First off, he was assigning them to null, not testing whether they are null, so the syntax is correct.
Secondly, while I know it's somewhat rude make an issue of sigs, you are aware that actual children are harmed in the making of child pornography, right?
The Supreme Court just ruled that U.S. states can now pretty much ignore international law at will. You misunderstood the ruling in Medellin. What it held was, absent a Congressional grant of power to do so, the President may not order a State to comply with a treaty. If the President want to bind a state court to the terms of a treaty, he has to get Congress unambiguously on board. To me, this seems like a good balance between the power of the Federal government to conduct foreign policy and the State's right not to be bound absent specific Congressional authorization.
Plus, I thought/.ers were supposed to be against executive power as a general matter. Think of the howls if the President ordered some blog to obey a international court ordering the take down a libelous or defamatory post.
First off, I'd hardly call Visual Studio either clunky or big - I have an XP VM almost solely for VS and it has always been responsive. The interface is fairly customizable and the debugging is top-notch (much better that SunStudio, the only other one I've tried).
If you don't like IDEs (which is apparent), then use your favorite text editor or shell to search for "functionName(*);" in all files. Shit, `grep -e WriteForce*\&* *.h*` ought to get you less than a screenfull of results, one of which is the prototype (regexp wizards can do better, I'm sure).
Finally (and I think most importantly), if you have to wonder whether a function changes a passed variable then either it is incorrectly named or you have more serious issues with program flow. A sensible naming scheme makes it clear what data is input versus output.
What I need is the extra RAM that the GNU libstdc++ implementation of the string class takes up. I develop for a handheld device, and its 4 MB of RAM is a lot smaller than the 1 GB of RAM that you're probably used to working with. Absolutely. I should have qualified my post by excepting cases like yours where RAM is super-tight (embedded/handheld/etc. . . ) and cases where performance is at a huge premium (tight loops/bottlenecks/mission-critical/real-time). That said, in the vast majority of cases the performance gain from using char* are vastly outweighed by the simplicity, clarity and power of std::string. Heck, automatic destruction when the string goes out of scope is, IMHO, worth the cost of admission alone.
I disagree on your dislike of references - when you type the function name, intellisense (or whatever) pops up the relevant prototypes and you can see immediately whether the parameters are type& or const type&. This gives the benefit of uniform function calling, since passing by reference or const reference should be the default unless there is an explicit need for pass-by-value.
As to char*, there are very few places I can see it being necessary (ifstream.read((char*)(&data),sizeof(data) does NOT count). You can pass char* to the string constructor and you can get back with string::c_str() -- what else do you need?
In most states, the jury swears an oath to uphold the law. I don't know if that oath means anything to you, but to me it was an important reminder that the trial is not about the jury or what they believe but about our duty to the law. If you believe so strongly that your conception of justice is more legitimate than the laws passed by legislature that you cannot take that oath in good faith then you should not sit on a jury.
Of course, the jury will always have the defacto power to nullify as there is no double-jeopardy and courts are generally unwilling to go into jury motivation. It just saddens me that we can have intelligent people encouraging people to swear a false oath in order to somehow produce a more just result.
Depending on where you get your figures, as much of 50% of US nuclear power is generated from recycled Soviet uranium, either extracted from decommissioned warheads or excess manufactured product that was in the pipeline at the time of collapse. The US also has a large number of vintage-era nuclear weapons that are no longer considered militarily viable (the trigger mechanisms decay quite a bit) and so could be recycled. Finally, if the going ever gets really bad, we can always reprocess our spent fuel for Plutonium and/or use breeder reactors to make the stuff - this is the primary mode in which the Japanese nuclear industry sustains itself without outside supply, although the cheap price of Uranium makes them feel kind of dumb.
In short, the US does not need to import a single gram of fissile material to run indefinitely. Solar/Wind/etc. . are fine ideas for the long term but do not meet our power needs today. We should absolutely invest in these alternative technologies and, while we are at it, invest in conservation and efficiency. Unfortunately, right now, we are making almost 50% of our power from coal that is massively environmentally destructive from the second it is strip-mined out of the ground to its large final carbon contribution. Nuclear power is the only technology currently available that can put a dent in coal usage. If you show me an alternative that can scale to 400 TerraWattHours, I'll withdraw that claim.
But a web address often has a 1-to-1 corespondence with its contents. Knowing the address is one simple - and undetectable - step from knowing the contents. They are doing an unconstitutional search here. Heavens I wouldn't want the feds to know I had visited http://mail.google.com/mail/ or https://www.paypal.com/us/cgi-bin/webscr?cmd=_acco unt - then they could figure out all my email and PayPal transactions! Address only implies content for publicly available resources - resources in which you have no reasonable expectation of privacy. If you want to keep something private, stick a login screen in front of it or encrypt it.
As far as your statement that this "search" (which it isn't) is unconstitutional, I defy you to show me where in the constitution the government is restricted from determined to whom a person is writing letters. We are guaranteed privacy for the content of the letter not anonymity in writing letters. These are not synonymous! The constitution is a very specific document - please do not pretend that it forbids practice X just because you think X is incompatible with your normative view of a free society.
Ok, I'm not a huge supporter of DRM but this sort of FUD about key revocation for hardware devices is foolish at best (at worst, it a distraction from more material concerns regarding DRM). Supposing, arguendo, that a hardware key is indeed compromised - an event I think is not very likely considering that crackers will prefer to attack software based system first. A simple (if somewhat multi-pronged) plan will more than satisfy the majority of consumers:
(1) Network enabled players will automatically update - built in WiFi with WPA/WEP support would cover the vast majority of people technically inclined enough to buy a next-gen player. This is the ideal solution but need to be enforced at the player level.
(2) Downloadable ISOs for those without network-enabled players that want it NOW. While they are downloading, offer them to sign up for (3)
(3) A web form and phone number that will (a) send out an upgrade disk to your home and (b) ask if you would like to automatically receive all updates in the future as soon as they are available. Of course, everyone that registered their product for warranty will get the update mailed as well.
(4) Ask/Require retail places that sell players/movies to keep a stock of these disks handy. All these places have sophisticated POS computers that can make sure that anyone buying a disk with a revoked key gets flagged and is given the opportunity to get an update disk.
Now, are there going to be fuckups? Of course but with this plan a vast majority of people would experience failure exactly once, after which they will sign themselves up for the automatic mailings (which will, of course, be timed to arrive before the retail release of the new disks). To make things a little sweeter, put some content on the update disks - an old movie would do very nicely and give the consumer a reason not to think that this is a chore (yes, it is a chore, but, due to the expense and technical difficulty, I can't imagine that hardware keys will be compromised all that often).
I bet I could get that off while the police are shouting "SEARCH WARRANT". Can't be aborted either since no applications (like shutdown -a) can initialize while windows is shutting down.
OTOH, my setup isn't very secure. I set up a TrueCrypt AES/BLOWFISH drive (using a file, not a partition) and moved some "sensitive" things on there (frequent flyer #s, finance records, etc. ..) but I leave it mounted.
For the average user (not the average/.er) there's just no need for anything like this.
BTW, my back of the envelope calculation: 8.5x11in * 1800dpi * 1800dpi * 256colors = 77 billion bits of information, or about 9GB. I think 1800dpi is the output you get with a high quality laser printer. I doubt that "theoretical maximum" is achievable, because even the best printers are designed only to impress the human eye.
A4 paper is not 8.5x11. RTFA or at least RTF title of the article!
My buddy worked at the Portland airport on installing banks of digital displays and asked about how the airport could afford hundreds of screens @ $2000 ea. Apparently, LG practically gave them all away to the airport as a way to get people (especially businesspeople) thinking about the utility of large displays and the reliability of their brand (after all, they use them in airports!).
I see this as a win-win situation. The public gets much better airport displays than we used to and the company gets nearly free advertising in a format that suits them.
I will admit that the rolodex-y thing is (was) pretty cool though.
With all due respect to the noble goals of such an enterprise, these anonymous networks quickly degenerate into places where child porn gets spread (browse Freenet for while).
Privacy is not a shield behind which to hide illegal activity - allowing them to do so works against our goals (by "our goals", I mean the right to keep you private legal activities private).
I'm surprised Ars Technica doesn't consider unDRM and Sidda, which work on all modern encrypted media (incl. video). Microsoft has tried very hard to suppress all mention of unDRM and I guess it worked (it does take some persistent googling to get the files).
Still, it works. I downloaded an encrypted rental from aebn and gave it a try . . .
The already did when they introduced the internet. Remember, the internet interpets censorship as damage and routes around it. If Google really wanted to they would have no problem keeping all their operations outside the great firewall (but they would have to forego some profitability in the process).
It's not like Google is short of the technical expertise on the matter. Meanwhile, I'm no CS major but I can think of a few steps that would help:
(1) FreeGoogle desktop application that allows you use your home computer as a proxy to fetch google content from mainland China. Use very weak SSL to obfuscate the content (but not destroy people's home computers). Aggregate a list of all these IPs and distribute accordingly. This could be extended to other sites on a per-computer whitelist basis (eg: Wikipedia, NYTimes, CNN).
(2) "Unofficial Google Servers" that essentially perform (1) but on a higher-bandwidth scale. Don't bind them to any DNS entries, just distribute the IPs. When the firewall blocks them, move on to a different IP. Lather Rinse Repeat.
(3) Google-News-Packs: Download all the content from the front page of news.google.com, strip the pictures and zip the contents. Distribute freely. Especially the ones about China.
How hard can the Chinese government make life for Google if they refuse to set a single foot inside the country (or Hong Kong)? Technically, I'm sure that Google would win this arms race if they only had half of the balls necessary to fight it.
You *should* pay more for your Linux distro downloading and massive P2P traffic because you are consuming more of the limited amount of bandwidth that the network offers and, ultimately, if everyone used their connection like yours, the whole thing would grind to a halt.
Most/.ers are exceedingly lucky in that their broadband is subsidised (in a sense) by the hordes of users that pay $50/month for much more bandwidth than they need (ie, web, email and an occasional episode of Lost from iTunes) while people like us (and I am particularly guilty in this respect) keep our outbound connections at >90%. We are using more than our share of the resources and should honestly be required to pay more.
To extend your car/road analogy, consider that you lived in a sub-division (ugh . . i know, bear with me) that connects to a super-highway a little ways away through a small toll loop. Initially, most houses aren't connected to the road but slowly they get connected. Some people just drive their Geo-Metro (SSH + PINE) to the email-store, most people drive medium sized cars and others still drive the SUV to the iTunes store and fill their trunk. A small percentage, however, have Mack Trucks making deliveries all day every day.
This analogy is apt because the local cable ciruit (in the case of cable to the last mile, by far the most common form of broadband in the USA) is a lot like the common road leading to the highway. Unused space is wasted but when it gets filled, everyone's traffic suffers. If everyone is to maintain the same quality of service in the face of increasing numbers of trucks they are going to either have to add lanes (capital intensive) or make some lanes that the trucks aren't allowed to enter (throttling).
Those trucks ought to pay a higher toll than everyone else or perhaps they should confine their deliveries to off-peak hours to as not to clog up the morning rush - that's just plain fairness to the other drivers. They are getting more out of it.
The parent's claim that the fact that the rest of the users aren't using their connection to the fullest misses the point for two reasons:
(1) Currently, the service in the US is unlimited transfer at a fixed bandwidth. This is like the subdivision road-builder sinking the cost of a two-lane road in capital invesetment and trying to recoup it by allowing any and all traffic for a fixed monthly fee. He has already built the road and its capacity is so underutilised that he doesn't care what sort of traffic he attracts so long as they pay him *something*. As traffic increases, those trucks will be more and more of a burden since he won't want to add more lanes for a small fraction of users while the rest complain of long commutes.
My prediction is that as network utilisation rises in the last-mile situations (especially as more legitimate media downloading sites come online), those providers will be keen on keeping their costs down for the majority of users that aren't saturating the pipe every second of every day. Furthermore, this is clearly in the interests of the majority of consumers since the bottom 90% of them use about 50% of the bandwidth. Again, this isn't a concern now because of network over-capacity but it will be soon.
Thos 90% of users would be perfeclty OK with traffic shaping and throttling because it will IMPROVE their quality of service and LOWER their prices. Joe Sixpack doesn't want to subsidize the delivery of your Linux Distros and he shouldn't have to. I have no problem paying for business-class services becuase I use it like business class service and I frankly think it would be unethical for me to get a home-class service and fill it like a business one.
In the end, the consumer will win. And by consumer, i mean the guy that uses his internet connection reasonably and in a way that does not over-tax the system and require massive capital upgrades (or will chip in to pay for those upgrades) not the whiny/. user that wants to get more than he pai
There's no difference in the memory management between pro and home. They differ only in a few minor features, notably managed-OS issues but also RDP, IIS, ASR. Read all about it here:
The comercial internet has always been a "sender pays" system at the middle ranks. Unless you a tier 1 provider you are going to be charged for every byte you inject into the network.
God help you if you try to introduce a "buyer pays" system!
"One can backdate e-mails through rolling back a computer's built-in clock"
For those that didn't RTFA, the next line was:
"'It will certainly prove that you sent the e-mail when you said you did,' Saltzman said. 'You can just blame the delay on the network.'"
The point is that a large time gap between sent and received headers will be invariably be interpreted as `a technological problem, not a dishonesty problem.
The move to digital TV will *make* the government money and will help finance your children's education.
The TV companies have a lock on the best part of the EM spectrum and they don't have to pay shit for it (relative to its worth). The revenues from auctioning it off will easily cover
Don't mean to interject, but there are two kinds of wind resistance: Laminar Flow and Turbulent Flow.
The former represent the smooth "moving air out the way of where you are going" resistance whereas the latter involve eddies and the like.
For a quick example, when you turn on your kitchen faucet on relatively low pressure you see a clear, constant stream of water - this is laminar flow. Turn the pressure all the way up, however, and you have turbulent flow - you will notice the boundary because the water is no longer transparent and continuous but rather all frothy.
As far as drags go, they aren't strictly polynomial but to a good approximation the resistance to laminar flow is proportional to the square of the velocity and the fourth power for turbulent flow.
This is a major reason why wind-tunneling is so necessary and why golf-balls have those dimples - the dimples cause the air to flow more smoothly and thus change the resistance from v^4 to v^2. Good deal.
Finally, back to the topic at hand, cars are way more efficient at 55 than 70 because at that level, almost all the flow becomes turbulent. This is also why it pays to put the windows up and run the AC (and forego the roof-rack).
Oren
PS When we studied this in high school "applied physics" our teacher started the whole unit of fluid dynamics with a large caveat: we don't understand shit about fluid dynamics. Even the simplest problems (water in a pipe to feed your house) is solved using tables and fudge factors. He even quoted Heisenbeg:
"When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first."
A net profit for everyone AND a whole boatload of nifty new wireless technlogies. 10 stations * 30 Hz * 480 * 320 > 10 Mbit/sec of bandwidth*!
*Actually the number is closer and i didn't even count color/grayscale bits. I'm being conservative because I'm not sure that all of space is actually usable for data transmission (you'll never notice a pixel being off but you might on the web).
Slashdot Mirror
First off, he was assigning them to null, not testing whether they are null, so the syntax is correct.
Secondly, while I know it's somewhat rude make an issue of sigs, you are aware that actual children are harmed in the making of child pornography, right?
Plus, I thought
First off, I'd hardly call Visual Studio either clunky or big - I have an XP VM almost solely for VS and it has always been responsive. The interface is fairly customizable and the debugging is top-notch (much better that SunStudio, the only other one I've tried).
If you don't like IDEs (which is apparent), then use your favorite text editor or shell to search for "functionName(*);" in all files. Shit, `grep -e WriteForce*\&* *.h*` ought to get you less than a screenfull of results, one of which is the prototype (regexp wizards can do better, I'm sure).
Finally (and I think most importantly), if you have to wonder whether a function changes a passed variable then either it is incorrectly named or you have more serious issues with program flow. A sensible naming scheme makes it clear what data is input versus output.
I disagree on your dislike of references - when you type the function name, intellisense (or whatever) pops up the relevant prototypes and you can see immediately whether the parameters are type& or const type&. This gives the benefit of uniform function calling, since passing by reference or const reference should be the default unless there is an explicit need for pass-by-value.
As to char*, there are very few places I can see it being necessary (ifstream.read((char*)(&data),sizeof(data) does NOT count). You can pass char* to the string constructor and you can get back with string::c_str() -- what else do you need?
In most states, the jury swears an oath to uphold the law. I don't know if that oath means anything to you, but to me it was an important reminder that the trial is not about the jury or what they believe but about our duty to the law. If you believe so strongly that your conception of justice is more legitimate than the laws passed by legislature that you cannot take that oath in good faith then you should not sit on a jury.
Of course, the jury will always have the defacto power to nullify as there is no double-jeopardy and courts are generally unwilling to go into jury motivation. It just saddens me that we can have intelligent people encouraging people to swear a false oath in order to somehow produce a more just result.
Sitting at home with any amount of computing power has to be more energy efficient than taking a car anywhere.
Depending on where you get your figures, as much of 50% of US nuclear power is generated from recycled Soviet uranium, either extracted from decommissioned warheads or excess manufactured product that was in the pipeline at the time of collapse. The US also has a large number of vintage-era nuclear weapons that are no longer considered militarily viable (the trigger mechanisms decay quite a bit) and so could be recycled. Finally, if the going ever gets really bad, we can always reprocess our spent fuel for Plutonium and/or use breeder reactors to make the stuff - this is the primary mode in which the Japanese nuclear industry sustains itself without outside supply, although the cheap price of Uranium makes them feel kind of dumb.
In short, the US does not need to import a single gram of fissile material to run indefinitely. Solar/Wind/etc. . are fine ideas for the long term but do not meet our power needs today. We should absolutely invest in these alternative technologies and, while we are at it, invest in conservation and efficiency. Unfortunately, right now, we are making almost 50% of our power from coal that is massively environmentally destructive from the second it is strip-mined out of the ground to its large final carbon contribution. Nuclear power is the only technology currently available that can put a dent in coal usage. If you show me an alternative that can scale to 400 TerraWattHours, I'll withdraw that claim.
References:
http://www.eia.doe.gov/cneaf/electricity/epm/epm_sum.html
http://www.usec.com/v2001_02/Content/News/NewsTemplate.asp?page=/v2001_02/Content/News/NewsFiles/04-13-03.htm
http://www.defencetalk.com/forums/archive/index.php/t-215.html
As far as your statement that this "search" (which it isn't) is unconstitutional, I defy you to show me where in the constitution the government is restricted from determined to whom a person is writing letters. We are guaranteed privacy for the content of the letter not anonymity in writing letters. These are not synonymous! The constitution is a very specific document - please do not pretend that it forbids practice X just because you think X is incompatible with your normative view of a free society.
Ok, I'm not a huge supporter of DRM but this sort of FUD about key revocation for hardware devices is foolish at best (at worst, it a distraction from more material concerns regarding DRM). Supposing, arguendo, that a hardware key is indeed compromised - an event I think is not very likely considering that crackers will prefer to attack software based system first. A simple (if somewhat multi-pronged) plan will more than satisfy the majority of consumers:
(1) Network enabled players will automatically update - built in WiFi with WPA/WEP support would cover the vast majority of people technically inclined enough to buy a next-gen player. This is the ideal solution but need to be enforced at the player level.
(2) Downloadable ISOs for those without network-enabled players that want it NOW. While they are downloading, offer them to sign up for (3)
(3) A web form and phone number that will (a) send out an upgrade disk to your home and (b) ask if you would like to automatically receive all updates in the future as soon as they are available. Of course, everyone that registered their product for warranty will get the update mailed as well.
(4) Ask/Require retail places that sell players/movies to keep a stock of these disks handy. All these places have sophisticated POS computers that can make sure that anyone buying a disk with a revoked key gets flagged and is given the opportunity to get an update disk.
Now, are there going to be fuckups? Of course but with this plan a vast majority of people would experience failure exactly once, after which they will sign themselves up for the automatic mailings (which will, of course, be timed to arrive before the retail release of the new disks). To make things a little sweeter, put some content on the update disks - an old movie would do very nicely and give the consumer a reason not to think that this is a chore (yes, it is a chore, but, due to the expense and technical difficulty, I can't imagine that hardware keys will be compromised all that often).
Windows+r , shutdown -s -f -t 01
.) but I leave it mounted.
/.er) there's just no need for anything like this.
I bet I could get that off while the police are shouting "SEARCH WARRANT". Can't be aborted either since no applications (like shutdown -a) can initialize while windows is shutting down.
OTOH, my setup isn't very secure. I set up a TrueCrypt AES/BLOWFISH drive (using a file, not a partition) and moved some "sensitive" things on there (frequent flyer #s, finance records, etc. .
For the average user (not the average
A4 paper is not 8.5x11. RTFA or at least RTF title of the article!
My buddy worked at the Portland airport on installing banks of digital displays and asked about how the airport could afford hundreds of screens @ $2000 ea. Apparently, LG practically gave them all away to the airport as a way to get people (especially businesspeople) thinking about the utility of large displays and the reliability of their brand (after all, they use them in airports!).
I see this as a win-win situation. The public gets much better airport displays than we used to and the company gets nearly free advertising in a format that suits them.
I will admit that the rolodex-y thing is (was) pretty cool though.
With all due respect to the noble goals of such an enterprise, these anonymous networks quickly degenerate into places where child porn gets spread (browse Freenet for while).
Privacy is not a shield behind which to hide illegal activity - allowing them to do so works against our goals (by "our goals", I mean the right to keep you private legal activities private).
>
I'm surprised Ars Technica doesn't consider unDRM and Sidda, which work on all modern encrypted media (incl. video). Microsoft has tried very hard to suppress all mention of unDRM and I guess it worked (it does take some persistent googling to get the files).
Still, it works. I downloaded an encrypted rental from aebn and gave it a try . . .
An unsafe user cannot be made safe by the system without serious frustration.
.)
No amount of software remediation will fix a a defective human peripheral (a clue-by-four, on the other hand . .
The already did when they introduced the internet. Remember, the internet interpets censorship as damage and routes around it. If Google really wanted to they would have no problem keeping all their operations outside the great firewall (but they would have to forego some profitability in the process).
It's not like Google is short of the technical expertise on the matter. Meanwhile, I'm no CS major but I can think of a few steps that would help:
(1) FreeGoogle desktop application that allows you use your home computer as a proxy to fetch google content from mainland China. Use very weak SSL to obfuscate the content (but not destroy people's home computers). Aggregate a list of all these IPs and distribute accordingly. This could be extended to other sites on a per-computer whitelist basis (eg: Wikipedia, NYTimes, CNN).
(2) "Unofficial Google Servers" that essentially perform (1) but on a higher-bandwidth scale. Don't bind them to any DNS entries, just distribute the IPs. When the firewall blocks them, move on to a different IP. Lather Rinse Repeat.
(3) Google-News-Packs: Download all the content from the front page of news.google.com, strip the pictures and zip the contents. Distribute freely. Especially the ones about China.
How hard can the Chinese government make life for Google if they refuse to set a single foot inside the country (or Hong Kong)? Technically, I'm sure that Google would win this arms race if they only had half of the balls necessary to fight it.
Powerstrip: http://www.entechtaiwan.com/ps.htm
duh . . .
You *should* pay more for your Linux distro downloading and massive P2P traffic because you are consuming more of the limited amount of bandwidth that the network offers and, ultimately, if everyone used their connection like yours, the whole thing would grind to a halt.
/.ers are exceedingly lucky in that their broadband is subsidised (in a sense) by the hordes of users that pay $50/month for much more bandwidth than they need (ie, web, email and an occasional episode of Lost from iTunes) while people like us (and I am particularly guilty in this respect) keep our outbound connections at >90%. We are using more than our share of the resources and should honestly be required to pay more.
/. user that wants to get more than he pai
Most
To extend your car/road analogy, consider that you lived in a sub-division (ugh . . i know, bear with me) that connects to a super-highway a little ways away through a small toll loop. Initially, most houses aren't connected to the road but slowly they get connected. Some people just drive their Geo-Metro (SSH + PINE) to the email-store, most people drive medium sized cars and others still drive the SUV to the iTunes store and fill their trunk. A small percentage, however, have Mack Trucks making deliveries all day every day.
This analogy is apt because the local cable ciruit (in the case of cable to the last mile, by far the most common form of broadband in the USA) is a lot like the common road leading to the highway. Unused space is wasted but when it gets filled, everyone's traffic suffers. If everyone is to maintain the same quality of service in the face of increasing numbers of trucks they are going to either have to add lanes (capital intensive) or make some lanes that the trucks aren't allowed to enter (throttling).
Those trucks ought to pay a higher toll than everyone else or perhaps they should confine their deliveries to off-peak hours to as not to clog up the morning rush - that's just plain fairness to the other drivers. They are getting more out of it.
The parent's claim that the fact that the rest of the users aren't using their connection to the fullest misses the point for two reasons:
(1) Currently, the service in the US is unlimited transfer at a fixed bandwidth. This is like the subdivision road-builder sinking the cost of a two-lane road in capital invesetment and trying to recoup it by allowing any and all traffic for a fixed monthly fee. He has already built the road and its capacity is so underutilised that he doesn't care what sort of traffic he attracts so long as they pay him *something*. As traffic increases, those trucks will be more and more of a burden since he won't want to add more lanes for a small fraction of users while the rest complain of long commutes.
My prediction is that as network utilisation rises in the last-mile situations (especially as more legitimate media downloading sites come online), those providers will be keen on keeping their costs down for the majority of users that aren't saturating the pipe every second of every day. Furthermore, this is clearly in the interests of the majority of consumers since the bottom 90% of them use about 50% of the bandwidth. Again, this isn't a concern now because of network over-capacity but it will be soon.
Thos 90% of users would be perfeclty OK with traffic shaping and throttling because it will IMPROVE their quality of service and LOWER their prices. Joe Sixpack doesn't want to subsidize the delivery of your Linux Distros and he shouldn't have to. I have no problem paying for business-class services becuase I use it like business class service and I frankly think it would be unethical for me to get a home-class service and fill it like a business one.
In the end, the consumer will win. And by consumer, i mean the guy that uses his internet connection reasonably and in a way that does not over-tax the system and require massive capital upgrades (or will chip in to pay for those upgrades) not the whiny
There's no difference in the memory management between pro and home. They differ only in a few minor features, notably managed-OS issues but also RDP, IIS, ASR. Read all about it here:
http://www.winsupersite.com/showcase/windowsxp_ho
The comercial internet has always been a "sender pays" system at the middle ranks. Unless you a tier 1 provider you are going to be charged for every byte you inject into the network.
God help you if you try to introduce a "buyer pays" system!
The move to digital TV will *make* the government money and will help finance your children's education.
The TV companies have a lock on the best part of the EM spectrum and they don't have to pay shit for it (relative to its worth). The revenues from auctioning it off will easily cover
Don't mean to interject, but there are two kinds of wind resistance: Laminar Flow and Turbulent Flow.
The former represent the smooth "moving air out the way of where you are going" resistance whereas the latter involve eddies and the like.
For a quick example, when you turn on your kitchen faucet on relatively low pressure you see a clear, constant stream of water - this is laminar flow. Turn the pressure all the way up, however, and you have turbulent flow - you will notice the boundary because the water is no longer transparent and continuous but rather all frothy.
As far as drags go, they aren't strictly polynomial but to a good approximation the resistance to laminar flow is proportional to the square of the velocity and the fourth power for turbulent flow.
This is a major reason why wind-tunneling is so necessary and why golf-balls have those dimples - the dimples cause the air to flow more smoothly and thus change the resistance from v^4 to v^2. Good deal.
Finally, back to the topic at hand, cars are way more efficient at 55 than 70 because at that level, almost all the flow becomes turbulent. This is also why it pays to put the windows up and run the AC (and forego the roof-rack).
Oren
PS When we studied this in high school "applied physics" our teacher started the whole unit of fluid dynamics with a large caveat: we don't understand shit about fluid dynamics. Even the simplest problems (water in a pipe to feed your house) is solved using tables and fudge factors. He even quoted Heisenbeg:
"When I meet God, I am going to ask him two questions: Why relativity? And why turbulence? I really believe he will have an answer for the first."
A net profit for everyone AND a whole boatload of nifty new wireless technlogies. 10 stations * 30 Hz * 480 * 320 > 10 Mbit/sec of bandwidth*!
*Actually the number is closer and i didn't even count color/grayscale bits. I'm being conservative because I'm not sure that all of space is actually usable for data transmission (you'll never notice a pixel being off but you might on the web).