Assuming a random distribution of meteors from every direction, the shadow of the earth isn't that impressive.
First, your initial assumption is begging the question, IMO; the statement that the distribution is random is equivalent to the statement that impacts will be equally distributed, therefore you're saying that impacts will be equally distributed because they are equally distributed.
You would actually expect the majority of meteor impacts in an orbital system to be caused by a fast-moving planet or moon in orbit catching up a slower-moving piece of debris in orbit. Meteor impacts, statistically speaking, tend to come overwhelmingly from the leading side of the planet, and are usually pretty close to falling parallel to the direction of orbit. In the case of moon impacts, that includes both the orbit of the Earth around the sun and the orbit of the moon around the Earth, with the caveat that the latter is largely uninteresting because the speed is so small relative to the former. Thus, there should, in theory, be slightly more impacts on the side of the moon in the direction of its travel, but the difference should be relatively tiny, probably to the point of being undetectable.
Second, it's not the shadow of the planet, but rather the shadow of the gravity well that is significant when it comes to impacts. Earth's gravity well has an effect that extends way beyond its physical surface. So basically, the question becomes one of whether its gravity well is more likely to pull a random hunk of rock into such a position that it would hit the moon or pull it into such a position that it would miss the moon. This question can be answered fairly simply: it will make it less likely to hit the side of the moon that faces Earth and more likely to hit the other side. Earth's gravity pulls random objects towards it. Therefore, if one side of the moon is always facing Earth, it is pulling objects away from that side, and pulling objects towards the opposite side.
This leaves a couple of questions:
1. Is Earth's gravity enough to make up for that disparity?
2. Are the maria disproportionately weighted towards the near side of the moon due to gravity in some way?
If the answer to both of those questions is no, then we can start to consider other curious concepts like a second moon.
That doesn't mean they don't know what to do. Most merely lack either the willpower, the motivation, the self esteem, or the time to do it. Or, in a few cases, they are so hopelessly genetically predisposed to obesity that there's not much they can do short of surgery. None of those things requires ignorance.
Does this we can't get above a certain percentage of correct predictions, or does it mean that more and more work will converge towards 100% (99.5, 99.9%, 99.999% etc. etc.)
Depending on what you mean by "work", either the latter or none of the above. The ability to detect a repeated state in a finite state machine is dependent on two things:
Number of states your storage can hold. The more complex the app, the more states you have. The bigger the set of possible states, the more states you can go through before you repeat a state. At some point, it starts to become rather prohibitive just to search through all the previous states.
Inclusiveness of each state. The more complex the app, the larger each state must be. For example, an app that zigzags back and forth through a file on disk, reading and writing, is going to be an absolute bear because ostensibly the entire file has to be considered part of each state. In practice, this can be simplified to a periodic snapshot plus a delta. Even so, this can become rather large.
I you want a really brutal example of why it's hard, consider a program that reads the last four bytes of a file, then writes them beginning two bytes from the EOF marker, appending two bytes to the end of the file every time. Technically, you never repeat a state, assuming you consider the file position to be part of your state. However, you can still be repeating every other aspect of the state while only one part is changing, and one might consider that to be an infinite loop if there is not a loop test anywhere that checks that part.
In practice, at some point, you run out of disk space, which means that any such app will eventually terminate. The question becomes one of whether you want to catch this otherwise infinite loop prior to that point, however. If you do, then you need to also detect patterns in the states (e.g. every value is now equal to every value plus k), and eventually, patterns in the patterns in the states.
So it is possible to converge towards 100% acuracy, but the total size of the state history becomes so prohibitively large in practice that it doesn't generally make sense to go much beyond a certain level of accuracy. At some point, it makes more sense to pop up a slow script dialog or whatever and ask the user to decide whether the process is really hung or not.
The state of an app is a lot more than just the state of the variables within the app.
For example, if you're reading from (and not writing to) a file, you have a file pointer that changes. If you get back to the same point in the FSM with a different file pointer value, it's not the exact same state. If the file pointer has the same value, it is. If you're writing to the file, it gets a lot more complicated rather quickly, but you get the basic idea.
But how often do you actually find infinite loops (that you KNOW are infinite loops) in software that's actually been released?
I found one in a shipping version of Xcode 4 last week. Steps to reproduce:
1. Create a project containing a library.
2. Drag the library into its own "Link Binary With Libraries" box.
3. Build.
It's not as uncommon as you think when you're dealing with apps that maintain complex, potentially cyclic (or at least heavily cross-linked) data structures under the hood.
Either it is taking in new data, in which case you will never reach an exact repeated state, or it isn't, in which case the programmer was incompetently using polling instead of a more reasonable notification mechanism, and you should buy or write better software.
Besides, for most graphical applications, a hang means an infinite loop in the main run loop. Kicking the app back out to the main run loop should almost never cause any catastrophic side effects, and will usually leave the app in a state that is clean enough to allow you to perform a "Save As" operation before quitting.
My core point was that Facebook chat, pokes, etc. are functionally no different than email and AIM, but it sounds like they are being singled out.
Also, the fact that the Facebook friendship even exists is evidence that the student considers that teacher to be a friend, not just a colleague, which means that the parents know to more carefully scrutinize how their children interact with that teacher. And that's what I mean about it being out in the open. It makes clear something that otherwise would be hidden. Thus, in effect, this law is likely to make it harder to catch predators, not easier.
Re:For a revolutionary workers party!
on
Debt Deal Reached
·
· Score: 1
One problem with abolishing business taxes is it allows corporations to be used as "money-cans" where the wealthy funnel their income into the business and then disperse it to themselves when needed and avoid the higher income taxes.
Okay, so continue business taxes for privately held businesses. Easily solved. Or don't worry about that corner case, knowing that eventually that money has to become somebody's income.
Additionally, business taxes actually subsidize employment. If a company is going to pay 25% of their profit in taxes, hiring an additional worker is effectively 25% cheaper.
Except that they pay a 15% payroll tax that almost negates that. Eliminate both, and you haven't lost much. And if you want to stimulate hiring, you just have to turn it around; instead of letting the business pay less in taxes, you give the business a chunk of change back from the taxes that their shareholders paid.
And don't forget your kids' real-life friends. One could argue that friending at least some of your kids' friends helps you know what calibre of people your kid hangs out with. There are a lot more corner cases than just your biological progeny.
The wrong side if they rule the way you think they will. The problem is that if a student and teacher are friends on Facebook, that is largely out in the open. It's the email messages, the phone calls late at night, the notes passed during school, etc. that are going to be the problem—the stuff that's under the table.
It is just plain arbitrary to single out Facebook and other social networking services while ignoring other Internet services like email and instant messaging, other telecommunications services like phone calls and texting, etc. It's like saying that students are not allowed to call teachers on Friday night because they might be arranging a date, ignoring the fact that they could call on Thursday night and trivially get around the whole problem.
No, where this sort of law fails is in the equal protection clause. One communications service should be treated the same as another, and this fails to do so by not being nearly broad enough, and a law that is broad enough would be struck down as unreasonable restraint of speech.
Either way, this law is what happens when technologically ignorant government officials see something bad happening and instead of asking, "Could we have reasonably prevented this?" or "How can we teach people to be better at spotting abusers?" instead ask, "What law can we pass so that this very narrow, specific case can never happen again (unless someone who is already going to break the law decides to break a second law)?" And that, in a single sentence, sums up everything that is wrong with criminal law today.
What this is actually telling us is that the top 1% is too large a bucket to clearly show the problem. According to Wikipedia, the top 1% is considered to be the split point between middle and upper classes. In effect, you're trying to compare whether the upper class pay more than the middle class. Yes, they do, on average. The question is whether the top part of the upper class pay more than the middle class. No, they don't, on average.
Worse, the split point isn't even necessarily upper class at $380,000, depending on whether you factor in cost of living and divide things up geographically. That's just the average nationwide split between middle and upper classes.
At $380,000, much of your income still comes from salaries and other fully taxable income. The higher up you go on that curve, the more of their income is being taxed at the much lower 15% long-term capital gains rate, and the lower their percentage of income paid in taxes becomes. You'll start to see this when you move your split point up to one or two million. And this is the disparity that people in favor of progressive taxation are complaining about.
Re:For a revolutionary workers party!
on
Debt Deal Reached
·
· Score: 1
So... If someone works his ass off (16 hour days) to save up for something he should pay more tax, in percent, than someone that works 8 hours per day? Progressive taxation may sound good but is not really fair.
If you're in the bracket of people who would still be willing to work 16 hours per day, you almost certainly won't be paying significantly more in taxes than someone working 8 hours per day with a well designed progressive tax system. Progressive taxation, if done properly, treats the long tail (which you would clearly be in at that point) at a fairly constant tax rate, using higher tax rates for only the upper parts of the curve.
Having different sales tax rates on different items isn't a very good solution because no matter what you are talking about (except real estate), the poor buy a lot more as a percentage of their income than the rich. You can live a fairly lavish lifestyle (outside of major cities) on $100,000 per year. You'd be hard pressed to spend over $250,000 per year on goods and services unless you are paying a staff (butler, maids, groundskeepers). So people making more than that typically don't pay any sales tax on money earned above that point.
By contrast, someone making $20,000 per year is likely to be spending very nearly 100% of whatever income is left over after housing expenses on taxable goods and services.
Sales tax cannot feasibly be made non-regressive. It's just the nature of the beast. Neither can business taxes. Businesses pay them and pass them on to consumers just like a sales tax. Therefore, the poor inherently pay more of it. We should, therefore, abolish business and sales taxes, then replace the capital gains tax with ordinary income tax (and adjust the rates as needed) to make up the difference. This ensures more fair taxation.
I guess it depends on whether the site license is a negotiated site license prior to the time of purchase or is merely a special package that you can buy.
If it is the former, then yes, but you could also presumably negotiate to strike those terms.
If it is the latter, then it can't be binding upon you before you use the software (or, possibly, before you open the little envelope with the media in it) because you can't be bound by something you haven't yet been presented with.
Re:For a revolutionary workers party!
on
Debt Deal Reached
·
· Score: 2
Since confiscation of all wealth for redistribution by the "workers government" removes all incentive (except the point of a Stalinist gun) for "workers" to slave for their political masters, they aren't productive.
Nobody with a clue suggests confiscating all wealth for redistribution. A proper system for redistribution of wealth should have a nominal tax rate for people within a standard deviation of the mean, and should start to slope one way or the other at that point. By two standard deviations, it should start to harshly penalize additional accumulation of wealth. By three standard deviations, it should be 0% or 100%.
The reasoning behind this is relatively simple. It is to the country's advantage to provide a fair amount of room for people to become wealthy. This provides the impetus needed to drive people to take risks, innovate, etc. However, once the distance between the poorest and the richest becomes too great, it ceases to be encouraging, and actually becomes discouraging to the people at the bottom.
Further, the greatest benefit to society comes not from making the middle class wealthy, but from bringing the poor up into the middle class. Our economic system currently behaves contrary to that goal; it does not provide enough opportunities for the poor to get out of poverty, while making it easier for the middle class to become wealthy, which creates a middle class vacuum. Such erosion of the middle class tends to destabilize societies.
Finally, there is absolutely no benefit whatsoever to allowing anyone to make more than a couple of million dollars in a single year (lottery earnings and inheritance notwithstanding). The way I look at it is this: the President of the United States makes an annual salary of $400,000 per year at the moment. No job is more important, nor is any job harder, so no job should pay more. If your job is paying more than that, then you're making more money than you deserve, pure and simple.
To that end, we would be much better off if we set up a system of taxation in which all income (earned or unearned) is treated as equal, and in which the percentage of taxes increases as income increases beyond some threshold. For example, you might make the first $100,000 be tax-free, the next $100,000 be taxed at 6.25%, the next $100,000 at 12.5%, etc. (These numbers are just pulled out of the air; the point isn't the specific numbers, but rather the concept of progressive taxation.)
Re:This is why we don't listen to your rants
on
Debt Deal Reached
·
· Score: 1
Their income tax rate is already far higher than anyone else. In fact, those 1%ers pay 38% of all federal income tax.
I have two problems with that statement:
That 38% number is meaningless out of context, as it doesn't take into account whether that 1% brings in more or less than 38% of all the taxable income.
The top 1% spend less of their income on things that are taxable. Therefore, they pay less of their income in other, non-income taxes.
As a percentage of income, the top 1% pay less of their income than anyone else in the top 10%. (Source: UCSC)
Binding arbitration with non disclosure clauses, it's built into the software they did buy and agree to. They also agreed to the audit.
Unless you signed a contract to that effect, the burden of proof is on the BSA to prove that you in fact are using the software. Unless you have installed and used the software, you have not agreed to the license. Therefore, unless you are using the software, the BSA has no right to audit you. Now, unless the apps you run have a "phone home" feature or use some other online key verification, there are only three ways for the BSA to prove that you are using the software: you can admit to using the software, you can let them come into your place of business and they can observe it, or they can file a lawsuit against you and force you to disclose it during discovery.
If you neither confirm nor deny that you are using any particular piece of software and refuse to let them in, their only option for obtaining proof that they have the right to perform the audit in the first place is to go to court, file a suit, and perform discovery. Thus, unless their evidence is fairly strong, they'll probably back down if the first thing that happens involves your lawyer telling their lawyer to fuck off.
If they do not back down, that's a sure sign that you have some serious compliance problems, and you need to get somebody in there to audit all of your systems ASAP. The folks at BSADefense.com recommend that you have an attorney conduct the audit. This places the results of the audit under attorney-client privilege, meaning that they cannot be obtained by the BSA during discovery. That seems like good advice to me.
Am I the only one who gets the urge to steal a car when I see those ads? I mean, the whole "Gone In 60 Seconds" style of that ad kind of glamorizes all sorts of bad stuff. You might even be able to legitimately argue that those ads drove you to steal someone's handbag.... Besides, if brutally ripping a purse away from some elderly lady is no worse than pirating a DVD off the Internet, then we might as well all rape and pillage. After all, it's all the same level of wrong.
Nothing sane or rational can come of that ad. Just saying.
Do computing science students these days get so close to particular languages that they actually even know what compiler flags are about?
I hope not. But they should have it drilled into their heads from day one that code has bugs, and you have to test code before you submit it to the teacher. If you decide to include that chunk of code that hasn't been tested when you turn it in to your teacher, you're risking getting a bad grade.
This is essentially what Oracle did when they enabled by default an option that hadn't been thoroughly vetted, and I maintain my original statement that this is something even a first year CS student should understand is a bad idea.
The random read performance depends on four factors, not one: the areal density, the seek/settle time, the rotational latency, and whether the data is in the drive's buffer cache already.
The modern 5400 RPM drives have higher areal density, making the actual read slightly faster and putting a lot more data in the buffer faster, faster head arm motors, making the seek/settle faster, and probably larger buffers, though to be honest, I don't remember the buffer sizes of the early 15K drives because they were out of my price range.:-)
Thus, the performance comparison is going to vary based on workload, and for many workloads, the new 5400 will beat the older 15k. For audio purposes (which is the context I'm usually describing performance in terms of), it's fairly sequential reading (of multiple mostly sequential stripes in alternation), which is why the modern 5400 RPM drives spank older 7200 RPM and 10k RPM drives.
I think you might be surprised how infrequently computers actually do highly random access to data that isn't coming out of a buffer cache these days (apart from application cold launch times, and in some OSes, boot times). There's a whole lotta cachin' goin' on (at least in Mac OS X; YMMV).
You don't need tests, you can definitely notice a quicker boot-up and snappier performance when using a 15,000 RPM drive. Those milliseconds add up fast!
Depends on the OS you're running. Mac OS X does a lot of work to make sure that booting consists largely of long sequential reads (the kext cache, etc.). If you saw a huge difference in boot times with Mac OS X (more than a couple of seconds), then you're probably seeing a sequential throughput difference rather than anything to do with seek penalties.
As an aside, I'd be curious to see boot time comparisons on an Air, just to see how much those boot caches contribute to Mac OS X booting almost twice as fast as Windows 7 on typical (spinning-drive-based) hardware.
Yeah. They were the ones I was referring to when I said that they had almost caught up, Unfortunately, I'm pushing the limits of my 500 GB drive at the moment, which means that would only give me a paltry 100 MB storage gain.
The big problem is that I take photos. Lots of photos. In RAW mode. They add up rather quickly, to the tune of ten or eleven megs apiece. On my last vacation, in a week, I shot somewhere on the order of two thousand pictures. There went fifteen or twenty gigs. Most vacations just net me one or two hundred pictures, but that's still a gig at a time. And, of course, I like to keep all my photos on my laptop.
With that in mind, I'd go through an extra hundred gigs in a matter of months, or maybe a year. Thus, if I'm going to move up to an SSD, it needs to be a lot more than a 100 gig bump. It doesn't really make sense to increase capacity by less than doubling at this point, and particularly if it costs more than a grand to do it.
I actually looked for 9.5mm terabyte drives just a few weeks ago and did not find any evidence that anybody had built one, so reading your post about the Samsung came as something of a surprise. Thanks for the tip. I was hoping somebody other than WD built one, as my previous experiences with WD drives have been acoustically unpleasant....
I still wish somebody built a 1 TB SSD drive (commercially available, as opposed to the pureSilicon hardware that seems to be vaporware), but at least SSDs have almost caught up now, so I'm hoping the Samsung I'm about to buy at Fry's tomorrow will be my last mechanical hard drive.
True, but just about every passing generation has faster seek/settle speed than the previous generation, too. At this point, that's just a very small part of the total seek time (I think), but IIRC, it used to be a much bigger part.
Slashdot Mirror
First, your initial assumption is begging the question, IMO; the statement that the distribution is random is equivalent to the statement that impacts will be equally distributed, therefore you're saying that impacts will be equally distributed because they are equally distributed.
You would actually expect the majority of meteor impacts in an orbital system to be caused by a fast-moving planet or moon in orbit catching up a slower-moving piece of debris in orbit. Meteor impacts, statistically speaking, tend to come overwhelmingly from the leading side of the planet, and are usually pretty close to falling parallel to the direction of orbit. In the case of moon impacts, that includes both the orbit of the Earth around the sun and the orbit of the moon around the Earth, with the caveat that the latter is largely uninteresting because the speed is so small relative to the former. Thus, there should, in theory, be slightly more impacts on the side of the moon in the direction of its travel, but the difference should be relatively tiny, probably to the point of being undetectable.
Second, it's not the shadow of the planet, but rather the shadow of the gravity well that is significant when it comes to impacts. Earth's gravity well has an effect that extends way beyond its physical surface. So basically, the question becomes one of whether its gravity well is more likely to pull a random hunk of rock into such a position that it would hit the moon or pull it into such a position that it would miss the moon. This question can be answered fairly simply: it will make it less likely to hit the side of the moon that faces Earth and more likely to hit the other side. Earth's gravity pulls random objects towards it. Therefore, if one side of the moon is always facing Earth, it is pulling objects away from that side, and pulling objects towards the opposite side.
This leaves a couple of questions:
1. Is Earth's gravity enough to make up for that disparity?
2. Are the maria disproportionately weighted towards the near side of the moon due to gravity in some way?
If the answer to both of those questions is no, then we can start to consider other curious concepts like a second moon.
That doesn't mean they don't know what to do. Most merely lack either the willpower, the motivation, the self esteem, or the time to do it. Or, in a few cases, they are so hopelessly genetically predisposed to obesity that there's not much they can do short of surgery. None of those things requires ignorance.
Or, more likely, evidence that the piece of land was previously above water at one time in Earth's past and got struck by a large meteorite.
Oh, and some random hunks of rock fell on top of it since then.
Depending on what you mean by "work", either the latter or none of the above. The ability to detect a repeated state in a finite state machine is dependent on two things:
I you want a really brutal example of why it's hard, consider a program that reads the last four bytes of a file, then writes them beginning two bytes from the EOF marker, appending two bytes to the end of the file every time. Technically, you never repeat a state, assuming you consider the file position to be part of your state. However, you can still be repeating every other aspect of the state while only one part is changing, and one might consider that to be an infinite loop if there is not a loop test anywhere that checks that part.
In practice, at some point, you run out of disk space, which means that any such app will eventually terminate. The question becomes one of whether you want to catch this otherwise infinite loop prior to that point, however. If you do, then you need to also detect patterns in the states (e.g. every value is now equal to every value plus k), and eventually, patterns in the patterns in the states.
So it is possible to converge towards 100% acuracy, but the total size of the state history becomes so prohibitively large in practice that it doesn't generally make sense to go much beyond a certain level of accuracy. At some point, it makes more sense to pop up a slow script dialog or whatever and ask the user to decide whether the process is really hung or not.
The state of an app is a lot more than just the state of the variables within the app.
For example, if you're reading from (and not writing to) a file, you have a file pointer that changes. If you get back to the same point in the FSM with a different file pointer value, it's not the exact same state. If the file pointer has the same value, it is. If you're writing to the file, it gets a lot more complicated rather quickly, but you get the basic idea.
I found one in a shipping version of Xcode 4 last week. Steps to reproduce:
1. Create a project containing a library.
2. Drag the library into its own "Link Binary With Libraries" box.
3. Build.
It's not as uncommon as you think when you're dealing with apps that maintain complex, potentially cyclic (or at least heavily cross-linked) data structures under the hood.
Either it is taking in new data, in which case you will never reach an exact repeated state, or it isn't, in which case the programmer was incompetently using polling instead of a more reasonable notification mechanism, and you should buy or write better software.
Besides, for most graphical applications, a hang means an infinite loop in the main run loop. Kicking the app back out to the main run loop should almost never cause any catastrophic side effects, and will usually leave the app in a state that is clean enough to allow you to perform a "Save As" operation before quitting.
My core point was that Facebook chat, pokes, etc. are functionally no different than email and AIM, but it sounds like they are being singled out.
Also, the fact that the Facebook friendship even exists is evidence that the student considers that teacher to be a friend, not just a colleague, which means that the parents know to more carefully scrutinize how their children interact with that teacher. And that's what I mean about it being out in the open. It makes clear something that otherwise would be hidden. Thus, in effect, this law is likely to make it harder to catch predators, not easier.
Okay, so continue business taxes for privately held businesses. Easily solved. Or don't worry about that corner case, knowing that eventually that money has to become somebody's income.
Except that they pay a 15% payroll tax that almost negates that. Eliminate both, and you haven't lost much. And if you want to stimulate hiring, you just have to turn it around; instead of letting the business pay less in taxes, you give the business a chunk of change back from the taxes that their shareholders paid.
And don't forget your kids' real-life friends. One could argue that friending at least some of your kids' friends helps you know what calibre of people your kid hangs out with. There are a lot more corner cases than just your biological progeny.
The wrong side if they rule the way you think they will. The problem is that if a student and teacher are friends on Facebook, that is largely out in the open. It's the email messages, the phone calls late at night, the notes passed during school, etc. that are going to be the problem—the stuff that's under the table.
It is just plain arbitrary to single out Facebook and other social networking services while ignoring other Internet services like email and instant messaging, other telecommunications services like phone calls and texting, etc. It's like saying that students are not allowed to call teachers on Friday night because they might be arranging a date, ignoring the fact that they could call on Thursday night and trivially get around the whole problem.
No, where this sort of law fails is in the equal protection clause. One communications service should be treated the same as another, and this fails to do so by not being nearly broad enough, and a law that is broad enough would be struck down as unreasonable restraint of speech.
Either way, this law is what happens when technologically ignorant government officials see something bad happening and instead of asking, "Could we have reasonably prevented this?" or "How can we teach people to be better at spotting abusers?" instead ask, "What law can we pass so that this very narrow, specific case can never happen again (unless someone who is already going to break the law decides to break a second law)?" And that, in a single sentence, sums up everything that is wrong with criminal law today.
What this is actually telling us is that the top 1% is too large a bucket to clearly show the problem. According to Wikipedia, the top 1% is considered to be the split point between middle and upper classes. In effect, you're trying to compare whether the upper class pay more than the middle class. Yes, they do, on average. The question is whether the top part of the upper class pay more than the middle class. No, they don't, on average.
Worse, the split point isn't even necessarily upper class at $380,000, depending on whether you factor in cost of living and divide things up geographically. That's just the average nationwide split between middle and upper classes.
At $380,000, much of your income still comes from salaries and other fully taxable income. The higher up you go on that curve, the more of their income is being taxed at the much lower 15% long-term capital gains rate, and the lower their percentage of income paid in taxes becomes. You'll start to see this when you move your split point up to one or two million. And this is the disparity that people in favor of progressive taxation are complaining about.
If you're in the bracket of people who would still be willing to work 16 hours per day, you almost certainly won't be paying significantly more in taxes than someone working 8 hours per day with a well designed progressive tax system. Progressive taxation, if done properly, treats the long tail (which you would clearly be in at that point) at a fairly constant tax rate, using higher tax rates for only the upper parts of the curve.
Having different sales tax rates on different items isn't a very good solution because no matter what you are talking about (except real estate), the poor buy a lot more as a percentage of their income than the rich. You can live a fairly lavish lifestyle (outside of major cities) on $100,000 per year. You'd be hard pressed to spend over $250,000 per year on goods and services unless you are paying a staff (butler, maids, groundskeepers). So people making more than that typically don't pay any sales tax on money earned above that point.
By contrast, someone making $20,000 per year is likely to be spending very nearly 100% of whatever income is left over after housing expenses on taxable goods and services.
Sales tax cannot feasibly be made non-regressive. It's just the nature of the beast. Neither can business taxes. Businesses pay them and pass them on to consumers just like a sales tax. Therefore, the poor inherently pay more of it. We should, therefore, abolish business and sales taxes, then replace the capital gains tax with ordinary income tax (and adjust the rates as needed) to make up the difference. This ensures more fair taxation.
I guess it depends on whether the site license is a negotiated site license prior to the time of purchase or is merely a special package that you can buy.
If it is the former, then yes, but you could also presumably negotiate to strike those terms.
If it is the latter, then it can't be binding upon you before you use the software (or, possibly, before you open the little envelope with the media in it) because you can't be bound by something you haven't yet been presented with.
Nobody with a clue suggests confiscating all wealth for redistribution. A proper system for redistribution of wealth should have a nominal tax rate for people within a standard deviation of the mean, and should start to slope one way or the other at that point. By two standard deviations, it should start to harshly penalize additional accumulation of wealth. By three standard deviations, it should be 0% or 100%.
The reasoning behind this is relatively simple. It is to the country's advantage to provide a fair amount of room for people to become wealthy. This provides the impetus needed to drive people to take risks, innovate, etc. However, once the distance between the poorest and the richest becomes too great, it ceases to be encouraging, and actually becomes discouraging to the people at the bottom.
Further, the greatest benefit to society comes not from making the middle class wealthy, but from bringing the poor up into the middle class. Our economic system currently behaves contrary to that goal; it does not provide enough opportunities for the poor to get out of poverty, while making it easier for the middle class to become wealthy, which creates a middle class vacuum. Such erosion of the middle class tends to destabilize societies.
Finally, there is absolutely no benefit whatsoever to allowing anyone to make more than a couple of million dollars in a single year (lottery earnings and inheritance notwithstanding). The way I look at it is this: the President of the United States makes an annual salary of $400,000 per year at the moment. No job is more important, nor is any job harder, so no job should pay more. If your job is paying more than that, then you're making more money than you deserve, pure and simple.
To that end, we would be much better off if we set up a system of taxation in which all income (earned or unearned) is treated as equal, and in which the percentage of taxes increases as income increases beyond some threshold. For example, you might make the first $100,000 be tax-free, the next $100,000 be taxed at 6.25%, the next $100,000 at 12.5%, etc. (These numbers are just pulled out of the air; the point isn't the specific numbers, but rather the concept of progressive taxation.)
I have two problems with that statement:
As a percentage of income, the top 1% pay less of their income than anyone else in the top 10%. (Source: UCSC)
Unless you signed a contract to that effect, the burden of proof is on the BSA to prove that you in fact are using the software. Unless you have installed and used the software, you have not agreed to the license. Therefore, unless you are using the software, the BSA has no right to audit you. Now, unless the apps you run have a "phone home" feature or use some other online key verification, there are only three ways for the BSA to prove that you are using the software: you can admit to using the software, you can let them come into your place of business and they can observe it, or they can file a lawsuit against you and force you to disclose it during discovery.
If you neither confirm nor deny that you are using any particular piece of software and refuse to let them in, their only option for obtaining proof that they have the right to perform the audit in the first place is to go to court, file a suit, and perform discovery. Thus, unless their evidence is fairly strong, they'll probably back down if the first thing that happens involves your lawyer telling their lawyer to fuck off.
If they do not back down, that's a sure sign that you have some serious compliance problems, and you need to get somebody in there to audit all of your systems ASAP. The folks at BSADefense.com recommend that you have an attorney conduct the audit. This places the results of the audit under attorney-client privilege, meaning that they cannot be obtained by the BSA during discovery. That seems like good advice to me.
As always, the usual caveats apply. IANALBIPOOSD.
Am I the only one who gets the urge to steal a car when I see those ads? I mean, the whole "Gone In 60 Seconds" style of that ad kind of glamorizes all sorts of bad stuff. You might even be able to legitimately argue that those ads drove you to steal someone's handbag.... Besides, if brutally ripping a purse away from some elderly lady is no worse than pirating a DVD off the Internet, then we might as well all rape and pillage. After all, it's all the same level of wrong.
Nothing sane or rational can come of that ad. Just saying.
I hope not. But they should have it drilled into their heads from day one that code has bugs, and you have to test code before you submit it to the teacher. If you decide to include that chunk of code that hasn't been tested when you turn it in to your teacher, you're risking getting a bad grade.
This is essentially what Oracle did when they enabled by default an option that hadn't been thoroughly vetted, and I maintain my original statement that this is something even a first year CS student should understand is a bad idea.
The random read performance depends on four factors, not one: the areal density, the seek/settle time, the rotational latency, and whether the data is in the drive's buffer cache already.
The modern 5400 RPM drives have higher areal density, making the actual read slightly faster and putting a lot more data in the buffer faster, faster head arm motors, making the seek/settle faster, and probably larger buffers, though to be honest, I don't remember the buffer sizes of the early 15K drives because they were out of my price range. :-)
Thus, the performance comparison is going to vary based on workload, and for many workloads, the new 5400 will beat the older 15k. For audio purposes (which is the context I'm usually describing performance in terms of), it's fairly sequential reading (of multiple mostly sequential stripes in alternation), which is why the modern 5400 RPM drives spank older 7200 RPM and 10k RPM drives.
I think you might be surprised how infrequently computers actually do highly random access to data that isn't coming out of a buffer cache these days (apart from application cold launch times, and in some OSes, boot times). There's a whole lotta cachin' goin' on (at least in Mac OS X; YMMV).
Depends on the OS you're running. Mac OS X does a lot of work to make sure that booting consists largely of long sequential reads (the kext cache, etc.). If you saw a huge difference in boot times with Mac OS X (more than a couple of seconds), then you're probably seeing a sequential throughput difference rather than anything to do with seek penalties.
As an aside, I'd be curious to see boot time comparisons on an Air, just to see how much those boot caches contribute to Mac OS X booting almost twice as fast as Windows 7 on typical (spinning-drive-based) hardware.
Yeah. They were the ones I was referring to when I said that they had almost caught up, Unfortunately, I'm pushing the limits of my 500 GB drive at the moment, which means that would only give me a paltry 100 MB storage gain.
The big problem is that I take photos. Lots of photos. In RAW mode. They add up rather quickly, to the tune of ten or eleven megs apiece. On my last vacation, in a week, I shot somewhere on the order of two thousand pictures. There went fifteen or twenty gigs. Most vacations just net me one or two hundred pictures, but that's still a gig at a time. And, of course, I like to keep all my photos on my laptop.
With that in mind, I'd go through an extra hundred gigs in a matter of months, or maybe a year. Thus, if I'm going to move up to an SSD, it needs to be a lot more than a 100 gig bump. It doesn't really make sense to increase capacity by less than doubling at this point, and particularly if it costs more than a grand to do it.
I actually looked for 9.5mm terabyte drives just a few weeks ago and did not find any evidence that anybody had built one, so reading your post about the Samsung came as something of a surprise. Thanks for the tip. I was hoping somebody other than WD built one, as my previous experiences with WD drives have been acoustically unpleasant....
I still wish somebody built a 1 TB SSD drive (commercially available, as opposed to the pureSilicon hardware that seems to be vaporware), but at least SSDs have almost caught up now, so I'm hoping the Samsung I'm about to buy at Fry's tomorrow will be my last mechanical hard drive.
In that case, there's nothing at all surprising about this drive. :-)
True, but just about every passing generation has faster seek/settle speed than the previous generation, too. At this point, that's just a very small part of the total seek time (I think), but IIRC, it used to be a much bigger part.