you do realize that much of the "hardware" we have today is little more than application specific instruction processors (ASIPs) and memory on a board (or SoC). For these hardware devices, much of the development work is in the firmware running on the processors. Oh, and much of that code was probably written by the processor vendor, and likely was obtained under a license agreement that doesn't allow you to release it. Now, if the hardware device contains flash or an eeprom, this isn't really an issue, as the code for these processors can be stored on there. However, many store the program data in the driver. This has a couple advantages, it's cheaper to manufacture the device (fewer components), more reliable (fewer components to fail) and if a bug is discovered in the ASIP code, the manufacturer can release new device drivers that automatically update the firmware of the device, without forcing the user to manually update it. Seems like device manufacturers would have to be stupid not to upload binary blobs to their devices. These binary blobs can't be open source for the reasons outlined above, and thus the device driver cannot be added to the linux kernel.
The issue isn't that people use cell phones on the road. The issue is that they're using a cell phone while on a section of road that needs their full attention. I can think of plenty of times when talking on my cell phone while driving isn't remotely dangerous. For instance, I might be driving on a stretch of interstate for 50 miles or so where I am not going through any major towns, and am not changing highways etc. In such an instance, talking on the cell phone is not such a distraction that you can't drive. In fact, it often helps keep you awake, as you really don't have much to do when cruise control is on and there's no one around.
Where it is dangerous is when you're driving around a city, or on crowded highways where you will likely have to be changing lanes and getting off at exits. In fact, I've been on the phone many times coming back from a concert or somewhere else that is a sizable distance away, where I tell whoever I'm talking to that I'll call them back in a bit, as I'm nearing home, and need to pay attention to the road again. It's just too difficult to try and pay attention to someone when you need to be watching everything around you. When you're talking to someone in the car, it can be as distracting as well, but normally, as they're aware of their surroundings, they know when not to talk, and what you can and can't talk about. Of course, having your hands free also helps, as holding a phone is one more thing to distract you.
I think laws outlawing all talking on the phone while driving are just too extreme. There are situations where it is appropriate, and situations where it isn't. Maybe it should just be enforced like seat belt laws (normally) are enforced. You won't be pulled over for talking on one, but if you are pulled over for something else (even if it was something minor that you normally wouldn't get a ticket for, but just a warning), and are talking on a cell phone, an additional fine will be added. I think this is fair, although it will likely be abused by law enforcement (although not as bad as the other way would be). An outright ban on cell phones just pisses people off who are legitimately safe when using their phones, and causes people who are on their phones to pay even less attention to the road, as they're now talking on the phone while looking out for cops, taking even more of their concentration.
Well,I can say I have had a couple of folks come into my shop wanting to pick my brain while I fixed their machine,and when I told them they couldn't back up their BD discs like they can their DVDs that was the end of that. Sony needs to realize that folks have kids,and no matter how careful you are kids can tear up a Sherman tank with a toothbrush without even meaning to. It would be one thing if they had made them cartridge style where it is much harder to scratch the thing,but they didn't. And even the most clueless home user has found out from their cousin/brother/uncle how to back up their DVDs so the kids don't accidentally trash them.
I'm calling bullshit on this one. VERY VERY VERY few clueless home users have found out how to back up their DVDs. Maybe the non DRM protected ones, but for the most part, people don't know how to do it. The DVD DRM has been very successful at preventing home users from making copies of their media. While some geeks know how to do it, and enough people probably have friends who can do it, or set up their computers so they can do it, this is not a common operation. Even if people have the software, it's often a pain to use, requiring the user to reencode their video etc.
Of course, when someone also posts that they think the economy is going straight into a depression, I have little reason to believe anything they say.
Lazarus Long was wise when he said: "A fake fortuneteller can be tolerated. But an authentic soothsayer should be shot on sight. Cassandra did not get half the kicking around she deserved."
And your test is completely meaningless when it comes to real-world performance of multiple processes. Few programs will launch 100 copies of themselves, it just doesn't make sense. Google's chrome might launch multiple processes (one for each tab), but that is not a big deal. Opening one new tab will take a couple miliseconds longer, which is not unreasonable. The real question with regards to multiple processes is how much does performance suffer in the steady state.
They better be able to pay for themselves in less than 10 years actually. Mostly because in 10 years many of your old lines will become obsolete. I imagine they can upgrade the signaling along them somewhat, but they're limited by the repeaters installed underwater. In 10 years, it's likely that a new line will be able to hold at least 10 times (probably a lot more) the traffic of the previous line (that costs about the same after normalizing for inflation, etc). This means that by the time the line is 10 years old its value has diminished significantly, and will soon cost more to operate than it is worth.
The specs aren't going to make a good graphics card driver with todays cards. This is not an arguable point if you know about the architecture of today's graphics cards. They look like a SIMD array of 100s of processors. They use internal memory (no caches), and run code on the card that transforms graphics requests into software that can run on the graphics card. The software is one of the most complex parts about the card, and just knowing how the hardware works is not enough. Your claim is like saying that "we have the specs of the x86 ISA, so designing a multithreaded video editing application with every feature imaginable, that is as fast as possible (fully utilizing sse2) is easy". I'm calling bullshit on this one. Software often requires large teams of programmers, and in the case of graphics card development, they must be experts at parallel programming, and fully understand the architecture of the cards they're working with.
You're never going to get "just works" with proprietary drivers. ATI drivers still have xv tearing and wine issues (wine fixable xv kind of not). nVidia's drivers have been having some serious issues, too. Not to mention their latest GPUs are in serious trouble from every end of the table.
That's interesting, my drivers "just work" in OSX, and last time I checked they were proprietary. Seems to me that you're just spreading FUD.
Phil
Because it makes THAT much of a difference that the code that operates your graphics card is compiled by nVidia, and not the ubuntu package maintainers. The whole argument for FOSS 3D video card drivers is just silly in my opinion. Very very very few people have the skills necessary to write good drivers for these chips (others can learn, but that takes months or years to do). The people who write these drivers do it as a full time job, and the drivers are some of the most important IP in a graphics card (if they were released under a gpl like license, it would be much easier for a new competitor to develop a product). Opening specs is somewhat possible, but costly to do, and a large pain to get the chance to do. At the end of the day it would allow the FOSS community to write drivers that half work on their cards. It could also break user's cards (which is an issue to the manufacturer, as many are warranted). It would take years to get a driver remotely comparable to the already available ones, and, that assumes a team of qualified people took up the challenge.
I just really love the fact that you consider it "evil" for them to not release source code for their products. For me, the nVidia drivers have always worked fine under linux. Of course, I don't game at all, and don't require powerful graphics cards in my machines. Personally I'd much rather have something that works well, even if I have to pay for it (which isn't the case with the nVidia drivers) than something that half works and is free and open.
the chance of any single card failing is still 20%. You are correct in this regard, however the parent said the chance that either card fails is 36%. The chance of one or more of your cards failing is not 20% when you have two cards. It's pretty obvious that if you had 100 cards, the chance of one or more of them failing is greater than 20%. The chance that any single card fails however is 20%. Of course, you could be trying to play a stupid semantics game where you imply that either card failing means the cards independently, but that was not the parents intention.
I seriously doubt that 1% of users are using 250GB of bandwidth a month, no less 400GB. That is just a tremendous amount of storage, more than most peoples hard drives. In a single year they would have to download over 3TB of data. The only people I can even imagine doing this are torrenting constantly, and are addicted to downloading, and don't even bother to watch/listen/play the things they are downloading. To download 250GB in a month, you need to sustain a download of ~96KB/s all month long. I just don't see it happening. I'm sure some slashdot users manage it where they have 4 or 5 geeks sharing a house and each torrenting 24/7, and playing WoW, etc constantly. But those are definitely the minority. I'd imagine the 1% figure is likely below 100GB at the moment.
Of course, the average will increase with time, but I don't see it as a bad thing to make people who use more bandwidth pay more. I wish there wasn't a monopoly controlling it, so pricing could be fairer, but I don't see why my parents who use their cable modem for email and web browsing should have to pay extra so that the cable company can upgrade their network for a few other people who are constantly downloading. Tiered service is likely the way of the future, it's no different than how most everything is sold. However, it is likely that ISPs will still have plans that are "unlimited within reason", and if they're smart, they will average a users downloading over more than a month to get a better average.
you do realize solar and wind power cannot really replace the base power loads of coal/nuclear plants (oil plants are rare now due to cost, and natural gas plants are generally run for supplemental power). Most plans so far have been looking into reducing the number of natural gas plants in use by using supplemental wind and solar plants. The real issue we will run into is the need to store power. Distributing it is great and helps allow us to rely on a larger percentage of "renewable" energy (as the odds that it's not sunny in places xyz, not windy in abc, etc starts to fall), but doesn't allow us to use the power 24/7. This is going to require an infrastructure capable of storing power for long time periods. The last thing we want is to have rolling blackouts on cloudy or windless days. It's not like you can just go "okay, for the next 3 hours this coal plant needs to produce more power", they just don't work that way (however natural gas plants are able to do this, which is why they're used for supplemental power despite being more expensive than coal).
Actually, the problem is not the same. Building a new factory needs new lines, sure, but the lines only have to go to a power plant (or rather to the last substation or whatever). This can be measured in 10s to 100s of miles. Not really that large scale, and there isn't a huge concern for power losses over this distance.
When building massive wind farms, the idea is that they're going to be built in areas without a large population center (say South Dakota). The power then needs to be delivered not 10s of miles, but rather 100s to 1000s of miles (the big demand for power is in the northeast and the west coast). This will require building huge lines that need to have low losses. This will likely mean building DC lines, and the cost of such infrastructure is huge compared to the cost of building lines 10s of miles. This means that despite the fact that wind power generation is currently less efficient (on a cost basis) than coal, the true cost will likely end up being even more. If the wind power is generated 1000 miles away, the real cost of the power has to factor in the cost of building the power plants (which still cannot be part of the base power load), the cost of building transmission lines, and the cost of the significant power losses that will occur when transferring the power 100s or thousands of miles.
True, food is basically just energy that we need to survive. But last time I checked, the only way we can affordably make food is by growing/raising it. Also, human nature means that we want to eat the foods we like, and not synthetic stuff. Last time I checked their was no way to grow corn, wheat, cows, pigs etc using nuclear power. I guess you could create artificial lights and greenhouses for the crops, but that still requires a lot of area to be done. Raising animals would require even more "land", and just wouldn't be feasible in such an environment.
Food alone is enough of a reason to say that an arcology holding 1+ million people cannot be self sustaining. Unless the arcology is built to a massive scale, it will not provide all the food for the people living in it to eat. Just as todays cities have to import food, so too would one of these. It just isn't economical to try and grow/raise all your food in one of these things. While technically a massive structure could grow all that food, it doesn't make sense to do it.
off the shelf components does not mean that the components can be bought at microwaves R us. It means that if you are looking to manufacture a microwave, you can contact wholesalers and buy 1000's of these parts, and that the parts are for the most part fairly simple. You just need someone who knows what parts are needed. The controller in a microwave is generally considered simple, so it's not the best example. How about the controllers involved in medical imaging equipment? These can be used for years as well. Also, the code in other devices mentioned earlier is by no means simple, and the control software used in many industrial settings are quite complex as well. They might not have huge performance requirements, but they often have realtime requirements, and have to account for many different factors.
Also, for the record, I was not arguing that all of this software should be patented, just that it's copyright should exist, and 14 years might not be long enough for it.
Uh, some of your arguments about code in embedded devices don't make sense. Sure these things require a mix of hardware and software, but most of the hardware is not that complex. They're generally made from commercial off the shelf parts, and are easy to recreate (often using reference hardware models). The idea of an individual recreating the hardware isn't the issue, but the idea of another company using the software from someone else's design is a big issue.
As for the airplane code, you're assuming that the code is always written by the airplane manufacturer and owned by them. If this is the case, then yes it isn't very useful. However, for many other devices this is most definitely not the case. Companies regularly buy components for their systems, this includes the intellectual property to use both their hardware (if applicable) and software designs. The code in microwaves, CD players, DVD players, etc isn't useful without the hardware or device, but the devices can be made, and the cost of the software is not always negligible (although for some of these devices I would imagine the software code is basically reference code that comes with the hardware). However, even here, the code needs copyright protection. Say reference code is given that the user of the chip is licensed to modify, and use in their devices. The code is written in C (say an h.264 decoder for an embedded platform), with some asm hooks to custom hardware chips unique to their device. A competitor could use that same source code to accelerate the creation of a competing product. Sure the asm hooks would need to be changed, and the hardware isn't the exact same, but the time and effort going into the code is often as great as the time and effort going into the design of the hardware. Graphics cards are a perfect example of this. The hardware is complex, but without the driver software (which is extremely complex), they would simply look like an array of vector processing units with a fast locally attached memory. The device would be incredibly difficult to use, and would cost a fraction of the cost to produce. However the code makes the cards useable, and has the ability to transform calls to directX and openGL into parallel software that is optimized for the GPU pipeline. As we head toward massively multicore designs, the enabling software can easily cost as much or more than the design of the device itself. Don't think competitors would not use someone else's code if it was free, and their were no legal ramifications for doing so.
you'd be surprised how much code is 14+ years old. Some of it is code that has been heavily modified over the years though. It's actually an interesting question... what would be the legal status of source code that was "stolen" from a project, but the project then fell out of copyright. Would someone be able to use it? Software is different as there is the source, and then the actual software you use. I'm not entirely sure how copyrights deal with changing works either.
As for the other legacy applications, I know the code in my car is over 14 years old (helps that my car is 15 years old), but there is also code in VCRs, cd players, microwaves, and a whole ton of embedded devices that would be relevant. The embedded community has tons of software in their products, and many parts of their products are in use for more than 14 years. Think about the code on airplanes and such. Getting the code to pass certifications and what not is a huge hurdle. Most of the code will still be used and sold for more than 14 years.
Your problem is that you limit yourself to thinking of software in the desktop sense. A large portion of code is not desktop-centric.
Making money off service and support can be a viable business model for some software. Especially software that have academic origins, and therefore already have many of the "hard" problems solved, and are "mature" software that doesn't need to have new features constantly added (think linux, {my,pg}sql, apache). However, there is a lot of other software out there that doesn't fit this model. Are you telling me Adobe could make the same money giving away photoshop while providing service and support for it? Especially when a half dozen other companies would jump in offering service and support as well. What incentive would they then have to continue developing photoshop, or their other products? They spend millions developing a product, and for what... to be in the same position everyone else is already in? Games are the same way. Very few open source games have been successful, and many OSS games are just clones or extensions to commercial games. These things just require too much hard work and are not "academic" enough to justify their development for free.
I do agree that copyright terms are for too long. However a 7 or 14 year copyright is just too short. Particularly with software. Many consumer devices sold are really little more than software (the hardware can be bought most anywhere). If someone could use the same binaries, it would be way too easy to recreate another product without expense. I also believe that profits for music and movies are justified. I think a more appropriate lifetime is in the 25-30 years. Enough to help justify the original work, and prevent others from simply duplicating someone else's work for profit.
The big issue with this commentary on growth rates is that 3% is higher than the natural growth rate in most developed countries. The vast majority of population growth in the US today is due to immigration, as families aren't having as many children as they used to. Most of europe actually has a negative growth rate (which will wreck havok with their social programs as the workforce retires).
Most of the population growth in the world is taking place in the developing world. India, China, asia in general has huge growth. In the more developed world, women are more likely to work, and effective birth control and social stigmas has pushed down the birth rate to a maintainable pace.
Actually, Hardware is closer to SW than you think when it comes to the pricing model. Intellectual property laws are what makes the hardware valuable. In reality there is little hardware that has a marginal cost approaching that of the hardwares price (at least among high-end hardware, high volume commodity RAMs, and flash chips behave slightly differently than processors and video cards). It's very similar to software in that regard, as the first unit costs millions or billions to make (designing the processor, creating a fab, designing masks, etc), and each subsequent chip costs relatively little to make. The manufacturers than charge a premium on the cost of the hardware (far greater than the marginal cost of the hardware) to make up their initial costs. The only difference with software is that the marginal cost of the software is essentially zero.
The argument that intellectual property shouldn't exist is just stupid. Without intellectual property games would cease to exist, we likely wouldn't have Moore's law driving the semiconductor industry, high-end productivity software (such as CAD, photoshop, etc) might exist in some free versions, but maybe not, as the free versions are often based on commercial products (attempting to clone them). In fact a large portion of the US economy today is based on intellectual property. It's the intellectual property that has allowed such things as FOSS to exist in the first place.
People really need to understand that just because some good has a low marginal cost, it is not "free" to produce. The current intellectual property situation does not produce ideal solutions, but the proposed methods (at least on/.) to solve these problems are FAR worse than the problem. Half the comments think it should be abolished completely (which would make todays economic downturn seem like nothing), and the other half thing draconian restrictions on how long copyright/patents etc should last are the answer. However not all intellectual property issues are the same. Some ideas take a long time to catch on, and some industries are more conservative on changing than others. If patents/copyrights expired so quickly, we'd see few new ideas come out for such industries, as the profit motive would be completely removed.
First, the only way to bring the changes Al Gore wanted to pass is through either totalitarian restrictions on energy usage, or higher prices on energy (for instance, the high cost of gas right now has actually reduced the amount of gas consumed in the US). The other option to reduce energy usage without directly effecting costs is to have huge government subsidies for alternative energy sources. However this increases the prices indirectly (as we all have to pay for them), and actually doesn't do as much to reduce usage, as the price of the energy (that is directly seen) doesn't increase.
The other issue is that even if we had cut our energy consumption by 5-10%, it wouldn't drastically reduce the cost of energy. The reason is that there has been a huge increase in demand for energy from the developing world (primarily india and china).
I will say most republican policies would have done little to help. Drilling in the arctic would help a little for a few years, and encourage more oil exploration, but overall, the high cost of oil isn't so horrible for the economy. It will push change to using more efficient cars, and use more "green" energy sources.
9.5/100,000/year seems EXTREMELY low. Unless it doesn't include natural deaths of old age and stuff. Considering the average human life is ~75 years, you can easily expect to see ~1/75 people die every year. If the probability of dying in any given year was ~1/10,000, there would be people living thousands of years...
Slashdot Mirror
you do realize that much of the "hardware" we have today is little more than application specific instruction processors (ASIPs) and memory on a board (or SoC). For these hardware devices, much of the development work is in the firmware running on the processors. Oh, and much of that code was probably written by the processor vendor, and likely was obtained under a license agreement that doesn't allow you to release it. Now, if the hardware device contains flash or an eeprom, this isn't really an issue, as the code for these processors can be stored on there. However, many store the program data in the driver. This has a couple advantages, it's cheaper to manufacture the device (fewer components), more reliable (fewer components to fail) and if a bug is discovered in the ASIP code, the manufacturer can release new device drivers that automatically update the firmware of the device, without forcing the user to manually update it. Seems like device manufacturers would have to be stupid not to upload binary blobs to their devices. These binary blobs can't be open source for the reasons outlined above, and thus the device driver cannot be added to the linux kernel.
Phil
The issue isn't that people use cell phones on the road. The issue is that they're using a cell phone while on a section of road that needs their full attention. I can think of plenty of times when talking on my cell phone while driving isn't remotely dangerous. For instance, I might be driving on a stretch of interstate for 50 miles or so where I am not going through any major towns, and am not changing highways etc. In such an instance, talking on the cell phone is not such a distraction that you can't drive. In fact, it often helps keep you awake, as you really don't have much to do when cruise control is on and there's no one around.
Where it is dangerous is when you're driving around a city, or on crowded highways where you will likely have to be changing lanes and getting off at exits. In fact, I've been on the phone many times coming back from a concert or somewhere else that is a sizable distance away, where I tell whoever I'm talking to that I'll call them back in a bit, as I'm nearing home, and need to pay attention to the road again. It's just too difficult to try and pay attention to someone when you need to be watching everything around you. When you're talking to someone in the car, it can be as distracting as well, but normally, as they're aware of their surroundings, they know when not to talk, and what you can and can't talk about. Of course, having your hands free also helps, as holding a phone is one more thing to distract you.
I think laws outlawing all talking on the phone while driving are just too extreme. There are situations where it is appropriate, and situations where it isn't. Maybe it should just be enforced like seat belt laws (normally) are enforced. You won't be pulled over for talking on one, but if you are pulled over for something else (even if it was something minor that you normally wouldn't get a ticket for, but just a warning), and are talking on a cell phone, an additional fine will be added. I think this is fair, although it will likely be abused by law enforcement (although not as bad as the other way would be). An outright ban on cell phones just pisses people off who are legitimately safe when using their phones, and causes people who are on their phones to pay even less attention to the road, as they're now talking on the phone while looking out for cops, taking even more of their concentration.
Phil
I'm calling bullshit on this one. VERY VERY VERY few clueless home users have found out how to back up their DVDs. Maybe the non DRM protected ones, but for the most part, people don't know how to do it. The DVD DRM has been very successful at preventing home users from making copies of their media. While some geeks know how to do it, and enough people probably have friends who can do it, or set up their computers so they can do it, this is not a common operation. Even if people have the software, it's often a pain to use, requiring the user to reencode their video etc.
Of course, when someone also posts that they think the economy is going straight into a depression, I have little reason to believe anything they say.
Phil
Lazarus Long was wise when he said: "A fake fortuneteller can be tolerated. But an authentic soothsayer should be shot on sight. Cassandra did not get half the kicking around she deserved."
Phil
uh huh, tens of thousands of lines of asm....
~/x264-snapshot-20080812-2245/common/x86$ wc -l *.asm
165 cabac-a.asm
91 cpu-32.asm
51 cpu-64.asm
437 dct-32.asm
223 dct-64.asm
316 dct-a.asm
874 deblock-a.asm
659 mc-a2.asm
933 mc-a.asm
428 pixel-32.asm
1615 pixel-a.asm
600 predict-a.asm
383 quant-a.asm
968 sad-a.asm
519 x86inc.asm
124 x86util.asm
8386 total
And your test is completely meaningless when it comes to real-world performance of multiple processes. Few programs will launch 100 copies of themselves, it just doesn't make sense. Google's chrome might launch multiple processes (one for each tab), but that is not a big deal. Opening one new tab will take a couple miliseconds longer, which is not unreasonable. The real question with regards to multiple processes is how much does performance suffer in the steady state.
Phil
Uh, in the UK you can be forced to (http://yro.slashdot.org/article.pl?sid=07/10/02/1237215)
phil
They better be able to pay for themselves in less than 10 years actually. Mostly because in 10 years many of your old lines will become obsolete. I imagine they can upgrade the signaling along them somewhat, but they're limited by the repeaters installed underwater. In 10 years, it's likely that a new line will be able to hold at least 10 times (probably a lot more) the traffic of the previous line (that costs about the same after normalizing for inflation, etc). This means that by the time the line is 10 years old its value has diminished significantly, and will soon cost more to operate than it is worth.
Phil
The specs aren't going to make a good graphics card driver with todays cards. This is not an arguable point if you know about the architecture of today's graphics cards. They look like a SIMD array of 100s of processors. They use internal memory (no caches), and run code on the card that transforms graphics requests into software that can run on the graphics card. The software is one of the most complex parts about the card, and just knowing how the hardware works is not enough. Your claim is like saying that "we have the specs of the x86 ISA, so designing a multithreaded video editing application with every feature imaginable, that is as fast as possible (fully utilizing sse2) is easy". I'm calling bullshit on this one. Software often requires large teams of programmers, and in the case of graphics card development, they must be experts at parallel programming, and fully understand the architecture of the cards they're working with.
Phil
That's interesting, my drivers "just work" in OSX, and last time I checked they were proprietary. Seems to me that you're just spreading FUD. Phil
Because it makes THAT much of a difference that the code that operates your graphics card is compiled by nVidia, and not the ubuntu package maintainers. The whole argument for FOSS 3D video card drivers is just silly in my opinion. Very very very few people have the skills necessary to write good drivers for these chips (others can learn, but that takes months or years to do). The people who write these drivers do it as a full time job, and the drivers are some of the most important IP in a graphics card (if they were released under a gpl like license, it would be much easier for a new competitor to develop a product). Opening specs is somewhat possible, but costly to do, and a large pain to get the chance to do. At the end of the day it would allow the FOSS community to write drivers that half work on their cards. It could also break user's cards (which is an issue to the manufacturer, as many are warranted). It would take years to get a driver remotely comparable to the already available ones, and, that assumes a team of qualified people took up the challenge.
I just really love the fact that you consider it "evil" for them to not release source code for their products. For me, the nVidia drivers have always worked fine under linux. Of course, I don't game at all, and don't require powerful graphics cards in my machines. Personally I'd much rather have something that works well, even if I have to pay for it (which isn't the case with the nVidia drivers) than something that half works and is free and open.
Phil
the chance of any single card failing is still 20%. You are correct in this regard, however the parent said the chance that either card fails is 36%. The chance of one or more of your cards failing is not 20% when you have two cards. It's pretty obvious that if you had 100 cards, the chance of one or more of them failing is greater than 20%. The chance that any single card fails however is 20%. Of course, you could be trying to play a stupid semantics game where you imply that either card failing means the cards independently, but that was not the parents intention.
Phil
I seriously doubt that 1% of users are using 250GB of bandwidth a month, no less 400GB. That is just a tremendous amount of storage, more than most peoples hard drives. In a single year they would have to download over 3TB of data. The only people I can even imagine doing this are torrenting constantly, and are addicted to downloading, and don't even bother to watch/listen/play the things they are downloading. To download 250GB in a month, you need to sustain a download of ~96KB/s all month long. I just don't see it happening. I'm sure some slashdot users manage it where they have 4 or 5 geeks sharing a house and each torrenting 24/7, and playing WoW, etc constantly. But those are definitely the minority. I'd imagine the 1% figure is likely below 100GB at the moment.
Of course, the average will increase with time, but I don't see it as a bad thing to make people who use more bandwidth pay more. I wish there wasn't a monopoly controlling it, so pricing could be fairer, but I don't see why my parents who use their cable modem for email and web browsing should have to pay extra so that the cable company can upgrade their network for a few other people who are constantly downloading. Tiered service is likely the way of the future, it's no different than how most everything is sold. However, it is likely that ISPs will still have plans that are "unlimited within reason", and if they're smart, they will average a users downloading over more than a month to get a better average.
Phil
you do realize solar and wind power cannot really replace the base power loads of coal/nuclear plants (oil plants are rare now due to cost, and natural gas plants are generally run for supplemental power). Most plans so far have been looking into reducing the number of natural gas plants in use by using supplemental wind and solar plants. The real issue we will run into is the need to store power. Distributing it is great and helps allow us to rely on a larger percentage of "renewable" energy (as the odds that it's not sunny in places xyz, not windy in abc, etc starts to fall), but doesn't allow us to use the power 24/7. This is going to require an infrastructure capable of storing power for long time periods. The last thing we want is to have rolling blackouts on cloudy or windless days. It's not like you can just go "okay, for the next 3 hours this coal plant needs to produce more power", they just don't work that way (however natural gas plants are able to do this, which is why they're used for supplemental power despite being more expensive than coal).
Phil
Actually, the problem is not the same. Building a new factory needs new lines, sure, but the lines only have to go to a power plant (or rather to the last substation or whatever). This can be measured in 10s to 100s of miles. Not really that large scale, and there isn't a huge concern for power losses over this distance.
When building massive wind farms, the idea is that they're going to be built in areas without a large population center (say South Dakota). The power then needs to be delivered not 10s of miles, but rather 100s to 1000s of miles (the big demand for power is in the northeast and the west coast). This will require building huge lines that need to have low losses. This will likely mean building DC lines, and the cost of such infrastructure is huge compared to the cost of building lines 10s of miles. This means that despite the fact that wind power generation is currently less efficient (on a cost basis) than coal, the true cost will likely end up being even more. If the wind power is generated 1000 miles away, the real cost of the power has to factor in the cost of building the power plants (which still cannot be part of the base power load), the cost of building transmission lines, and the cost of the significant power losses that will occur when transferring the power 100s or thousands of miles.
Phil
True, food is basically just energy that we need to survive. But last time I checked, the only way we can affordably make food is by growing/raising it. Also, human nature means that we want to eat the foods we like, and not synthetic stuff. Last time I checked their was no way to grow corn, wheat, cows, pigs etc using nuclear power. I guess you could create artificial lights and greenhouses for the crops, but that still requires a lot of area to be done. Raising animals would require even more "land", and just wouldn't be feasible in such an environment.
Phil
Food alone is enough of a reason to say that an arcology holding 1+ million people cannot be self sustaining. Unless the arcology is built to a massive scale, it will not provide all the food for the people living in it to eat. Just as todays cities have to import food, so too would one of these. It just isn't economical to try and grow/raise all your food in one of these things. While technically a massive structure could grow all that food, it doesn't make sense to do it.
Phil
off the shelf components does not mean that the components can be bought at microwaves R us. It means that if you are looking to manufacture a microwave, you can contact wholesalers and buy 1000's of these parts, and that the parts are for the most part fairly simple. You just need someone who knows what parts are needed. The controller in a microwave is generally considered simple, so it's not the best example. How about the controllers involved in medical imaging equipment? These can be used for years as well. Also, the code in other devices mentioned earlier is by no means simple, and the control software used in many industrial settings are quite complex as well. They might not have huge performance requirements, but they often have realtime requirements, and have to account for many different factors.
Also, for the record, I was not arguing that all of this software should be patented, just that it's copyright should exist, and 14 years might not be long enough for it.
Phil
Uh, some of your arguments about code in embedded devices don't make sense. Sure these things require a mix of hardware and software, but most of the hardware is not that complex. They're generally made from commercial off the shelf parts, and are easy to recreate (often using reference hardware models). The idea of an individual recreating the hardware isn't the issue, but the idea of another company using the software from someone else's design is a big issue.
As for the airplane code, you're assuming that the code is always written by the airplane manufacturer and owned by them. If this is the case, then yes it isn't very useful. However, for many other devices this is most definitely not the case. Companies regularly buy components for their systems, this includes the intellectual property to use both their hardware (if applicable) and software designs. The code in microwaves, CD players, DVD players, etc isn't useful without the hardware or device, but the devices can be made, and the cost of the software is not always negligible (although for some of these devices I would imagine the software code is basically reference code that comes with the hardware). However, even here, the code needs copyright protection. Say reference code is given that the user of the chip is licensed to modify, and use in their devices. The code is written in C (say an h.264 decoder for an embedded platform), with some asm hooks to custom hardware chips unique to their device. A competitor could use that same source code to accelerate the creation of a competing product. Sure the asm hooks would need to be changed, and the hardware isn't the exact same, but the time and effort going into the code is often as great as the time and effort going into the design of the hardware. Graphics cards are a perfect example of this. The hardware is complex, but without the driver software (which is extremely complex), they would simply look like an array of vector processing units with a fast locally attached memory. The device would be incredibly difficult to use, and would cost a fraction of the cost to produce. However the code makes the cards useable, and has the ability to transform calls to directX and openGL into parallel software that is optimized for the GPU pipeline. As we head toward massively multicore designs, the enabling software can easily cost as much or more than the design of the device itself. Don't think competitors would not use someone else's code if it was free, and their were no legal ramifications for doing so.
Phil
you'd be surprised how much code is 14+ years old. Some of it is code that has been heavily modified over the years though. It's actually an interesting question... what would be the legal status of source code that was "stolen" from a project, but the project then fell out of copyright. Would someone be able to use it? Software is different as there is the source, and then the actual software you use. I'm not entirely sure how copyrights deal with changing works either.
As for the other legacy applications, I know the code in my car is over 14 years old (helps that my car is 15 years old), but there is also code in VCRs, cd players, microwaves, and a whole ton of embedded devices that would be relevant. The embedded community has tons of software in their products, and many parts of their products are in use for more than 14 years. Think about the code on airplanes and such. Getting the code to pass certifications and what not is a huge hurdle. Most of the code will still be used and sold for more than 14 years.
Your problem is that you limit yourself to thinking of software in the desktop sense. A large portion of code is not desktop-centric.
Phil
Making money off service and support can be a viable business model for some software. Especially software that have academic origins, and therefore already have many of the "hard" problems solved, and are "mature" software that doesn't need to have new features constantly added (think linux, {my,pg}sql, apache). However, there is a lot of other software out there that doesn't fit this model. Are you telling me Adobe could make the same money giving away photoshop while providing service and support for it? Especially when a half dozen other companies would jump in offering service and support as well. What incentive would they then have to continue developing photoshop, or their other products? They spend millions developing a product, and for what... to be in the same position everyone else is already in? Games are the same way. Very few open source games have been successful, and many OSS games are just clones or extensions to commercial games. These things just require too much hard work and are not "academic" enough to justify their development for free.
I do agree that copyright terms are for too long. However a 7 or 14 year copyright is just too short. Particularly with software. Many consumer devices sold are really little more than software (the hardware can be bought most anywhere). If someone could use the same binaries, it would be way too easy to recreate another product without expense. I also believe that profits for music and movies are justified. I think a more appropriate lifetime is in the 25-30 years. Enough to help justify the original work, and prevent others from simply duplicating someone else's work for profit.
Phil
The big issue with this commentary on growth rates is that 3% is higher than the natural growth rate in most developed countries. The vast majority of population growth in the US today is due to immigration, as families aren't having as many children as they used to. Most of europe actually has a negative growth rate (which will wreck havok with their social programs as the workforce retires).
Most of the population growth in the world is taking place in the developing world. India, China, asia in general has huge growth. In the more developed world, women are more likely to work, and effective birth control and social stigmas has pushed down the birth rate to a maintainable pace.
phil
Actually, Hardware is closer to SW than you think when it comes to the pricing model. Intellectual property laws are what makes the hardware valuable. In reality there is little hardware that has a marginal cost approaching that of the hardwares price (at least among high-end hardware, high volume commodity RAMs, and flash chips behave slightly differently than processors and video cards). It's very similar to software in that regard, as the first unit costs millions or billions to make (designing the processor, creating a fab, designing masks, etc), and each subsequent chip costs relatively little to make. The manufacturers than charge a premium on the cost of the hardware (far greater than the marginal cost of the hardware) to make up their initial costs. The only difference with software is that the marginal cost of the software is essentially zero.
The argument that intellectual property shouldn't exist is just stupid. Without intellectual property games would cease to exist, we likely wouldn't have Moore's law driving the semiconductor industry, high-end productivity software (such as CAD, photoshop, etc) might exist in some free versions, but maybe not, as the free versions are often based on commercial products (attempting to clone them). In fact a large portion of the US economy today is based on intellectual property. It's the intellectual property that has allowed such things as FOSS to exist in the first place.
People really need to understand that just because some good has a low marginal cost, it is not "free" to produce. The current intellectual property situation does not produce ideal solutions, but the proposed methods (at least on /.) to solve these problems are FAR worse than the problem. Half the comments think it should be abolished completely (which would make todays economic downturn seem like nothing), and the other half thing draconian restrictions on how long copyright/patents etc should last are the answer. However not all intellectual property issues are the same. Some ideas take a long time to catch on, and some industries are more conservative on changing than others. If patents/copyrights expired so quickly, we'd see few new ideas come out for such industries, as the profit motive would be completely removed.
Phil
First, the only way to bring the changes Al Gore wanted to pass is through either totalitarian restrictions on energy usage, or higher prices on energy (for instance, the high cost of gas right now has actually reduced the amount of gas consumed in the US). The other option to reduce energy usage without directly effecting costs is to have huge government subsidies for alternative energy sources. However this increases the prices indirectly (as we all have to pay for them), and actually doesn't do as much to reduce usage, as the price of the energy (that is directly seen) doesn't increase.
The other issue is that even if we had cut our energy consumption by 5-10%, it wouldn't drastically reduce the cost of energy. The reason is that there has been a huge increase in demand for energy from the developing world (primarily india and china).
I will say most republican policies would have done little to help. Drilling in the arctic would help a little for a few years, and encourage more oil exploration, but overall, the high cost of oil isn't so horrible for the economy. It will push change to using more efficient cars, and use more "green" energy sources.
Phil
9.5/100,000/year seems EXTREMELY low. Unless it doesn't include natural deaths of old age and stuff. Considering the average human life is ~75 years, you can easily expect to see ~1/75 people die every year. If the probability of dying in any given year was ~1/10,000, there would be people living thousands of years...
Phil