Does anyone have any data whatsoever showing that with downloading on the internet, the small artists are being hurt? I'm sure there's plenty of data showing that the big guys are being hurt by online piracy, but that's really not a huge deal. The big artists will continue making music for only millions of dollars instead of tens of millions. Besides, even amongst the big artists, the big money tends to be in their concerts. As far as I can tell, the small artists have been selling just as many albums as before.
The one big thing that is currently hurting small artists (far more than illegal downloading) is the cost of gas. Now that they're paying $4/gallon, it's a lot more expensive to tour and promote your music. Breaking even, or making a profit on a tour was never an assured thing for small bands, and it's a lot harder to do today.
Just because Linux can be the best tool for a job doesn't mean it always is. If you can setup a windows MCE box in 2 hours, and it would take 10 hours to setup and tweak a linux box to do the same thing, then Linux might not be the better choice. You can make all the claims you want about it being able to do the same things as windows (not entirely true though), and it being free, but you've managed to completely dismiss the time element.
Additionally, a Windows MCE box (at the moment) has advantages over a myth box. For one, you don't need to pay to get listings. Those are automatic. Additionally, if you want to view Netflix on demand movies, you can install plugins on a Windows MCE box to let you browse and watch them from your TV. Also, as the parent said, the video codecs for playing back h.264 video on windows are better than those available on Linux (to the extent that many machines can watch HD h.264 video in Windows but not under Linux). Not all of these problems are insurmountable, and they may be fixed in the future. However, many people want their machine to do what they want RIGHT NOW, and not have to wait months or years to get things up to speed.
Both windows and linux have their strengths and weaknesses. For many tasks linux is better, and for some windows is. If I want a fileserver, webserver, router, etc, I'm going to use Linux. However, if I want to set up a media center box, I've found windows (gasp) Vista MCE to do the job perfectly fine. The interface is decent (not perfect, and not as customizable as myth apparently), it plays back the media I want (although I did have to install a few plugins to get everything I wanted), and for the most part, I don't have to actually use the keyboard/mouse. I can do most everything I need to do using the remote that comes with it.
Calling Bush "far right" is a huge laugh. One of the reasons Bush is so widely disliked across the spectrum is because he is not strongly conservative, and not strongly liberal. He is pushing a huge war in the middle east that is the exact opposite of one of his early campaign promises (to keep US troops at home and stay away from peace keeping missions). A "far right" president may have supported the war in Iraq, deposing Saddam, and throwing in a dictator friendly to the US governments interests. Sticking it out and wasting billions of dollars on bringing democracy to the middle east is not a "far right" idea, that's the idea of peace keeping which Bill Clinton was always big on. It's a feel good idea to help make people feel better for fighting a war.
Bush has also not pushed for any significant conservative legislation. His education "reforms" are very liberal, and gave the federal government more control over education. His views on immigration are the same as the democrats (open borders and amnesty). The only issues that people can point at him and say he is conservative are social issues. Bush is a strong Christian, which is an automatically bad thing. He is anti-abortion, which I can understand why people dislike that. However after the past 8 years, have any laws really reflected this? Some government money might now go to church-based charities, instead of only to non-denominational ones. Aside from that, has anything really changed on this front? Social issues like this are always big talking points before an election, but at the end of the day politicians rarely enact real and meaningful change in regard to them.
Bush's stance on the war on terror tends to be "right wing", but the libertarian fringe of the republican party (who are considered further right) tend to be against these movements. However the actions of the patriot act are things that both Republicans and Democrats have pushed for at one time or another. Both tend to seek increasing the governments power (although Republicans were, once upon a time the small government party), however they only tend to want to increase the power when they're in charge, hence the democrats objections to the patriot act while the republicans are in charge.
I would consider many Democratic and Republican actions to be "on the left". Both tend to favor big government, the distinction being how big. One wants socialized health care, the other wants this hideous bureaucratic and inefficient mess that they've created as "compromises". One wants the government to fully pay for peoples retirement, the other wants the people to have some say in how the government pays for their retirement. For most issues, neither party is looking to smaller government, but thankfully neither party is looking fully into socialism either.
This isn't that much of an issue really. The public has never been that well-informed, and even when they are people will claim they aren't for their own political purposes. The current solution that seems to be in use is to end democracy and the republic system that we have. It's what the well-meaning people of the USA have been pushing for years. Of course, at the same time they claim to want to make things more democratic. It seems like the two goals are contradictory, but they seem to work quite well together. Some well-meaning (often liberal, but not always) leader has an idea that he knows the public won't go for yet. The idea is well regarded by the intelligentsia, and they want it made into law.
They have two options to do this. One is to convince judges to make it law. However, this is hard to do and can be unreliable (but with the right judges it is much easier to do). The other is to use grass roots campaigns that won't stop fighting. If you have an idea that only 20% of the population likes, but they like it enough (particularly on the local/state level), they can push it through. They do this by continuously revoting on the issue. The idea being that if the first vote fails, get another one, and another one and another one . . . If they're lucky they can get the vote to occur at an election that most people don't care about (some local only election with traditionally low voter turnout). Eventually they can get it to pass, and now their plan is passed "democratically" through a referendum. The opposition is often not organized enough to stop it (most people don't support it, but aren't passionately opposed).
I always find it amusing how direct-voter referendums tend to be the LEAST democratic method of determining laws because of things like this. I'm sure people all over have seen the same sort of ideas steam-rolled through. It helps the intelligentsia pass their ideas without having an informed public, and still maintaining the idea that the people have their say.
I think you're missing the point that even without a welfare system, there are plenty of people who would take the job at the prevailing wage. Teenagers for instance.
You act like there is some God given right for people to earn good wages. The fact of the matter is there isn't. In all likelihood, if given the choice between paying the few people who need the job a living wage, or firing them, the companies would likely choose to fire the workers. The work they do just isn't worth the amount requested. As there exists a social welfare system, what is better from the government's perspective. Someone who needs help to make ends meet as their job doesn't quite pay enough, or someone who earns no money because they don't have enough skills to get any job.
Economic forces are so much more complicated than "if we pay someone more things would be better". But keep living in your little world. I'm sure some day economic forces and human nature will stop working, and be the way you want them.
If your job can be done by a 15 year old kid who has never had a job in his life, isn't dangerous, is easy to do, and you can find a replacement worker in a matter of days when someone quits, it really isn't a good job. It is not hard to get a better job than working in fastfood. Hell, if you show up on time for a month at one of those places people will think you're the best worker ever. But they don't really need the worlds best workers, and in fact don't want good workers. They'd much rather have cheap workers. They're easier to replace, and don't cost much.
I would venture to say that most people who are working fast food (and aren't at least managers) that seriously need the job for their income have something wrong with them that prevents them from getting a real job. Maybe it's something as simple as they have a couple kids, and have to regularly take time off for them, and so they're considered unreliable at other places. Sometimes they might be mentally challenged, or just really slow/lazy. Maybe they're just irresponsible adults who don't want a real job. Despite the economy being "bad" right now, with a high school education, and a hard work attitude (which means being willing to accept idiots in authority for a while when you're being paid shit), you are able to advance. You can get to a position in less than a year where you're making more than a living wage for yourself. You might not be living the high life, but it is not that difficult for a single person to live on less than $20,000/year (of course, if there are health considerations to be taken into account this changes, as the state of health care is a disaster).
This is spoken from someone working on a macbook, so don't mark me down as an anti-apple zealot. OSX eats up memory. 2GB is the minimum I'd want on a leopard machine. If you plan on using parallels at all, 3GB is likely not enough. Of course, I tend to always have tons of tabs open in safari, I run mail, a terminal, xchat, adium, textedit, itunes,and other stuff at the same time.
At the end of the day, RAM is dirt cheap. I can buy a 2GB stick for about $40. If my OS eats memory it really isn't a big deal. The real problem is that my machine can only accept so much RAM.
I think they have that feature already. There are a lot of FOSS people (at least on slashdot) who are also gamers. The ability to play almost every game in windows (combined with the fact that their video card drivers are generally more mature and offer slightly better performance) means that many FOSS people HAVE to have microsoft windows on their computers. Should microsoft care that they also have linux installed and use that sometimes, maybe even predominately? As long as these users have still purchased a legitimate copy of MS Windows, Microsoft is making their profits, regardless of how much their product is used on the machine.
Of course, if the gamers are pirating windows, then that's a different story, and they have no right to complain about Microsoft's ethics.
the number of cups varies, but normally there are 15 cups. 5 in the back row forming the triangle, and of course 2 water cups on each side. The cups should be relatively small (none of the giant 16oz cups). The game should also be played 3v3.
Of course, many different variants exist, but this is the common way I know.
Any time a country is in an economic slump, the best way to relieve it is by instituting widespread public works projects. Not only do they create short term wealth and job opportunities, but they have sustained maintenance work as well as the overall betterment of society through the finalization of said public work.
Whoa there, this statement IS NOT a fact. Public works projects can help a slumping economy, but only if the public works project is needed, and absolutely helps expand the economy. There is more to it than that, but creating jobs does not necessarily expand the economy but can result in simple wealth redistribution. For example, if the government hired 10,000 people to dig a giant ditch, and than hired another 10,000 people to fill in the ditch, jobs would be created, but would it help the economy? The government doesn't magically have money, they need to obtain it somewhere. In this instance they've created 20,000 jobs, but added nothing to the economy. In fact, under such a situation, they've likely decreased the economy. Even if unemployment is really high, some of these people are likely not doing other (productive) jobs to dig a ditch and fill it in instead. This decreases the net value of the economy. Additionally, where is the money to pay these workers coming from? They either tax the people (reducing the money they have to create new jobs, and buy goods, decreasing the size of the economy) or print money, causing inflation, resulting in an inflation tax instead.
Of course, the real world is much more difficult, and I am not an economist, but I know not all economists believe that public works projects are good for the economy. The publics works projects in the great depression did not cure the depression, however government military spending did help bring us out of the depression (although, I imagine the average standard of living decreased during the war years, as the money was going into the war). One factor of public works projects that can also helps the economy (beyond the help the public works project itself does) in the long term, is the training that workers might receive working on the project, making them more productive afterward.
What I do know about pushing people into public works projects on renewable resources is that it would create jobs, and result in more renewable energy. However, if the cost of the energy is greater, than everyone is paying in higher overall costs (or taxes). It must also be noted that in a slumping economy, the costs of implementing large public works projects is cheaper, as there are often large numbers of unemployed people (who in the US are often earning money from the government already from the welfare system). This means the net cost of implementing these projects is cheaper due to being able to pay lower wages, and even cheaper still because you don't have to pay these people welfare benefits.
Maybe a real economist could plug through the numbers and predict if your proposed projects would help the economy (even than they'd be guessing). However, claiming it's a fact that public works projects help the economy is definitely not true.
this is partially true. It's true that the main load power stations always produce the same amount of power. Coal, hydro, nuclear can't change dynamically to use a smaller load. However, most power grids also have supplemental power from natural gas or whatever else they choose. These plants are designed to provide peak loads. They generally cost more to generate a watt of electricity, but they don't have to be running at full capacity 24/7/365 like the base power plants, resulting in overall cheaper prices for electricity. Also, the grids plan for the future, if it looks like we'll need 10% more power next year, they better be able to generate it.
of course, you're assuming that the cost of installing a panel in arizona to meet your electricity needs is the same as in California. If you're spending $400/month on electricity, you probably need more panels than this guy did. By this logic, google could cut the power costs of their data center to 1/5th of the original cost by installing $36k worth of solar panels. . .
One thing that can make extended warranties worthwhile is that the cost to repair a broken item is much higher to you than it is to the company offering the warranty. For instance, if you have a laptop and something on the logic board breaks, you can easily spend $400 finding a replacement logic board. However the company making the machine really spends less than $200 on it (probably far less than that). The issue is that they are the only source of replacement parts, the whole vendor lock-in problem.
I've learned my lesson, and now buy extended warranties on laptops. The extended warranty on my Macbook has more than paid for itself already, and in the end got me an upgrade to a model released a year and a half after I bought mine. Hopefully I don't need to use the warranty again, but it's very nice knowing I don't have to worry about it. Plus, while it might be cheaper to repair some things on your own, you really need to value your time on getting something fixed. How long does it take to find the parts, a place to fix it etc. How long will your item be out of commission. It's quite convenient to call one number and get a box to return the item in the next day.
And as we start adding more cores to the mix, how much does performance improve? Right now branch predictors are easily in the 90%+ range for many codes. I'll use the 10% number for when the prediction fails, and assume that on a fail, we take 10 times as long to run the instruction. I'll also assume that branches occur every 10 instructions (I can't remember the exact number, so I'm using intentionally pessimistic numbers). Assuming a base CPI of 1 (although the analysis works the same for any cpi), we have a branch misprediction once every 90 instruction, and then must add 10 cycles for the penalty, so our overall speed is limited to 90% of the theoretical maximum. Even if we make all these numbers much worse, we can see that the problem is not branch prediction. If a branch was equally likely to be taken as not taken, two cores (each running a separate branch) could yield an appreciable speedup, this just isn't the case. A 10% speedup by using the wrong path branch scheme isn't horrible performance, but far shy of the performance 2 cores could give, and if we expand the number of cores, we greatly reduce that number.
This leads us to the other problem, in that you assumed that processor speed was the limiting factor. This in fact is not really true. Many processors can run significantly faster then they do, however the speeds aren't sustainable due to power limitations. When power is our concern, we've doubled the power of the processor and improved performance 10%. This is really an awful tradeoff. While power consumption grows with the square of the processor's frequency, this isn't nearly as bad. We can get a 10% performance gain by increasing the processor's power by only 21%.
As much as these new technologies look great, the question is how easy will it be to use? If the answer is harder than a single core processor (which it most assuredly would be), than the question becomes how much harder is it to use?
In recent times, there has been no end to proposed tools and languages to help express parallelism. These are made by extremely bright people, and many have some neat and interesting features. However, so far, few people can really take advantage of them. Experts can design programs on them and use them, but these experts are a far cry from your run of the mill people. These are not the programmers you can hire for $40k or even $80k oftentimes.
New technologies are needed to take advantage of parallel computing. However these technologies must be as easy to use as Visual C++ is (really it needs to be as easy as VB, but that's another story). So far they all have problems, and a programmer cannot have a serial mindset when programming these architectures. Unfortunately, the brain does not seem to be very good at expressing parallelism, and the tools we currently have do not do enough to prevent developers from shooting off their legs.
Will these new technologies be useful in snow leopard? Possibly, they will probably be used in Quicktime, and some of Apple's video software. It's possible that open source video codecs might take advantage of them, but that depends on whether people make research projects out of them. Photoshop might make use of it for some of their operations, but don't expect everything to be done that way, as it's expensive to rewrite complex algorithms in parallel.
I just laugh when I read everyone clamoring about how this technology will change the world... It is a step in the right direction, but there is no panacea to make parallel programming easy. The first step involves making libraries of many of the compute intensive functions available to programmers. Joe programmer can call library routines. . . at least if they fit into normal programming paradigms. Expect these libraries to be expensive though. Writing highly parallel optimized code to do the compute intensive operations people need is expensive. The experts capable of doing it are extremely expensive, and it isn't like they can do this work overnight, or in a week sometimes. Also, expect HDL coders to be in demand. They understand parallelism and might be capable of using these new tools.
It's actually a pretty smart way for Sony to prevent people from hacking the PS3. I'm sure part of this was at IBMs request, as IBM wants the CELL to be useful for other purposes than just gaming. By allowing Linux to easily run in a limited form, it means that people can play with the machine in Linux, but don't really have control over it. However, it also means fewer people are willing to spend the time required to hack the box and make a real linux media center machine out of it. I wouldn't doubt if this would already be available for the PS3, if not for the fact that it is so easy to get Linux installed on the device already.
As it stands, researchers already have access to play around with the Cell SPEs, and can do enough that there's no need to break it for their own needs. The general hobbyist who wants all the other stuff tend to not have the knowledge and resources to break in, and install linux, and thus no one has done it. Not a bad tradeoff, the research community gets to use the PS3 to play with Cell processors (helping IBMs goal of encouraging Cell development), and the hacking community has far fewer resources available to break the system, and less demand for it.
First, there are criminal acts (that are not civil charges) that requires someone to press charges for them to go through. I don't know the circumstances with these particular computer crimes, but they might require that the victim party presses charges for anyone to be punished. I don't see why someone couldn't make a civil agreement whereby one party agrees not to press criminal charges against another. I think this happens in lawsuits often enough where a party was harmed by another, and they agree to drop civil and criminal charges in response to payment. It's pretty clear cut when a criminal action is perpetrated directly against one or entity. Again, i'm not a lawyer or anything, but I'm pretty sure this sort of stuff is legal.
When I mean where the guilt lies, I mean where in the corporation, and how the charges should be filed. Just because we know a company did XYZ, we don't know where the orders came from, and who did what, and why.
Your comment is hilarious... Do you think that the Bush administration somehow controls the FBI and tells them that they can't prosecute a company for this? Do you honestly think the president of the united states is that powerful, that he has the power of a dictator? Even if he does have that power, would he use it so openly for this?
The FBI likely will investigate this matter, and the only way I could see media sentry getting out of criminal charges is by paying off Rev 3 enough that they drop all criminal charges against media sentry. Companies have been convicted of crimes before, and will be again. Sure the *AA has favor from congress, and the executive, but that doesn't exactly stretch over to the courts. At this point, I bet the *AA's first reaction will be (if they have any intelligence whatsoever) "i didn't know mediasentry did this", and back out of deals. The last thing they want is to be hit up with more lawsuits for hiring a criminal company.
The likely situation from all this is that media sentry is going to get hit, and hard. While you can make comments that you or I would already be arrested for this, it's only been a week, and a company the size of media sentry isn't exactly a flight risk. The hardest part of the case will be determining where exactly the guilt lies.
The futures market provides quite a few GREAT things for the economy. Sure, it may raise the cost of gas a bit at times, but that is a small price to pay for a steady supply. Without these markets in place we'd likely end up with many short-term shortages, and with it, much more variable pricing of gas. The importance of having a steady supply is FAR more important than saving 10 cents a gallon on gas.
you neglect to mention that fact that IMMEDIATELY following this statement the constitution lays out a list enumerating exactly what those powers are. If the line "common defense and general welfare" was taken to grant congress power over everything related to the above there would be no need to explicitly list what congress is allowed to do.
Additionally, if this line was to mean congress could do anything not explicitly forbidden by the first 9 amendments, there would be no need for the 10th amendment which states: "The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people." It is not arguing semantics to say that when the authors of the constitution wrote the line about "common defense and general welfare" they meant the powers that they were explicitly granting following that statement.
However, the real argument that has been used for years is at the end of enumerating congress' rights which states " To make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof." This has fueled the famous "necessary and proper" debate that has raged between strict and loose constructionists for years. However, I don't think anyone of the founders would argue that this statement gives the federal government the power to do anything not denied them in the first 9 amendments (hence the 10th amendment).
Of course, one of the main reasons that NASA exists in the first place is to show our military might. If we could send people to the moon it is obvious that we have accurate missiles that can make it across the world and still hit their target. NASA was the "peaceful" way of showing our military power and technological innovations, and it served its purpose quite well.
The problem today is that without a cold war, NASA doesn't have as much of a purpose. It's still around, and still doing neat things, but it isn't where innovation is being pushed as far.
While you make some good points in your comment, there are parts that are off. First, UNIX pipes are not an effective way to parallelize an application. UNIX pipes provide a method that tends to be inefficient, and will involve much "needless" copying of data (from your application to the pipe, the OS will then read in the data and write it to the other process which will then likely read the data into its address space). Additionally, UNIX pipes work well for steady state, but tend to have problems with starting up and stopping. This is true of many pipeline systems. They also lack the ability to easily communicate "side channel" information. There are other streaming programming languages that are being tried that will hopefully fix some of these problems, for instance, see the StreamIT project at MIT.
Additionally your examples given in your post about UNIX pipes neglect to mention the fact that a pipeline is only as fast as it's slowest component. The idea of breaking tar -czf into tar cf|gzip and fully utilizing 2 CPUs is laughable. tar (without compression) is an extremely simple program that doesn't require much CPU time to run, while gzip will likely run for a while on the data. You'll likely get the situation where gzip is running 100% of the time (assuming your I/O is fast enough), and tar is running 5-10% of the time. This is quite far from the 200% number quoted.
Pipelining is a fairly easy technique to exploit parallelism for some applications, but to really utilize many processors, SPMD (single program multiple data) techniques are necessary, as you hinted at with the worker threads (one way to achieve this). A well parallelized program should take advantage of both types of parallelism to maximize performance.
-How do semaphores affect SMP cache efficiency? Is the CPU notified to keep the data in shared cache?
Not specially, they are simply a special case of the problem: How to access data Several threads may compete for the same data, but if they are accessing the same data in one cache-line, it will lead to lots of communication (thrashing the cache).</blockquote>
I think you have this wrong. Sharing data in one cache line between processors is not always bad. In fact in multicores this can be a very good thing. What causes problems is when multiple threads are writing to a cache line. This will happen with the cache line that holds the semaphore, and can cause multiple invalidate messages being sent out if multiple threads are spinning on a lock, and the lock is poorly coded (a proper lock will just perform reads while spinning, which does not cause invalidate messages). In a multicore architecture locks can be done much more efficiently. The data being locked is likely in one processors L1 cache, and written back to the L2. Reads can be read from the L2 cache by any processor, and cached in their L1. Upon a write to the lock, the local copies are invalidated, and the update is propagated to the L2. As the L2 can be read by all processors relatively fast, this should not cause cache thrashing.
Cache thrashing on a multicore is associated with one thread having a much larger cache footprint than another, causing the other threads data to be evicted. The big problem with locks and semaphores is that if there is regularly contention for it, performance will suffer greatly due to a thread having to spin (or block to the OS) as it waits to obtain the lock.
Slashdot Mirror
Does anyone have any data whatsoever showing that with downloading on the internet, the small artists are being hurt? I'm sure there's plenty of data showing that the big guys are being hurt by online piracy, but that's really not a huge deal. The big artists will continue making music for only millions of dollars instead of tens of millions. Besides, even amongst the big artists, the big money tends to be in their concerts. As far as I can tell, the small artists have been selling just as many albums as before.
The one big thing that is currently hurting small artists (far more than illegal downloading) is the cost of gas. Now that they're paying $4/gallon, it's a lot more expensive to tour and promote your music. Breaking even, or making a profit on a tour was never an assured thing for small bands, and it's a lot harder to do today.
Phil
Just because Linux can be the best tool for a job doesn't mean it always is. If you can setup a windows MCE box in 2 hours, and it would take 10 hours to setup and tweak a linux box to do the same thing, then Linux might not be the better choice. You can make all the claims you want about it being able to do the same things as windows (not entirely true though), and it being free, but you've managed to completely dismiss the time element.
Additionally, a Windows MCE box (at the moment) has advantages over a myth box. For one, you don't need to pay to get listings. Those are automatic. Additionally, if you want to view Netflix on demand movies, you can install plugins on a Windows MCE box to let you browse and watch them from your TV. Also, as the parent said, the video codecs for playing back h.264 video on windows are better than those available on Linux (to the extent that many machines can watch HD h.264 video in Windows but not under Linux). Not all of these problems are insurmountable, and they may be fixed in the future. However, many people want their machine to do what they want RIGHT NOW, and not have to wait months or years to get things up to speed.
Both windows and linux have their strengths and weaknesses. For many tasks linux is better, and for some windows is. If I want a fileserver, webserver, router, etc, I'm going to use Linux. However, if I want to set up a media center box, I've found windows (gasp) Vista MCE to do the job perfectly fine. The interface is decent (not perfect, and not as customizable as myth apparently), it plays back the media I want (although I did have to install a few plugins to get everything I wanted), and for the most part, I don't have to actually use the keyboard/mouse. I can do most everything I need to do using the remote that comes with it.
Phil
Calling Bush "far right" is a huge laugh. One of the reasons Bush is so widely disliked across the spectrum is because he is not strongly conservative, and not strongly liberal. He is pushing a huge war in the middle east that is the exact opposite of one of his early campaign promises (to keep US troops at home and stay away from peace keeping missions). A "far right" president may have supported the war in Iraq, deposing Saddam, and throwing in a dictator friendly to the US governments interests. Sticking it out and wasting billions of dollars on bringing democracy to the middle east is not a "far right" idea, that's the idea of peace keeping which Bill Clinton was always big on. It's a feel good idea to help make people feel better for fighting a war.
Bush has also not pushed for any significant conservative legislation. His education "reforms" are very liberal, and gave the federal government more control over education. His views on immigration are the same as the democrats (open borders and amnesty). The only issues that people can point at him and say he is conservative are social issues. Bush is a strong Christian, which is an automatically bad thing. He is anti-abortion, which I can understand why people dislike that. However after the past 8 years, have any laws really reflected this? Some government money might now go to church-based charities, instead of only to non-denominational ones. Aside from that, has anything really changed on this front? Social issues like this are always big talking points before an election, but at the end of the day politicians rarely enact real and meaningful change in regard to them.
Bush's stance on the war on terror tends to be "right wing", but the libertarian fringe of the republican party (who are considered further right) tend to be against these movements. However the actions of the patriot act are things that both Republicans and Democrats have pushed for at one time or another. Both tend to seek increasing the governments power (although Republicans were, once upon a time the small government party), however they only tend to want to increase the power when they're in charge, hence the democrats objections to the patriot act while the republicans are in charge.
I would consider many Democratic and Republican actions to be "on the left". Both tend to favor big government, the distinction being how big. One wants socialized health care, the other wants this hideous bureaucratic and inefficient mess that they've created as "compromises". One wants the government to fully pay for peoples retirement, the other wants the people to have some say in how the government pays for their retirement. For most issues, neither party is looking to smaller government, but thankfully neither party is looking fully into socialism either.
Phil
This isn't that much of an issue really. The public has never been that well-informed, and even when they are people will claim they aren't for their own political purposes. The current solution that seems to be in use is to end democracy and the republic system that we have. It's what the well-meaning people of the USA have been pushing for years. Of course, at the same time they claim to want to make things more democratic. It seems like the two goals are contradictory, but they seem to work quite well together. Some well-meaning (often liberal, but not always) leader has an idea that he knows the public won't go for yet. The idea is well regarded by the intelligentsia, and they want it made into law.
They have two options to do this. One is to convince judges to make it law. However, this is hard to do and can be unreliable (but with the right judges it is much easier to do). The other is to use grass roots campaigns that won't stop fighting. If you have an idea that only 20% of the population likes, but they like it enough (particularly on the local/state level), they can push it through. They do this by continuously revoting on the issue. The idea being that if the first vote fails, get another one, and another one and another one . . . If they're lucky they can get the vote to occur at an election that most people don't care about (some local only election with traditionally low voter turnout). Eventually they can get it to pass, and now their plan is passed "democratically" through a referendum. The opposition is often not organized enough to stop it (most people don't support it, but aren't passionately opposed).
I always find it amusing how direct-voter referendums tend to be the LEAST democratic method of determining laws because of things like this. I'm sure people all over have seen the same sort of ideas steam-rolled through. It helps the intelligentsia pass their ideas without having an informed public, and still maintaining the idea that the people have their say.
Phil
I think you're missing the point that even without a welfare system, there are plenty of people who would take the job at the prevailing wage. Teenagers for instance.
You act like there is some God given right for people to earn good wages. The fact of the matter is there isn't. In all likelihood, if given the choice between paying the few people who need the job a living wage, or firing them, the companies would likely choose to fire the workers. The work they do just isn't worth the amount requested. As there exists a social welfare system, what is better from the government's perspective. Someone who needs help to make ends meet as their job doesn't quite pay enough, or someone who earns no money because they don't have enough skills to get any job.
Economic forces are so much more complicated than "if we pay someone more things would be better". But keep living in your little world. I'm sure some day economic forces and human nature will stop working, and be the way you want them.
Phil
If your job can be done by a 15 year old kid who has never had a job in his life, isn't dangerous, is easy to do, and you can find a replacement worker in a matter of days when someone quits, it really isn't a good job. It is not hard to get a better job than working in fastfood. Hell, if you show up on time for a month at one of those places people will think you're the best worker ever. But they don't really need the worlds best workers, and in fact don't want good workers. They'd much rather have cheap workers. They're easier to replace, and don't cost much.
I would venture to say that most people who are working fast food (and aren't at least managers) that seriously need the job for their income have something wrong with them that prevents them from getting a real job. Maybe it's something as simple as they have a couple kids, and have to regularly take time off for them, and so they're considered unreliable at other places. Sometimes they might be mentally challenged, or just really slow/lazy. Maybe they're just irresponsible adults who don't want a real job. Despite the economy being "bad" right now, with a high school education, and a hard work attitude (which means being willing to accept idiots in authority for a while when you're being paid shit), you are able to advance. You can get to a position in less than a year where you're making more than a living wage for yourself. You might not be living the high life, but it is not that difficult for a single person to live on less than $20,000/year (of course, if there are health considerations to be taken into account this changes, as the state of health care is a disaster).
Phil
What about all the OSX users?
This is spoken from someone working on a macbook, so don't mark me down as an anti-apple zealot. OSX eats up memory. 2GB is the minimum I'd want on a leopard machine. If you plan on using parallels at all, 3GB is likely not enough. Of course, I tend to always have tons of tabs open in safari, I run mail, a terminal, xchat, adium, textedit, itunes,and other stuff at the same time.
At the end of the day, RAM is dirt cheap. I can buy a 2GB stick for about $40. If my OS eats memory it really isn't a big deal. The real problem is that my machine can only accept so much RAM.
Phil
I think they have that feature already. There are a lot of FOSS people (at least on slashdot) who are also gamers. The ability to play almost every game in windows (combined with the fact that their video card drivers are generally more mature and offer slightly better performance) means that many FOSS people HAVE to have microsoft windows on their computers. Should microsoft care that they also have linux installed and use that sometimes, maybe even predominately? As long as these users have still purchased a legitimate copy of MS Windows, Microsoft is making their profits, regardless of how much their product is used on the machine.
Of course, if the gamers are pirating windows, then that's a different story, and they have no right to complain about Microsoft's ethics.
Phil
the number of cups varies, but normally there are 15 cups. 5 in the back row forming the triangle, and of course 2 water cups on each side. The cups should be relatively small (none of the giant 16oz cups). The game should also be played 3v3.
Of course, many different variants exist, but this is the common way I know.
Phil
Beer pong is a game played with paddles, throwing ping pong balls into triangles of cups is beirut.
Phil
and have you ever heard of the broken window fallacy (http://en.wikipedia.org/wiki/Parable_of_the_broken_window)?
Phil
Whoa there, this statement IS NOT a fact. Public works projects can help a slumping economy, but only if the public works project is needed, and absolutely helps expand the economy. There is more to it than that, but creating jobs does not necessarily expand the economy but can result in simple wealth redistribution. For example, if the government hired 10,000 people to dig a giant ditch, and than hired another 10,000 people to fill in the ditch, jobs would be created, but would it help the economy? The government doesn't magically have money, they need to obtain it somewhere. In this instance they've created 20,000 jobs, but added nothing to the economy. In fact, under such a situation, they've likely decreased the economy. Even if unemployment is really high, some of these people are likely not doing other (productive) jobs to dig a ditch and fill it in instead. This decreases the net value of the economy. Additionally, where is the money to pay these workers coming from? They either tax the people (reducing the money they have to create new jobs, and buy goods, decreasing the size of the economy) or print money, causing inflation, resulting in an inflation tax instead.
Of course, the real world is much more difficult, and I am not an economist, but I know not all economists believe that public works projects are good for the economy. The publics works projects in the great depression did not cure the depression, however government military spending did help bring us out of the depression (although, I imagine the average standard of living decreased during the war years, as the money was going into the war). One factor of public works projects that can also helps the economy (beyond the help the public works project itself does) in the long term, is the training that workers might receive working on the project, making them more productive afterward.
What I do know about pushing people into public works projects on renewable resources is that it would create jobs, and result in more renewable energy. However, if the cost of the energy is greater, than everyone is paying in higher overall costs (or taxes). It must also be noted that in a slumping economy, the costs of implementing large public works projects is cheaper, as there are often large numbers of unemployed people (who in the US are often earning money from the government already from the welfare system). This means the net cost of implementing these projects is cheaper due to being able to pay lower wages, and even cheaper still because you don't have to pay these people welfare benefits.
Maybe a real economist could plug through the numbers and predict if your proposed projects would help the economy (even than they'd be guessing). However, claiming it's a fact that public works projects help the economy is definitely not true.
Phil
this is partially true. It's true that the main load power stations always produce the same amount of power. Coal, hydro, nuclear can't change dynamically to use a smaller load. However, most power grids also have supplemental power from natural gas or whatever else they choose. These plants are designed to provide peak loads. They generally cost more to generate a watt of electricity, but they don't have to be running at full capacity 24/7/365 like the base power plants, resulting in overall cheaper prices for electricity. Also, the grids plan for the future, if it looks like we'll need 10% more power next year, they better be able to generate it.
of course, you're assuming that the cost of installing a panel in arizona to meet your electricity needs is the same as in California. If you're spending $400/month on electricity, you probably need more panels than this guy did. By this logic, google could cut the power costs of their data center to 1/5th of the original cost by installing $36k worth of solar panels. . .
Phil
One thing that can make extended warranties worthwhile is that the cost to repair a broken item is much higher to you than it is to the company offering the warranty. For instance, if you have a laptop and something on the logic board breaks, you can easily spend $400 finding a replacement logic board. However the company making the machine really spends less than $200 on it (probably far less than that). The issue is that they are the only source of replacement parts, the whole vendor lock-in problem.
I've learned my lesson, and now buy extended warranties on laptops. The extended warranty on my Macbook has more than paid for itself already, and in the end got me an upgrade to a model released a year and a half after I bought mine. Hopefully I don't need to use the warranty again, but it's very nice knowing I don't have to worry about it. Plus, while it might be cheaper to repair some things on your own, you really need to value your time on getting something fixed. How long does it take to find the parts, a place to fix it etc. How long will your item be out of commission. It's quite convenient to call one number and get a box to return the item in the next day.
Phil
And as we start adding more cores to the mix, how much does performance improve? Right now branch predictors are easily in the 90%+ range for many codes. I'll use the 10% number for when the prediction fails, and assume that on a fail, we take 10 times as long to run the instruction. I'll also assume that branches occur every 10 instructions (I can't remember the exact number, so I'm using intentionally pessimistic numbers). Assuming a base CPI of 1 (although the analysis works the same for any cpi), we have a branch misprediction once every 90 instruction, and then must add 10 cycles for the penalty, so our overall speed is limited to 90% of the theoretical maximum. Even if we make all these numbers much worse, we can see that the problem is not branch prediction. If a branch was equally likely to be taken as not taken, two cores (each running a separate branch) could yield an appreciable speedup, this just isn't the case. A 10% speedup by using the wrong path branch scheme isn't horrible performance, but far shy of the performance 2 cores could give, and if we expand the number of cores, we greatly reduce that number.
This leads us to the other problem, in that you assumed that processor speed was the limiting factor. This in fact is not really true. Many processors can run significantly faster then they do, however the speeds aren't sustainable due to power limitations. When power is our concern, we've doubled the power of the processor and improved performance 10%. This is really an awful tradeoff. While power consumption grows with the square of the processor's frequency, this isn't nearly as bad. We can get a 10% performance gain by increasing the processor's power by only 21%.
Phil
As much as these new technologies look great, the question is how easy will it be to use? If the answer is harder than a single core processor (which it most assuredly would be), than the question becomes how much harder is it to use?
In recent times, there has been no end to proposed tools and languages to help express parallelism. These are made by extremely bright people, and many have some neat and interesting features. However, so far, few people can really take advantage of them. Experts can design programs on them and use them, but these experts are a far cry from your run of the mill people. These are not the programmers you can hire for $40k or even $80k oftentimes.
New technologies are needed to take advantage of parallel computing. However these technologies must be as easy to use as Visual C++ is (really it needs to be as easy as VB, but that's another story). So far they all have problems, and a programmer cannot have a serial mindset when programming these architectures. Unfortunately, the brain does not seem to be very good at expressing parallelism, and the tools we currently have do not do enough to prevent developers from shooting off their legs.
Will these new technologies be useful in snow leopard? Possibly, they will probably be used in Quicktime, and some of Apple's video software. It's possible that open source video codecs might take advantage of them, but that depends on whether people make research projects out of them. Photoshop might make use of it for some of their operations, but don't expect everything to be done that way, as it's expensive to rewrite complex algorithms in parallel.
I just laugh when I read everyone clamoring about how this technology will change the world... It is a step in the right direction, but there is no panacea to make parallel programming easy. The first step involves making libraries of many of the compute intensive functions available to programmers. Joe programmer can call library routines. . . at least if they fit into normal programming paradigms. Expect these libraries to be expensive though. Writing highly parallel optimized code to do the compute intensive operations people need is expensive. The experts capable of doing it are extremely expensive, and it isn't like they can do this work overnight, or in a week sometimes. Also, expect HDL coders to be in demand. They understand parallelism and might be capable of using these new tools.
Phil
It's actually a pretty smart way for Sony to prevent people from hacking the PS3. I'm sure part of this was at IBMs request, as IBM wants the CELL to be useful for other purposes than just gaming. By allowing Linux to easily run in a limited form, it means that people can play with the machine in Linux, but don't really have control over it. However, it also means fewer people are willing to spend the time required to hack the box and make a real linux media center machine out of it. I wouldn't doubt if this would already be available for the PS3, if not for the fact that it is so easy to get Linux installed on the device already.
As it stands, researchers already have access to play around with the Cell SPEs, and can do enough that there's no need to break it for their own needs. The general hobbyist who wants all the other stuff tend to not have the knowledge and resources to break in, and install linux, and thus no one has done it. Not a bad tradeoff, the research community gets to use the PS3 to play with Cell processors (helping IBMs goal of encouraging Cell development), and the hacking community has far fewer resources available to break the system, and less demand for it.
Phil
First, there are criminal acts (that are not civil charges) that requires someone to press charges for them to go through. I don't know the circumstances with these particular computer crimes, but they might require that the victim party presses charges for anyone to be punished. I don't see why someone couldn't make a civil agreement whereby one party agrees not to press criminal charges against another. I think this happens in lawsuits often enough where a party was harmed by another, and they agree to drop civil and criminal charges in response to payment. It's pretty clear cut when a criminal action is perpetrated directly against one or entity. Again, i'm not a lawyer or anything, but I'm pretty sure this sort of stuff is legal.
When I mean where the guilt lies, I mean where in the corporation, and how the charges should be filed. Just because we know a company did XYZ, we don't know where the orders came from, and who did what, and why.
Phil
Your comment is hilarious... Do you think that the Bush administration somehow controls the FBI and tells them that they can't prosecute a company for this? Do you honestly think the president of the united states is that powerful, that he has the power of a dictator? Even if he does have that power, would he use it so openly for this?
The FBI likely will investigate this matter, and the only way I could see media sentry getting out of criminal charges is by paying off Rev 3 enough that they drop all criminal charges against media sentry. Companies have been convicted of crimes before, and will be again. Sure the *AA has favor from congress, and the executive, but that doesn't exactly stretch over to the courts. At this point, I bet the *AA's first reaction will be (if they have any intelligence whatsoever) "i didn't know mediasentry did this", and back out of deals. The last thing they want is to be hit up with more lawsuits for hiring a criminal company.
The likely situation from all this is that media sentry is going to get hit, and hard. While you can make comments that you or I would already be arrested for this, it's only been a week, and a company the size of media sentry isn't exactly a flight risk. The hardest part of the case will be determining where exactly the guilt lies.
Phil
The futures market provides quite a few GREAT things for the economy. Sure, it may raise the cost of gas a bit at times, but that is a small price to pay for a steady supply. Without these markets in place we'd likely end up with many short-term shortages, and with it, much more variable pricing of gas. The importance of having a steady supply is FAR more important than saving 10 cents a gallon on gas.
phil
you neglect to mention that fact that IMMEDIATELY following this statement the constitution lays out a list enumerating exactly what those powers are. If the line "common defense and general welfare" was taken to grant congress power over everything related to the above there would be no need to explicitly list what congress is allowed to do.
Additionally, if this line was to mean congress could do anything not explicitly forbidden by the first 9 amendments, there would be no need for the 10th amendment which states: "The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people." It is not arguing semantics to say that when the authors of the constitution wrote the line about "common defense and general welfare" they meant the powers that they were explicitly granting following that statement.
However, the real argument that has been used for years is at the end of enumerating congress' rights which states " To make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof." This has fueled the famous "necessary and proper" debate that has raged between strict and loose constructionists for years. However, I don't think anyone of the founders would argue that this statement gives the federal government the power to do anything not denied them in the first 9 amendments (hence the 10th amendment).
Phil
Of course, one of the main reasons that NASA exists in the first place is to show our military might. If we could send people to the moon it is obvious that we have accurate missiles that can make it across the world and still hit their target. NASA was the "peaceful" way of showing our military power and technological innovations, and it served its purpose quite well.
The problem today is that without a cold war, NASA doesn't have as much of a purpose. It's still around, and still doing neat things, but it isn't where innovation is being pushed as far.
Phil
While you make some good points in your comment, there are parts that are off. First, UNIX pipes are not an effective way to parallelize an application. UNIX pipes provide a method that tends to be inefficient, and will involve much "needless" copying of data (from your application to the pipe, the OS will then read in the data and write it to the other process which will then likely read the data into its address space). Additionally, UNIX pipes work well for steady state, but tend to have problems with starting up and stopping. This is true of many pipeline systems. They also lack the ability to easily communicate "side channel" information. There are other streaming programming languages that are being tried that will hopefully fix some of these problems, for instance, see the StreamIT project at MIT.
Additionally your examples given in your post about UNIX pipes neglect to mention the fact that a pipeline is only as fast as it's slowest component. The idea of breaking tar -czf into tar cf|gzip and fully utilizing 2 CPUs is laughable. tar (without compression) is an extremely simple program that doesn't require much CPU time to run, while gzip will likely run for a while on the data. You'll likely get the situation where gzip is running 100% of the time (assuming your I/O is fast enough), and tar is running 5-10% of the time. This is quite far from the 200% number quoted.
Pipelining is a fairly easy technique to exploit parallelism for some applications, but to really utilize many processors, SPMD (single program multiple data) techniques are necessary, as you hinted at with the worker threads (one way to achieve this). A well parallelized program should take advantage of both types of parallelism to maximize performance.
Phil