort of ILLEGAL.
which is also true.
But you fail to make the logical connection here, and then spout:
I fail to see how ANYONE can think that having a website that talks about fucking and/or eating corpses is protected speech.
TALKING about necrophilia and cannibalism and COMMITTING necrophilia and cannibalism are NOT THE SAME THING. How does talking about a crime violate ANYONE's rights?
(I'm ignoring the question, for now, of whether necrophilia *really* violates someone's rights).
What if those bits are a collection of child pronography? I would say someone's rights and liberties were violated to create that content.
Yes, someone's rights were violated in creating that content. But we're talking about *distributing* it. Lots of people's rights were violated to create pictures of 9-11 occuring, but it's legal to distribute those pictures.
Distribution of that content is continued abrogation of that person's rights.
The only rights I can see that are violated by the *distribution* of child pornography are privacy rights, unless the children's faces are obscured. The violation of rights that led to the taking of the picture only happened once, and doesn't continue as the picture is distributed. So I don't see how it's "continued abrogation ["abolishment, or annulment, especially by authority"], of that person's rights."
Besides, current law doesn't only make it a crime to distribute pictures, but to simply *posess* them.
Now I could see criminalizing *selling* the pictures -- making money from a violation of someone's rights should be a crime, imho. But that's a whole 'nother bag of worms.
I agree with you about libel and "free speech" in general, though.
If this really is economic math, then economic math is bullshit that's designed to inflate numbers as much as possible with no actual reasoning behind it.
If you pay someone $20 every day with and expectation of getting $40 back every day, then on normal days you net +$20.
If someone hinders your guy from doing his job, then you net -$20.
The amount that you lost, by any reasonable definition of the term is 20 - -20 = $40, or the opportunity cost of the guy not doing his job.
It seems that what you, and previous posters, have been computing is some sort of gross losses, rather than net losses that are *due* to some cause. In other words, you're adding together opportunity cost losses + standard running costs, when the standard running costs are *not* due to the virus/mugger/whatnot.
I think the problem here is that "productive value" is too abstract for some people to work with preciecely. Let me try to modify the problem a bit to make it more clear what's going on:
Let's say that every day you buy an "employment unit" for $120, and at the end of the day you can cash that employment unit in for $150. This is economically equivalent to hiring someone and gaining productivity out of them, but it's in more concrete terms.
If I buy the $120 unit, and can sell it for $150, and someone steals it from me, how much money did I lose? I lost $150, as that was how much this unit was worth to me, regardless of how much i paid for it.
I did *not* lose $270. You're double counting the real cost when you come up with that figure.
Some more extremes to make this more clear: Let's say I bought something for $120 and could sell it for exactly $120. How much do I lose if someone steals it? $240? I think not. The thing was equivalent in value to the cash I expended for it. It might as well have been more cash, in fact. I lost $120 when it was stolen from me.
Another example: Say I give a dollar to someone to purchase 4 quarters. Then someone steals those 4 quarters from me. Did I lose $2 or $1? If you say $2, then you're out of your freaking mind. I lost a DOLLAR.
One more, as food for thought: Say I buy something for $120, and it turns out it's worth nothing. Nada. Zip. I can only give it away. I get absolutely no value or worth out of it by keeping it. People won't even pay a penny for it.
Then it gets stolen.
How much did I lose by the fact that it got stolen?
My answer: NONE. I may have lost $120 by making the dumb investment in the first place, but the fact that it got stolen changed *nothing* about my current wealth or wealth opportunities, and therefore cost me *nothing*.
----
Summary: "opportunity cost" is a really tricky subject that people throw around to inflate numbers, but it doesn't end up being logically consistent if you're not very careful about it.
The amount lost to a virus should be $cost of opportunities lost due to virus + $cost of *extra* expenditures required to fight virus (overtime, products required, outsourcing help, etc.). Your *normal* operating costs should *not* appear in this equation -- you're already counting for productivity lost in the "opportunities lost" part.
Sorry for my munged HTML. I should have previewed. Here's a better version:
You're comparing apples and oranges.
Firstly, the original hypothesis is that a more popular system will have more viruses/worms written for it, not that it will have more security holes.
Secondly, the viruses/worms in question do not even exploit any security holes to proliferate. And they wouldn't need to on linux, either. Therefore number of security holes has nothing to do with the hypothesis in question, and your whole comparison is at best a red herring, and at worst a complete non sequitur. Unfortunately, it's a very often-repeated argument that for some reason is very convincing to linux zealots.
If you really are a scientist, I worry for the quality of the research done at your lab, if this kind of thinking is common there.
I'd also like to mention that your attempts to pretty up your argument by inserting phrases such as "testing theories annoys people because it makes them question their deepest held beliefs" and "but I like reality" only weaken what you're saying when put up to a more critical eye. Do you commonly write like this when you're submitting a paper with weak supporting data, hoping that it distracts those doing peer review?
Firstly, the original hypothesis is that a more popular system will have more viruses/worms written for it, not that it will have more security holes.
Secondly, the viruses/worms in question do not even exploit any security holes to proliferate. And they wouldn't need to on linux, either. Therefore number of security holes has nothing to do with the hypothesis in question, and your whole comparison is at best a red herring, and at worst a complete non sequitur. Unfortunately, it's a very often-repeated argument that for some reason is very convincing to linux zealots.
If you really are a scientist, I worry for the quality of the research done at your lab, if this kind of thinking is common there.
Where do you get your data? The "Bush-is-good-no-matter-what economic report"? It's not even *close* to reality. Unemployment had been steadily decreasing since the early 90s, down to around 4%, and right after bush took office there was a sharp *increase* in unemployment, which has been hanging steadily around 6% since then.
I just realized that the Qt license is for commercial use, not closed source use. In this case it isn't much an ethical thing as a if-you-make-money-with-this-pay-us thing, which does make sense to me.
You won't give away the apps you write, so why complain about others?
He's not complaining about others "not giving away the apps they write".. he's complaining about others not letting him use their libraries unless his app conforms to their ethical specifications.
Even MICROSOFT lets people link to their core libraries free of cost or strings.
there's also a difference between "taking the code for software" and linking to an open source library, which is what this discussion was originally about.
-- Your comment about linux's "clone/improve" dev cycle implies that this doesn't happen at all in the commercial arena, which is a very amusing concept.
-- I don't see how linux being developed in a "cleanroom" or not is at all relevant to the SCO case. SCO is claiming that linux stole specific code practically line-by-line, not that some linux developer was tainted by NDA code. However, they haven't been able to show *any* legit examples of this as of yet. But they *are* trying to force people to pay licenses to *them* for linux, with *no proof* so far that they have any right to. This is why many people believe the case to be 110% bullshit.
And the only difference between closed and open source software in this regard is that it's harder to find when closed software steals code. Due to this, I'd think that OS software would be *stronger* on the IP front. It's all out there for everyone to see, it can't hide anything.
The point is, if you have admin on a windows machine, there's a billion ways to circumvent Ctrl-alt-del (how about changing the GINA driver, for another?), so this particular registry key isn't a big deal.
I use multitrack/midi sequencing software (Sonar) that uses it to toggle whether the window follows the cursor when you're playing back the song or not.
If you check the VNC source you'd see that the way they send "Ctrl Alt Del" is a hack -- it doesn't just send the keyboard events along. It actually switches over to the security desktop with some magic I don't quite remember.
You have a point about 3rd party programs being able to redirect to/from the security desktop, however. But VNC requires some sort of driver or hook to be installed by administrator, doesn't it? (As administrator, you can just hijack MSGINA.DLL and run the whole login with your own code anyway)
Slashdot Mirror
Therefore, the screen will appear to humans to be black, just like the other black objects in the world.
Unless the ambient light in your room happened to have energy in one of the frequencies that the screen reflected. Then it wouldn't look black.
Would you call a red ball in a room with only cyan lights a black ball, just because it looked black?
Where does the President live? "The White House" or "the Red, Orange, Yellow, Green, Blue, Indigo, and Violet House"?
The white house is accurate because "white" means "reflects all colors of visible light."
At night, do you call it the "dark gray" house? During a flaming sunset, do you call it the "red house"? No, it's still the white house.
Similarly, this veil isn't really black, because it does reflect certain visible frequencies -- these frequencies just aren't usually present in ambient light.
Chef is a scientologist?
/burn-karma
Man, I've missed too many episodes.
Hm.
The next time you're pulling your hair out because your new videogame takes 20 minutes to load every level, ask yourself again if 0.176 seconds per texture "seems quick enough" (and double-check to make sure your game isn't shipping with an Erlang interpreter...)
Or hell, just buy a shiny new 3Ghz system -- how dare you make the programmer's lives more difficult by expecting their code to run decently on anything less?
Your own citation of dictionary.com on your "complete rant" disagrees with you. "The low-level software which handles the interface to peripheral hardware, schedules tasks, allocates storage, and presents a default interface to the user when no application program is running
/ The foundation software of a machine; that which schedules tasks, allocates storage, and presents a default interface to the user between applications"
Default interface to the user? Sounds like a shell to me.
So there you go -- operating system = kernel + shell. (One could argue for a bootloader as well)
Windows wasn't able to "carry out multiple tasks in parallel" (non-preemptive multitasking) until 1995, 10 years after the Amiga did it.
1) You're confusing the implementation of parallel tasks with the interface. To the user, there's no difference at all between cooperative and preemptive multitasking, except that the latter is a bit slower and the former locks up the whole system more often. They both involve parallel tasks running at the same time.
2) Even at the implementation level, pre-emptive multitasking doesn't "carry out multiple tasks in parallel" any more than cooperative does. Both types of multitasking only have one program's code running at any one time. The difference is, in pre-emptive multitasking, the operating system can jump in and cause a task switch without the program requesting one (and this usually occurs in response to a timer interrupt, so there's still only one task running at any time).
Perhaps you're thinking of multiprocessor machines (or, to a lesser extent, hyperthreading)?
No, the parent is correct.
Think of it in terms of cryptography. XORing *any* stream of bits with a purely random (unbiased) stream will result in a purely random unbiased stream for output, no matter how biased the input was. And XOR with random bits is a very good parallel to "the ground flipping the coin 50% of the time."
I sometimes think probability and statistics are the *most* misunderstood of all the maths.
First you quote:
a right is only one so long as it does not infringe upon the rights of another
which I agree with.
You also say:
Last I checked, both cannibalism and necrophilia were sort of ILLEGAL.
which is also true.
But you fail to make the logical connection here, and then spout:
I fail to see how ANYONE can think that having a website that talks about fucking and/or eating corpses is protected speech.
TALKING about necrophilia and cannibalism and COMMITTING necrophilia and cannibalism are NOT THE SAME THING. How does talking about a crime violate ANYONE's rights?
(I'm ignoring the question, for now, of whether necrophilia *really* violates someone's rights).
What if those bits are a collection of child pronography? I would say someone's rights and liberties were violated to create that content.
Yes, someone's rights were violated in creating that content. But we're talking about *distributing* it. Lots of people's rights were violated to create pictures of 9-11 occuring, but it's legal to distribute those pictures.
Distribution of that content is continued abrogation of that person's rights.
The only rights I can see that are violated by the *distribution* of child pornography are privacy rights, unless the children's faces are obscured. The violation of rights that led to the taking of the picture only happened once, and doesn't continue as the picture is distributed. So I don't see how it's "continued abrogation ["abolishment, or annulment, especially by authority"], of that person's rights."
Besides, current law doesn't only make it a crime to distribute pictures, but to simply *posess* them.
Now I could see criminalizing *selling* the pictures -- making money from a violation of someone's rights should be a crime, imho. But that's a whole 'nother bag of worms.
I agree with you about libel and "free speech" in general, though.
But.. but... the pentium bug had nothing do to with normal floating point rounding error. Why does everyone keep making this joke?
If this really is economic math, then economic math is bullshit that's designed to inflate numbers as much as possible with no actual reasoning behind it.
If you pay someone $20 every day with and expectation of getting $40 back every day, then on normal days you net +$20.
If someone hinders your guy from doing his job, then you net -$20.
The amount that you lost, by any reasonable definition of the term is 20 - -20 = $40, or the opportunity cost of the guy not doing his job.
It seems that what you, and previous posters, have been computing is some sort of gross losses, rather than net losses that are *due* to some cause. In other words, you're adding together opportunity cost losses + standard running costs, when the standard running costs are *not* due to the virus/mugger/whatnot.
I think the problem here is that "productive value" is too abstract for some people to work with preciecely. Let me try to modify the problem a bit to make it more clear what's going on:
Let's say that every day you buy an "employment unit" for $120, and at the end of the day you can cash that employment unit in for $150. This is economically equivalent to hiring someone and gaining productivity out of them, but it's in more concrete terms.
If I buy the $120 unit, and can sell it for $150, and someone steals it from me, how much money did I lose? I lost $150, as that was how much this unit was worth to me, regardless of how much i paid for it.
I did *not* lose $270. You're double counting the real cost when you come up with that figure.
Some more extremes to make this more clear: Let's say I bought something for $120 and could sell it for exactly $120. How much do I lose if someone steals it? $240? I think not. The thing was equivalent in value to the cash I expended for it. It might as well have been more cash, in fact. I lost $120 when it was stolen from me.
Another example: Say I give a dollar to someone to purchase 4 quarters. Then someone steals those 4 quarters from me. Did I lose $2 or $1? If you say $2, then you're out of your freaking mind. I lost a DOLLAR.
One more, as food for thought: Say I buy something for $120, and it turns out it's worth nothing. Nada. Zip. I can only give it away. I get absolutely no value or worth out of it by keeping it. People won't even pay a penny for it.
Then it gets stolen.
How much did I lose by the fact that it got stolen?
My answer: NONE. I may have lost $120 by making the dumb investment in the first place, but the fact that it got stolen changed *nothing* about my current wealth or wealth opportunities, and therefore cost me *nothing*.
----
Summary: "opportunity cost" is a really tricky subject that people throw around to inflate numbers, but it doesn't end up being logically consistent if you're not very careful about it.
The amount lost to a virus should be $cost of opportunities lost due to virus + $cost of *extra* expenditures required to fight virus (overtime, products required, outsourcing help, etc.). Your *normal* operating costs should *not* appear in this equation -- you're already counting for productivity lost in the "opportunities lost" part.
Sorry for my munged HTML. I should have previewed. Here's a better version:
You're comparing apples and oranges.
Firstly, the original hypothesis is that a more popular system will have more viruses/worms written for it, not that it will have more security holes.
Secondly, the viruses/worms in question do not even exploit any security holes to proliferate. And they wouldn't need to on linux, either. Therefore number of security holes has nothing to do with the hypothesis in question, and your whole comparison is at best a red herring, and at worst a complete non sequitur. Unfortunately, it's a very often-repeated argument that for some reason is very convincing to linux zealots.
If you really are a scientist, I worry for the quality of the research done at your lab, if this kind of thinking is common there.
I'd also like to mention that your attempts to pretty up your argument by inserting phrases such as "testing theories annoys people because it makes them question their deepest held beliefs" and "but I like reality" only weaken what you're saying when put up to a more critical eye. Do you commonly write like this when you're submitting a paper with weak supporting data, hoping that it distracts those doing peer review?
You're comparing apples and oranges.
Firstly, the original hypothesis is that a more popular system will have more viruses/worms written for it, not that it will have more security holes.
Secondly, the viruses/worms in question do not even exploit any security holes to proliferate. And they wouldn't need to on linux, either. Therefore number of security holes has nothing to do with the hypothesis in question, and your whole comparison is at best a red herring, and at worst a complete non sequitur. Unfortunately, it's a very often-repeated argument that for some reason is very convincing to linux zealots.
If you really are a scientist, I worry for the quality of the research done at your lab, if this kind of thinking is common there.
unemployment rate the lowest in years, eh?
Where do you get your data? The "Bush-is-good-no-matter-what economic report"? It's not even *close* to reality. Unemployment had been steadily decreasing since the early 90s, down to around 4%, and right after bush took office there was a sharp *increase* in unemployment, which has been hanging steadily around 6% since then.
Why do people design systems with epochs like 1904? Do they expect the thing to travel backwards in time?
If I break rocks for a month, why shouldn't I have the exclusive right to break rocks? Hmmm, it doesn't sound quite as convincing that way.
You changed the sentence structure around. It should be: "If I break rocks for a month, why shouldn't I have the exclusive right to those rocks?"
And maybe you should, since they are the product of your labour (as long as you didn't steal the original rocks from someone else).
Consumer groups should know that thanks to lower frequencies that can penetrate through walls and trees
Yeah, those damn radio stations.
I just realized that the Qt license is for commercial use, not closed source use. In this case it isn't much an ethical thing as a if-you-make-money-with-this-pay-us thing, which does make sense to me.
You won't give away the apps you write, so why complain about others?
.. he's complaining about others not letting him use their libraries unless his app conforms to their ethical specifications.
He's not complaining about others "not giving away the apps they write"
Even MICROSOFT lets people link to their core libraries free of cost or strings.
there's also a difference between "taking the code for software" and linking to an open source library, which is what this discussion was originally about.
Some comments about point 1a)
-- Your comment about linux's "clone/improve" dev cycle implies that this doesn't happen at all in the commercial arena, which is a very amusing concept.
-- I don't see how linux being developed in a "cleanroom" or not is at all relevant to the SCO case. SCO is claiming that linux stole specific code practically line-by-line, not that some linux developer was tainted by NDA code. However, they haven't been able to show *any* legit examples of this as of yet. But they *are* trying to force people to pay licenses to *them* for linux, with *no proof* so far that they have any right to. This is why many people believe the case to be 110% bullshit.
And the only difference between closed and open source software in this regard is that it's harder to find when closed software steals code. Due to this, I'd think that OS software would be *stronger* on the IP front. It's all out there for everyone to see, it can't hide anything.
I've bought and read plenty of sheet music for other people's songs. Does this mean I can never write my own songs?
I've also seen lots of proprietary source code from companys I've worked at. Does this mean I can never write code again?
The point is, if you have admin on a windows machine, there's a billion ways to circumvent Ctrl-alt-del (how about changing the GINA driver, for another?), so this particular registry key isn't a big deal.
I use multitrack/midi sequencing software (Sonar) that uses it to toggle whether the window follows the cursor when you're playing back the song or not.
If you check the VNC source you'd see that the way they send "Ctrl Alt Del" is a hack -- it doesn't just send the keyboard events along. It actually switches over to the security desktop with some magic I don't quite remember.
You have a point about 3rd party programs being able to redirect to/from the security desktop, however. But VNC requires some sort of driver or hook to be installed by administrator, doesn't it? (As administrator, you can just hijack MSGINA.DLL and run the whole login with your own code anyway)