It seems to me that the problem of memory fragmentation could be addressed at some level by the operating system by giving each process its own 32-bit virtual address space. By giving each process a full 32-bit address space, it's unlikely that a process could sufficently fragment a 32-bit address space to the point where large allocations can no longer be made. In a 64-bit address space, it would be nearly impossible.
Of course, this scheme would impact performance as it would require the operating system to do a TLB flush after each context switch. It would also require more memory to map virtual pages to physical memory differently for each process. Still, it seems like it could be a workable solution for most memory fragmentation problems. Is there something I'm missing?
The problem with your logic is that any event becomes negligable if you average the effects over a long enough time frame. The death of every person in Asia would probably be negligable if you averaged the effects over the span of human existance.
I understand your argument that 9/11 might just be an isolated statistical spike that doesn't really indicate a change in the overall death rate caused by terrorist acts against Americans. On the other hand, it might mark the start of a period where there are significantly more terrorism related deaths than the historical average. My opinion is that 9/11 was a signigicant spike in a general upward trend, but not enough time has elapsed to tell for sure.
Sept. 11th, in and of itself, was really a fairly negligable event - ten times that number die each year in car accidents in the US, and more than a thousand times that number are currently in prison in the US for violent crimes
Roughly 3000 people died from the terrorist attacks on 9/11. If those deaths had been spread throughout the United States over the course of a year, then you might have a point... but they weren't. On average there are about 6500 deaths in the United States every day. Even looking at a national level, an extra 3000 deaths in one day is significant. If you only look at New York, then the spike in the death rate is huge. If one morning there was a giant car crash in downtown Manhattan that killed 3000 people and caused $20 billion in property damage, would you consider that event negligible?
What are we going to have to do to convince "ordinary users" to visit WindowsUpdate once in a while?
Better yet, what are we going to have to do to convince "ordinary users" not to run executable email attachments?
Some users are smart enough not to run executable attachments. Some users are too dumb to know how to open any attachment. It's all the other users that cause most of the problem. Unfortunately, there are a lot of them.
Running as root is dangerous, but is more dangerous than the average home user is used to? Probably not. The average user probably runs windows from a single user account with admin rights. For most people, the recycle bin is the only protection from stupid mistakes.
Anybody know how this works? I've got a gen 4 iPod and it doesn't act like a normal USB storage device. On Windows PCs, the iPod doesn't show up as a drive unless iTunes has been installed. A little investigation shows iTunes is bundled with a driver for the iPod and installs an iPod service. I admit that I haven't actually tried to boot from the iPod, but I can't see how it would work. Does anyone know if the iPod mini is different in this respect and functions as a standard USB storage device?
My personal experience has been that windows driver support is still much better than linux, although linux driver support is much better than it used to be. That's not suprising since most hardware manufacturers are going to develop and test to their windows drivers since that probably accounts for 90% of the systems their hardware will be used in.
As an example, I recently installed linux on a spare windows box I had with the intention of making it a file server. The motherboard had an ITE8212 raid controller I was planning to use as an additional IDE interface (without RAID), but it wasn't supported. Some googling eventually turned up a kernel patch that added the driver. After recompiling the kernel, the drives did show up, but if I tried to access them the system would hang. I gave up and got a couple of cheap IDE cards with the SiI680 chipset which is supported by the kernel. With the driver supplied with the 2.4.28 kernel, only drives attached to the first controller showed up even though lspci showed both controllers were detected. I then upgraded to the 2.6.10 kernel which uses a different driver for that chipset and drives attached to both controllers showed up. However, if I tried to read or write much data to the drives, I got DMA errors and the driver would reset and enter PIO mode. Sometimes it would even hang the system. All of these cards worked flawlessly in Windows. I've been trying to get something to work under linux for two weeks and still no luck.
I'm about to buy different IDE cards again. If anybody knows of a card that will definitely work, let me know. I don't want to try a fourth set of cards. This is starting to get expensive.
The current gen XBOX already has hardware support for HDTV at 720p and 1080i. The only problem is that most of the games don't support the higher resolutions. Still, on all games you can play in 480p instead of 480i if you have a HDTV (or a display capable of progressive scan).
Rather, they should be liable for the actions of the corporation when an illegal act is performed.
If this was the case, you wouldn't find anyone who would take a job as CEO. If an employee commits a crime without the consent or knowledge of the CEO, should the CEO really go to jail? CEOs can already be charged criminally if they permit employess to commit crimes on behalf of the corporation.
If I RTFA correctly, private industry wants to build a road using private funding in the hopes of making a profit from the toll revenue. How does this contradict conservative principles? In theory, conservatives are in favor of smaller government and privatization of government services whenever it makes sense. This sounds exactly like the kind of project conservatives would support.
If the proposal was to increase property taxes on Texas's wealthiest land owners to build something nobody wants in order to create a few jobs for the people building it, then you can bet that conservatives would be against it. You might find a few liberals who would support it, though.
TFA is short on details. I have a hard time believing the tin-foil hat crowd's analysis that Tenet is advocating restricting internet access to approved users who take security seriously. I don't think anyone would support mass restrictions of internet access. After all, the article said he advocated measures taken by the private-sector to improve security.
Since are no real deatils in the article, I guess you can assume whatever you want. However, my guess is that he's probably talking about stuff that a lot of people on/. might support. For example, if a ton of spam, viruses, hacking attempts, etc is coming from a particular netblock because the owners of that netblock don't enforce basic TOS policies to prevent this sort of thing, maybe ICANN should take back that net block. Just the threat of ICANN reclaiming blocks of IPs would probably be enough to close up a lot of open proxies and mail servers.
Then again, maybe he wants to eliminate internet access for everyone without a government security clearance. There's no way to know since by his own admission, the writer of the article didn't hear the speech since the press wasn't invited. I wouldn't get too worked about about a vague article written from second-hand information about a speech given by a guy who no longer holds any power.
If you're the government, that's also just fine: less chance of Joe Sixpack retiring early on a long-term capital gain
When companies grant stock options to employees, they typically grant non-qualified (also called non-statutory) options. In most cases, the fair market value of these options can't be readily determined since they're not traded on an exchange and they typically come with restrictions like vesting schedules and non-transferability clauses. As a result, you don't pay any taxes on them when they are granted. However, when you exercise the options, the money you make gets treated as income not a long-term capital gain. See IRS publication 525. So Uncle Sam isn't unhappy. In fact, if the options you exercise are worth a lot of money, it will probably push your income into the 35% tax bracket... much higher than the 15% tax rate for long-term capital gains.
And just to add another data point, my employer grants options to about 30% of all employees each year based on individual performance. The number of options vary based on your position and performance level. Since the economy has improved, they've also started offering cash bonuses again to go along with the options. As with options, cash bonuses vary based on position and job performance.
Just splitting the electoral votes without any other changes might have some unintended consequences. The way the law currently reads, if no one obtains a majority of the electoral votes, the election gets decided by the House of Representatives. If you split the vote in all states, it would be much more likely for a 3rd party candidate to get enough electoral votes that no one obtains a majority. This probably would have happened when Perot ran in 1992. IIRC, he got about 20% of the popular vote the first time he ran with Bush and Clinton both close to 40% (Clinton slightly higher than Bush). However, because most states don't split electoral votes this was never an issue. I don't think Perot got any electoral votes (if he did, it wasn't more than 1 or 2).
I don't argue that splitting electoral votes might be a good idea if we remove the 50% majority requirement. I only point out that we would need a constitutional amendment to overhaul the current electoral system. Aside from it being unfair for some states to split votes and not others, if several states decide to start splitting electoral votes like it was proposed in Colorado, we may end up needing a constitutional amendment anyway. My guess is that the amendment would come the day after the House of Representatives picks the president.
Again, I'm not disagreeing with you, and I know that you weren't proposing to split electoral votes in some states and not others. I'm merely pointing some facts on this issue that are not often discussed.
If they aren't citizens, they shouldn't have any effect on gov't policy, should they?
Non-citizens shouldn't determine gov't policy, but I believe it is a good idea for states to have a voice proportional to the total population of the state. Even though non-citizens can't vote, they are still provided government services. By basing representation on total population instead of number of voters, it helps ensure that government services are apportioned based on the needs of the population, not just the needs of voting citizens.
If there are errors, then it's not clear - those errors could change the electoral vote anyway
The election can remain unclear until the electoral college actually votes. When the vote is taken, the electors must either vote for a candidate or refuse to vote. If no candidate obtains a majority, the House of Representitives chooses from the top 3 candidates. If no majority is obtained there, the Senate choses from the top 2 candidates. If there is a tie in the senate, the current Vice President breaks the tie. By law, votes by the House and Senate if needed are done after the electors votes are read by the president of the Senate on January 6 -- 14 days before the president is sworn in on January 20. People might not be happy about the outcome. Questions may remain over who got the most votes. However, we will always have a clear winner by January 6. Without the electoral collage, you could have a missing box of ballots show up in February that changes the outcome of the election.
There are a number of good reasons to keep the electoral college. Here are a few that come to mind. I'll leave out the ones about ensuring representation for rural states, since that one usually comes up in the discussion.
It isolates voting irregularities to a single state. This can be important. For example, if Diebold voting machines showed 3 billion people voted in Montana, it wouldn't have a drastic effect on the outcome since Montana only has 3 electoral votes.
It balances differences in voter turnout. New York is roughly twice the size of North Carolina. However, lets assume that New York gets hit by thunderstorms and has massive flooding on election day making it less convenient for people to vote. As a result, New York might have 30% voter turnout while North Carolina might have 60% voter turnout. This would mean North Carolina would have roughly the same representation as New York -- a state twice its size. The electoral college reduces the impact of weather, disasters, and even regional voter apathy on the final election results.
Not everyone that lives in a state may be eligible to vote because they may not be citizens. If a state has a large immigrant population, it is important the state's interests are represented in proportion to its size even though many of its residents may be unable to vote. The electoral college ensures this since electoral representation is determined based on raw population data from the census. A nationwide popular election would short-change states with lots of immigrants, or lots of children, or any other sizeable block of ineligible voters.
The electoral college ensures elections will always have a definite outcome. Even in 2000 when election results were unclear and court challeges delayed the outcome, the electoral college ensured we would eventually get a result that could not be legally disputed. Even if Gore had continued the court challenges and things were undecided until the day the electors cast their votes, once the electors voted, the outcome would be definite. By having the votes of a few hundred electors chosen by the states determine the final outcome, there is no room for errors in voting or tabulation. It is always clear how each of the electors vote.
I agree that some of the things included in the report are probably not "terrorism", and clearly not all of the terrorist acts involed muslim terrorists. The point I was trying to make is that terrorism is not a problem that is unique to the United States and its allies as many people have been led to believe. Al-Qaeda doesn't have a monopoly on terrorism. Terrorism is a tactic used all over the world to achieve a wide variety of political and social objectives. Anyone who harms innocent people to use fear as a bargaining chip is a terrorist. Even if you're being oppressed, that's never a justifiable recourse.
You dont see every country in the world being attacked by militant islamic extremist foreigners now do you?
Not every country, but there are a lot of them. Here are a few terrorist attacks from 2003 (the 2004 report isn't out yet). I excluded attacks on Americans, British, and Jews, because everyone knows we are evil and deserve it. I also left a lot of others out because I got tired of typing. All told, there were 208 significant terrorist attacks in 2003 resulting in 625 deaths and 3646 injuries. None of them occurred on U.S. soil. (source: Patterns of Global Terrorism 2003)
2/25/03 - Venezuela - 2 bombs explode simultaneously at spanish and columbian embassies. 1 Columbian and 3 Venezualans killed.
They don't discharge anywhere near that fast just sitting there unconnected, because the R is effectively infinite. Resistivity of dry air is *very* high.
That's true if you're talking about an ideal capacitor, but real electrolytic capacitors have some internal leakage current, so R can never be infinite. Without looking at a datasheet, I'm not sure if that internal leakage is equivalent to 10Mohms, but it's probably in the ballpark.
The voltage (the main item of concern) decreases by the function V=emf*(e^(-t/RC))... In other words, electric potential across the terminals decreases exponentially.
But if the capacitor is disconnected then the R in your equation is going to be a huge number. The voltage will still decay exponentially, but because the time constant is so big, it can take a while. As an example, if R is 10 megaohms and C is 100 microfarads, then it will take 1000 seconds for the capacitor to discarge to 37% of its starting voltage. So in this example, if the capacitor was initially charged to 600V, then 15 minutes after it was disconnected it would still have well over 200V left to shock you with.
Windows has JScript and VBScript built in. Last time I checked, Windows CDs also come with qbasic if you bother to copy it off. On top of that, anyone with an internet connection can download Perl or Java. DJGPP is a very good C/C++ compiler for DOS enviornments and is totally free. Borland gives away an old version of Turbo C for anyone to download on their website. Add Cygwin to the mix, and you've got another whole set of totally free compilers and assemblers that run under windows. I could keep listing, but you get the point.
Now... compare that to when I was younger. The only "free" langauges I had for DOS/Windows 3.1 was qbasic and debug (and writing assembly in debug was a bit over my head when I was a kid). There were also a few shareware/freeware assemblers that were farily good, but I never found any good free compilers for high-level langauges (and I did look). Since I wan't a rich kid, langauges like C and Pascal were inaccessable to me. Any kid who wants to learn how to program today is far better off than I was when I was a kid.
But most of the spectrum alloted for TV is still unused. Since most people do have cable or satellite, there isn't much incentive for corporations to invest the $$ to build a bunch of new broadcast towers for more over-the-air TV channels. If anything, the number of over-the-air channels will probably decrease, not increase. Because of the FCC's over-allotment, most of the UHF spectrum, and in most markets, much of the VHF spectrum is just going to waste. Why not compress the alloted spectrum into a few channels and free up the rest for other uses? Stations might have to move to different channels, but you'd still have your free TV.
It's McDLT. It stood for McDonald's Lettuce & Tomato. It came in a special container with the meat and half the bun on one side that stayed warm and the top of the bun with the lettuce and tomato on the other side to keep it cool.
BTW, you left out one small detail in the song. It should go "...a coffee, decaf too, a lowfat milk..." And yes, it is pathetic that I remember this stuff.
Well, for one thing, the processor isn't the only thing that needs a clock. Remember, the data bus needs to be synchronized to the processor clock too. I suppose you could put an oscillator and clock driver in the same package as the processor and have it drive the whole system, but think what that would do. Want to build a laptop that runs at a slower clock when idle to save power? Too bad. The clock is fixed in the processor. Want to make a motherboard with 2 processors? Guess you'll need a bunch of extra hardware to handle the differences in clocks since there's no way to use a common clock for both processors.
There's a bunch of compelling reasons not to embed the oscillator in the processor package. However, the only compelling reason I can see for putting it inside the processor package is to frustrate all the evil overclockers.
You shouldn't be able to patent processes that are obvious given current prior art. If you look at the diagram in the article, what's being proposed isn't anything particularly unique or new.
According to the diagram, it looks like they use an input clock to drive a counter. Then, after a set number of cycles of the internal crystal oscillator, you look at the value of the counter. If it's above a certain number, you know the input clock is too fast (somebody is overclocking it).
This is EXACTLY how a frequency counter works. Only frequency counters do some extra math so they can display the frequency in Hz or MHz, or whatever is appropriate. This is a simpler case because you're only concerned with crossing a set threshold.
So really, what you have is a patent for a design that has been around as long as crystals and flip-flops existed. The only thing that's really new here is that they're using it to prevent people from overclocking their processors. In my opinion, you shouldn't be able to get a patent for that. But what do I know? I didn't think Amazon should have been able to patent a one-click checkout even if they were the first ones to do it.
Slashdot Mirror
It seems to me that the problem of memory fragmentation could be addressed at some level by the operating system by giving each process its own 32-bit virtual address space. By giving each process a full 32-bit address space, it's unlikely that a process could sufficently fragment a 32-bit address space to the point where large allocations can no longer be made. In a 64-bit address space, it would be nearly impossible. Of course, this scheme would impact performance as it would require the operating system to do a TLB flush after each context switch. It would also require more memory to map virtual pages to physical memory differently for each process. Still, it seems like it could be a workable solution for most memory fragmentation problems. Is there something I'm missing?
I understand your argument that 9/11 might just be an isolated statistical spike that doesn't really indicate a change in the overall death rate caused by terrorist acts against Americans. On the other hand, it might mark the start of a period where there are significantly more terrorism related deaths than the historical average. My opinion is that 9/11 was a signigicant spike in a general upward trend, but not enough time has elapsed to tell for sure.
Roughly 3000 people died from the terrorist attacks on 9/11. If those deaths had been spread throughout the United States over the course of a year, then you might have a point... but they weren't. On average there are about 6500 deaths in the United States every day. Even looking at a national level, an extra 3000 deaths in one day is significant. If you only look at New York, then the spike in the death rate is huge. If one morning there was a giant car crash in downtown Manhattan that killed 3000 people and caused $20 billion in property damage, would you consider that event negligible?
Running as root is dangerous, but is more dangerous than the average home user is used to? Probably not. The average user probably runs windows from a single user account with admin rights. For most people, the recycle bin is the only protection from stupid mistakes.
Anybody know how this works? I've got a gen 4 iPod and it doesn't act like a normal USB storage device. On Windows PCs, the iPod doesn't show up as a drive unless iTunes has been installed. A little investigation shows iTunes is bundled with a driver for the iPod and installs an iPod service. I admit that I haven't actually tried to boot from the iPod, but I can't see how it would work. Does anyone know if the iPod mini is different in this respect and functions as a standard USB storage device?
My personal experience has been that windows driver support is still much better than linux, although linux driver support is much better than it used to be. That's not suprising since most hardware manufacturers are going to develop and test to their windows drivers since that probably accounts for 90% of the systems their hardware will be used in.
As an example, I recently installed linux on a spare windows box I had with the intention of making it a file server. The motherboard had an ITE8212 raid controller I was planning to use as an additional IDE interface (without RAID), but it wasn't supported. Some googling eventually turned up a kernel patch that added the driver. After recompiling the kernel, the drives did show up, but if I tried to access them the system would hang. I gave up and got a couple of cheap IDE cards with the SiI680 chipset which is supported by the kernel. With the driver supplied with the 2.4.28 kernel, only drives attached to the first controller showed up even though lspci showed both controllers were detected. I then upgraded to the 2.6.10 kernel which uses a different driver for that chipset and drives attached to both controllers showed up. However, if I tried to read or write much data to the drives, I got DMA errors and the driver would reset and enter PIO mode. Sometimes it would even hang the system. All of these cards worked flawlessly in Windows. I've been trying to get something to work under linux for two weeks and still no luck.
I'm about to buy different IDE cards again. If anybody knows of a card that will definitely work, let me know. I don't want to try a fourth set of cards. This is starting to get expensive.
The current gen XBOX already has hardware support for HDTV at 720p and 1080i. The only problem is that most of the games don't support the higher resolutions. Still, on all games you can play in 480p instead of 480i if you have a HDTV (or a display capable of progressive scan).
If the proposal was to increase property taxes on Texas's wealthiest land owners to build something nobody wants in order to create a few jobs for the people building it, then you can bet that conservatives would be against it. You might find a few liberals who would support it, though.
Since are no real deatils in the article, I guess you can assume whatever you want. However, my guess is that he's probably talking about stuff that a lot of people on /. might support. For example, if a ton of spam, viruses, hacking attempts, etc is coming from a particular netblock because the owners of that netblock don't enforce basic TOS policies to prevent this sort of thing, maybe ICANN should take back that net block. Just the threat of ICANN reclaiming blocks of IPs would probably be enough to close up a lot of open proxies and mail servers.
Then again, maybe he wants to eliminate internet access for everyone without a government security clearance. There's no way to know since by his own admission, the writer of the article didn't hear the speech since the press wasn't invited. I wouldn't get too worked about about a vague article written from second-hand information about a speech given by a guy who no longer holds any power.
And just to add another data point, my employer grants options to about 30% of all employees each year based on individual performance. The number of options vary based on your position and performance level. Since the economy has improved, they've also started offering cash bonuses again to go along with the options. As with options, cash bonuses vary based on position and job performance.
I don't argue that splitting electoral votes might be a good idea if we remove the 50% majority requirement. I only point out that we would need a constitutional amendment to overhaul the current electoral system. Aside from it being unfair for some states to split votes and not others, if several states decide to start splitting electoral votes like it was proposed in Colorado, we may end up needing a constitutional amendment anyway. My guess is that the amendment would come the day after the House of Representatives picks the president.
Again, I'm not disagreeing with you, and I know that you weren't proposing to split electoral votes in some states and not others. I'm merely pointing some facts on this issue that are not often discussed.
It isolates voting irregularities to a single state. This can be important. For example, if Diebold voting machines showed 3 billion people voted in Montana, it wouldn't have a drastic effect on the outcome since Montana only has 3 electoral votes.
It balances differences in voter turnout. New York is roughly twice the size of North Carolina. However, lets assume that New York gets hit by thunderstorms and has massive flooding on election day making it less convenient for people to vote. As a result, New York might have 30% voter turnout while North Carolina might have 60% voter turnout. This would mean North Carolina would have roughly the same representation as New York -- a state twice its size. The electoral college reduces the impact of weather, disasters, and even regional voter apathy on the final election results.
Not everyone that lives in a state may be eligible to vote because they may not be citizens. If a state has a large immigrant population, it is important the state's interests are represented in proportion to its size even though many of its residents may be unable to vote. The electoral college ensures this since electoral representation is determined based on raw population data from the census. A nationwide popular election would short-change states with lots of immigrants, or lots of children, or any other sizeable block of ineligible voters.
The electoral college ensures elections will always have a definite outcome. Even in 2000 when election results were unclear and court challeges delayed the outcome, the electoral college ensured we would eventually get a result that could not be legally disputed. Even if Gore had continued the court challenges and things were undecided until the day the electors cast their votes, once the electors voted, the outcome would be definite. By having the votes of a few hundred electors chosen by the states determine the final outcome, there is no room for errors in voting or tabulation. It is always clear how each of the electors vote.
I agree that some of the things included in the report are probably not "terrorism", and clearly not all of the terrorist acts involed muslim terrorists. The point I was trying to make is that terrorism is not a problem that is unique to the United States and its allies as many people have been led to believe. Al-Qaeda doesn't have a monopoly on terrorism. Terrorism is a tactic used all over the world to achieve a wide variety of political and social objectives. Anyone who harms innocent people to use fear as a bargaining chip is a terrorist. Even if you're being oppressed, that's never a justifiable recourse.
2/25/03 - Venezuela - 2 bombs explode simultaneously at spanish and columbian embassies. 1 Columbian and 3 Venezualans killed.
3/4/03 - Philippines - bomb explodes at airport. 21 killed, 149 injured.
3/20/03 - Lebanon - bomb explodes in apartment building. 2 killed, 9 wounded.
3/22/03 - Greece - bomb explodes at ATM.
3/24/03 - India - 11 men, 11 women, and 2 boys shot execution style by armed militants
3/25/03 - Serbia - 4 bomb attacks on UN interim administration
3/26/03 - Chile - bomb explodes at bank
3/29/03 - Greece - hand grenade tossed into a McDonalds
3/31/03 - Cuba - plane carrying 46 passengers hijacked
4/2/03 - Philippines - bomb explodes on passanger warf. 16 killed, 55 wounded.
4/5/03 - Lebanon - two bombs explode at restaurant. 10 wounded. undetonated C-4, TNT, and gas containers found.
4/8/03 - Algeria - 1 Swede and Dutch citizen kidnapped
4/11/03 - Algeria - 2 Austrians kidnapped
4/12/03 - India - multiple grenade attacks kill 1, wound 43.
4/12/03 - Venezuala - C-4 bomb explodes at OAS office
4/14/03 - France - militants set fire to car and destroy restaurant
4/15/03 - Turkey - bombs explode at 2 different McDonalds. 1 injured.
4/22/03 - India - bomb explodes at dairy. 6 killed, 12 wounded.
4/25/03 - India - bomb explodes at courthouse. 3 killed, 34 wounded.
5/5/03 - India - bomb and grenade attacks kill 1, injure 26.
5/16/03 - Morocco - 5 bombs explode simultaneously damaging Belgian consulate. 33 killed, 101 wounded.
6/4/03 - Belgium - Letters found containing the nerve agent adamsite. 10 hospitalized.
6/9/03 - Peru - 71 workers of an Argentine company kidnapped
6/17/03 - Italy - bomb explodes in front of spanish school
6/18/03 - France - militants destroy 2 villas with bomb blasts
6/26/03 - Kenya - aid workers attacked with hand grenades
7/3/03 - Columbia - 5 swiss citizens kidnapped
8/5/03 - Indonesia - bomb explodes in front of hotel 12 killed 149 wounded. Al-Qaida claims responsibility.
8/8/03 - Spain - mail bomb sent to Greek consulate
10/5/03 - Malaysia - 3 Indonesians and 2 Filipinos kidnapped. 1 escaped, 4 found executed.
11/11/03 - Greece - bomb found outside bank
They don't discharge anywhere near that fast just sitting there unconnected, because the R is effectively infinite. Resistivity of dry air is *very* high.
That's true if you're talking about an ideal capacitor, but real electrolytic capacitors have some internal leakage current, so R can never be infinite. Without looking at a datasheet, I'm not sure if that internal leakage is equivalent to 10Mohms, but it's probably in the ballpark.
But if the capacitor is disconnected then the R in your equation is going to be a huge number. The voltage will still decay exponentially, but because the time constant is so big, it can take a while. As an example, if R is 10 megaohms and C is 100 microfarads, then it will take 1000 seconds for the capacitor to discarge to 37% of its starting voltage. So in this example, if the capacitor was initially charged to 600V, then 15 minutes after it was disconnected it would still have well over 200V left to shock you with.
Now... compare that to when I was younger. The only "free" langauges I had for DOS/Windows 3.1 was qbasic and debug (and writing assembly in debug was a bit over my head when I was a kid). There were also a few shareware/freeware assemblers that were farily good, but I never found any good free compilers for high-level langauges (and I did look). Since I wan't a rich kid, langauges like C and Pascal were inaccessable to me. Any kid who wants to learn how to program today is far better off than I was when I was a kid.
But most of the spectrum alloted for TV is still unused. Since most people do have cable or satellite, there isn't much incentive for corporations to invest the $$ to build a bunch of new broadcast towers for more over-the-air TV channels. If anything, the number of over-the-air channels will probably decrease, not increase. Because of the FCC's over-allotment, most of the UHF spectrum, and in most markets, much of the VHF spectrum is just going to waste. Why not compress the alloted spectrum into a few channels and free up the rest for other uses? Stations might have to move to different channels, but you'd still have your free TV.
It's McDLT. It stood for McDonald's Lettuce & Tomato. It came in a special container with the meat and half the bun on one side that stayed warm and the top of the bun with the lettuce and tomato on the other side to keep it cool. BTW, you left out one small detail in the song. It should go "...a coffee, decaf too, a lowfat milk..." And yes, it is pathetic that I remember this stuff.
There's a bunch of compelling reasons not to embed the oscillator in the processor package. However, the only compelling reason I can see for putting it inside the processor package is to frustrate all the evil overclockers.
According to the diagram, it looks like they use an input clock to drive a counter. Then, after a set number of cycles of the internal crystal oscillator, you look at the value of the counter. If it's above a certain number, you know the input clock is too fast (somebody is overclocking it).
This is EXACTLY how a frequency counter works. Only frequency counters do some extra math so they can display the frequency in Hz or MHz, or whatever is appropriate. This is a simpler case because you're only concerned with crossing a set threshold.
So really, what you have is a patent for a design that has been around as long as crystals and flip-flops existed. The only thing that's really new here is that they're using it to prevent people from overclocking their processors. In my opinion, you shouldn't be able to get a patent for that. But what do I know? I didn't think Amazon should have been able to patent a one-click checkout even if they were the first ones to do it.