It varies by state. Where I live (Wisconsin), you can get a learner's permit at 15.5 years old (meaning you can drive only with an adult), and a license at 16. However, this requires having a sponsor (usually a parent) and taking a drivers' education class - at 18, those requirements are removed. Also, after I got my license, they started a "graduated licensing" system, where a new teen driver doesn't have complete driving privileges: specifically, they can't drive between midnight and 5 AM (except between home, school, and work), and they can only have one passenger in the car (excluding family or responsible adults).
It seems to work, for the most part - and you can bet there are a lot of relieved parents who no longer have to worry about driving their kids to school activities. My mom was thankful when I got my license, because my dad no longer had the opportunity to work on Saturdays while I was playing in a youth orchestra. My dad was not so happy, as it meant that he had to do yardwork and similar chores.
...allows us to walk in on voting day. Not only that, but if I'm registered, I can say "Yeah, he's with me!" and my friend gets registered on the spot too - no proof of address needed. It's like going to a popular nightclub. Except all the employees are over 55, and you're in an elementary school gym.
People do not vote on the laws directly. Rather, they elect a body of people who then vote on the laws. It's fine to like a different system, but you have to understand that the US is a Republic and is working as intended.
For a few years, I've thought it would be interesting to augment the system to allow individual voters to have input. For example, the representatives elected from your district will cast their votes as they do today. However, if you disagree with one of your representatives' votes, you can cast your own in a different manner, effecting a minute change in the vote totals (ie a fractional vote - a politician representing 10,000 people would then have 0.9999 of his vote in one category, and the other 0.0001 would be representing your dissenting voice). So, if something unpopular comes along (like telecom immunity), people might actually be able to make a difference (though likely not when the spread is more than just a couple votes).
However, my thought of this ends at "interesting." I can think of approximately ten thousand reasons why it would never work. For example, a readily available system for performing this voting would be needed (eg the Internet), however if we can't trust a Slashdot poll, how could we ever trust real voting? There's the usual uneducated voter argument (though maybe you could do something like divide the politician's vote according to the number of votes in the last election among all candidates - if you don't vote, you can't participate later in the process, either). And it's not likely that it would be all that effective - who's going to take the time to read the legislation if there's not a good chance of being able to make a difference? And without enough people doing so, there's no way that change can come about.
My mom likes the Quakers' idea; every decision must be unanimous. People will quickly realize that to get anything done, you have to be fair and compromise.
My plan: buy a small island, secede from whatever country, and start my own. Now accepting applications from those interested in funding my vision!
Gotta put in the requisite "why is this news?" reply... So some website registers the most hits on its web pages of netbooks, and suddenly Slashdot declares that "Netbooks move up in notebook rankings"? I'm sure it's a valid headline, but at least back it up with a real ranking (eg sales numbers), and not the number of hits on a website. By that metric, Rick Astley should be this year's top artist.
Re:Done before, using different sensory organ
on
Seeing With Your Skin?
·
· Score: 2, Informative
...and a project (also from the UW) involving several guys I know, called Visual Taste does that as well. There are pictures and videos, if the average slashdot reader can be troubled to follow the link...
No, it's not. The certificate has a small tear taken out of it. The other IDs didn't look to have fared quite as well, but also don't show any obvious burn marks - they're torn, tattered, and dirty. A year of exposure to the elements probably caused all the damage you see to those papers and the bills.
More plausible: He was able to get the plane on the ground, probably destroyed it in the process, suffered significant injuries, and was unable to get help. He may have left some items behind, dropped them, or perhaps his body was disturbed by wildlife. I expect they'll find more evidence upon searching more thoroughly.
It's 3 partial cycles, but in terms of battery lifetime, it's roughly equivalent to a full discharge-charge cycle. For example, see the second from last paragraph on Apple's battery site.
Others have given more technical details, but for a real-world example; the iPod Classic charges to 80% in 2 hours, and 100% in 4. The other iPods are the same, except the nano charges to 80% in 1.5 hours and 100% in 3. Any lithium ion battery should be charged in a similar manner.
You can find more information about charging LiIon batteries here.
Sure, Apple's decisions aren't always popular, but that doesn't mean that the iPhone is going to fade away into the sunset. They have an amazing amount of brand power (pronounced: "fanboys"). And they have a pretty good product from the point of view of the consumer - and that's where the money comes from.
The iPod is an even more closed platform (developers by invitation only!), yet it has been wildly popular. And this is in the face of a similar market - a late entry, a more expensive product, and countless others trying to imitate or improve upon it.
That's not to say that the iPhone is a guaranteed success, either. But so far, it has done pretty well - I doubt any single new phone will outdo it any time soon. On the other hand, an extension to an existing line (like from RIM) or an avalanche of Android phones might bury it. For the short term, I see no reason to discount the iPhone.
Having tons of liquified CO2 sitting around does not sound like a long term solution.
So maybe we should plan to reverse these reactions we've been relying on for energy for the last few centuries? All sorts of fuel exist, and don't really seem to have a big effect on the environment - so let's take our CO2 and a bunch of energy, and produce fuel. It can be safely stored away where we got it from, until somebody forgets why we did that in the first place, and they go burn it all again.
The disadvantage - this would require us to make not just all of our current energy consumption green, but also a significant percentage above what we consume in order to reverse the damage done. So this might be a long term solution, but it's a long time away from being feasible.
But you know damned well that 99% of the population is going to say: "Ok, great! We've got these CO2 filter thingies now, so the problem is solved. Now where's the keys to my Hummer?"
And the better solution is never going to come along.
On the other hand, their kids will see this, and many of them might just realize that there's a better way (especially if we teach them that there is!). Breaking habits is very difficult. You won't get most of the population to change their ways any time soon - the real change will happen over generations.
I wouldn't write that technique off completely... There might be times when it could be quite effective. Like if function A calls one of 10 different functions (call them 1 through 10), and each one of those might call function B, possibly multiple times. Now, there might be a case where function B wants to return all the way to function A. Do you really want to add if statements to each of functions 1 through 10, possibly in multiple locations of each, to check B's return value, and possibly then return to A? I could envision using an exception here instead, as long as performance is not critical, or the event is uncommon.
There are other ways one might implement this; say implement functions 1 through 10 as a single function, and the parts that are different could be implemented, say, as case statements. That significantly reduces the number of places you'd have to check the return value, but also gives you one very large function. I'm sure there are many more ways one could reorganize the code, perhaps avoiding the need to pop up to function A at all, depending on the specific algorithm.
But, the point is, don't dismiss using an exception just because it doesn't "seem" exception-like (eg, an I/O error). According to Sun:
An exception is an event, which occurs during the execution of a program, that disrupts the normal flow of the program's instructions.
OK, so that leaves some more questions - like what's the "normal flow" of a program? In any case, there's nothing here to say that you can't use an exception to break out of deep nesting. If it works, use it. If there's a better way (in terms of maintenance, speed, readability, code size, coding ease, etc.), use that instead. Often, you'll find that absent a specific guideline ("speed is more important than anything else"), several options might work well. Don't discount a technique just because it's not the "usual" thing you're used to seeing.
On the other hand, if you're not used to seeing it, that's a mark against it for readability and maintenance - so it had better have some benefits in the other categories.
I agree with the sibling post that recumbents are fine on hills - at least, anecdotally. I've only tried them briefly, but I biked across the US about 11 years ago (with an organized group), and one of the faster riders rode a recumbent. This included more than a few hills, such as the Rockies (which, incidentally, aren't too bad, since they're not nearly as steep as shorter hills tend to be). Of course, two of the slower riders also rode recumbents, and did have a lot of trouble on hills, so it does take some conditioning, as the sibling post also notes.
In any case, I never stand while pedaling up a hill - I've always felt that it's horribly inefficient. It might be OK for a quick boost of power, but it seems to waste a lot of energy in lateral and other motion that doesn't go into the pedals. Plus, it usually seems to me like there's a dead spot at the bottom of the pedal motion - after falling with the pedal, it takes a tiny bit of time to shift your weight to the other pedal and push on it instead. When just on the seat, I can push much more consistently on the pedals, because I can use the seat to balance myself, even if I'm putting a lot of my weight into the pedals anyway. I've found that I can almost always beat somebody standing on the pedals up a hill by either downshifting, putting a little more power into it, or both. I imagine that, on a recumbent, you'd need to be able to press into the back of the seat to get as much power, since you don't have gravity acting on your weight to help. But you should still be able to get up a hill just fine.
I was recently cleaning out some really old stuff, and came across a physics lab from high school. We measured our power output by timing how long it took to run or walk up a flight of stairs; my peak was running 2 steps at a time, when I reached 865 Watts! Granted, this was only sustained for 3.5 seconds, but just imagine if the human body could sustain that amount of output...
I was thinking the exact same thing. I don't know how many people work in their headquarters, but it's quite possible that all the technology (PCs, projectors, printers, etc.) OUTSIDE the datacenter was worth more than the servers, routers, etc. inside (plus the infrastructure, which is also quite valuable).
Of course, in either case, the real estate is probably worth the most. And, real estate won't depreciate as quickly as the data center will - in a few years, the servers and routers will be too slow and not have enough capacity; the cooling capabilities will probably not be enough to match the power density of new servers; and the power distribution network might not be up to the task of feeding all the servers that could fill that data center.
Please remember the quantum demonstration that whether a quantum packet is a particle or a wave depends on how you measure it.
No, it doesn't. Sometimes, matter behaves like we expect a wave to. Sometimes, it behaves like we expect a particle to. It's just convenient to think of its behavior that way, because waves and particles are things we understand.
There was a good quantum explanation linked from Slashdot (I think) a while ago. I think he has finished the series of explanations at this point - they're quite good, but quantum can be very difficult to understand.
And, now that we've reached about -10 offtopic, I think I'll stop writing.
Okay, I'm thinking of a number between 0000 0000 0000 0000 and 9999 9999 9999 9999. Once a psychic gets that right, we can move on to the expiration date.
This type of test is fairly commonplace in certain engineering fields, and should be. Specific technical knowledge, trivial to test for, are much more important and much simpler to test for in engineers and technicians than they are in professional workers.
Having gone through a good number of engineering interviews, I second this. My various interviews (some lasting up to a full 8 hours) have all included questions that I'd expect from some of the early undergrad-level courses I took (even after getting a Master's degree... good student does not equal good employee). Often, they ask a couple more-advanced questions, but not always. In any case, they could easily put these questions on paper if they wanted - and I encountered one company that did, as a quick screening before they talk to you further.
Really, the companies may be cheating themselves if they give you a written test; some things are definite and can be written down, but many of the questions I've been asked require a dialog. Maybe the interviewer doesn't give you enough information to solve the problem, and they want to make sure you ask questions, and ask the right questions. Or perhaps there are several possible solutions, so the interviewer might encourage you to find another after your first solution. In any case, these are probably far better than a test, though harder to score.
Bottom line: stop whining, answer a few questions, and try to avoid making a written test if you ever have to interview somebody. It's far better to find out how they think than whether they have all the answers stored in their brain. Answers can be found with Google, so an employee that can think about the problems (and has internet access!) will do better than one that just has the answers.
So if you have code that isn't SIMD-able you are really only using 1/32 available threads per unit of branching code.
In addition to what's already been said, there are other techniques that can be used when your code does in fact need to branch. For example, you can take BOTH paths, and then later pick the result from the path you want. This is common when you have lots of parallel hardware, whether made for you in a GPU, or in hardware you're designing yourself, like an ASIC or FPGA. So if you have if( A ) {
Z = B + C; } else {
Z = B - C; }
then you have instructions (or hardware) that perform B+C, separate instructions (or additional hardware) that perform B-C, and then a special instruction (in a GPU) or a multiplexer (for an FPGA or ASIC) that selects one of those two results (based on A) and puts it in Z. (I just wanted to add another set of parentheses here because I didn't think there were enough, so you can ignore this side remark.)
In this way, you do most of your computation in parallel (assuming your computation is a little bit more complex than just an addition/subtraction), although you end up doing twice as much work. Obviously, there's a limit to how many different paths you can compute this way before you lose the advantage.
Actually, *both* parties pay in the US, at least for any plan I'm familiar with. In my case, I pay 20 cents to send or receive a text. If I send a text to somebody else with a pay-per-use text plan, then *they* pay 20 cents, too. So, AT&T has just gouged its two customers 40 cents for 140 bytes of data. That's $2,995.93/MB, and it's very nearly pure profit.
Sure, there's some overhead too that this doesn't account for, but it's entirely a profit center for the cell phone providers, much like land lines are for the more old-school telecom companies. Basically, they want everybody to pay $5 more per month, because that recurring revenue stream is a lot more valuable to them.
I applaud Kohl - as a Wisconsin resident, I'm proud of both him and our other Senator, Russ Feingold. They almost convince me that politicians don't have to be sleazy...
Yes, if you have access to the challenge/response pairs, then you can program that information into some RFID chip without their fancy PUF thingy, and it could masquerade as any of the RFID chips for which it has that data. It's dependent on the reader only asking one of the known challenges, of course. For that matter, if you have access to only the challenges (or if a programmer always seeds the random number generator with the same number, or something stupid like that), and can also get access to the RFID tag, you can query it with all of those challenges and build your own database.
The difficulty in directly cloning this chip is that, if you don't know what the challenges will be, it's nearly impossible. The fastest RFID rate I found (which is from a 2008 academic paper, so probably not actually implemented) was 4 Mb/s; typical rates seem to be hundreds of kilobits/sec. Even at the faster rate, you can only read 65536 64-bit responses per second (assuming you had full-duplex communications to send challenges at the same time, could generate responses as fast as they could be sent, and basically every other unreasonable assumption possible). At that rate, it would take almost 9 million years to gather every possible challenge/response pair for a single RFID tag.
So, if you're given one of these RFID chips and want it to be (fairly) secure, generate your own challenges (and gather the responses inside a Faraday cage, so nobody else can listen in). And make sure they're RANDOM. Heck, when verifying the chip, challenge it several times, and make it get them all right. And, most importantly, keep that challenge/response database secure. If the database is compromised, the only way to regain security is to take each of your RFID tags, verify that it is accurate (by a method other than RFID), and generate new challenge/response pairs, and delete the old ones.
More details can be found for the geekily-minded in their academic paper (PDF warning!).
Basically, it's a series of multiplexers. The challenge selects exactly what pair of paths through the multiplexers are taken, and the output is a 0 or 1 depending on which path is faster. Presumably, this then gets replicated or reused several times to make a multi-bit response. They show an LFSR in their diagram, but don't explicitly say what they use it for - my guess would be they initialize it with the challenge, then use it to generate the programming bits to select a path through the multiplexers.
So yeah, it's pretty difficult to manufacture a circuit that exactly matches it. And it would probably take too long to exhaustively try all challenges to discover what the responses are. However, I still see several possible weaknesses.
First, the challenge/response pairs that are stored (which are outside the RFID chip, used to verify that it is valid) must be selected randomly. If an attacker can reduce the number of possible challenges from 2^64 down to a much smaller number, it's no longer secure: he can interrogate the RFID chip for its responses to those challenges, and then program those into a new chip. It's not completely cloned, but as far as anybody can tell from the stored challenge/response pairs, it is identical.
Second, the paper shows that about 11 bits out of every 128 are different each time you use the *same* challenge with the *same* chip. To catch most false negatives with the fewest false positives (ie highest security possible), the threshold would have to be probably only 104 correct bits out of 128. (The same challenge with different chips is close to the ideal of 64 changed bits out of 128 total). Presumably, these numbers are approximately halved when using 64-bit challenges and responses. This makes the chip weaker than something that really has 2^64 combinations; you don't have to get all 64 bits right, you just have to get maybe 52 of them right. In the paper, they suggest a threshold of 96 correct bits - or presumably 48-bits with the 64-bit implementation. That effectively knocks a good 5 orders of magnitude off the number of possible responses.
Third, what's to stop somebody from figuring out the timing parameters of a particular RFID, and emulating the circuit? They say in the paper that they "scramble its output to thwart such 'model building' attacks." OK, how? Is this why the LFSR is in the design? Obviously, they're trying to prevent their competitors from copying their work, but are they also trying to get security through obscurity? We all know how well *that* works.
Fourth, the challenge/response pairs have to be stored securely. If an attacker can get them, it's game over. Considering most companies still haven't figured out how to secure their customers' credit card numbers, the only thing keeping an attacker at bay is a lack of motivation. Make the payoff good enough, and this is probably the weak point in the system that would be hacked first.
Fifth, if I'm a malicious supplier of RFID chips, I might be able to find two similar chips. I sell one to somebody else, and keep the second for my own malicious purposes. Since it doesn't have to be exactly identical (within a few bits is fine), and I can use the principles of the birthday attack, this shouldn't be a terribly difficult thing to do. Now, if I did my math right, a malicious supplier would have to buy around 83 million RFID chips to have a 50% chance of getting one pair that are considered to be matches, *if* the threshold is set at the most secure level possible. I'd bet a typical threshold would drop that by another order of magnitude or so. That's a lot of RFID tags, but given RFID's target (low-cost, high-volume), it's not so unreasonable.
The paper, like many involving an actual company, lacks a lot o
OK, so the reward is now $500, they can take the drive apart, there's no deposit, and it's been publicized on Slashdot.
The reward is still waaaaay too small. A normal fee would be well over $1,000. And that's only if your drive is broken. This probably requires specialized equipment beyond what most recovery companies have. So yeah, zeroing your drive is probably usually safe, as long as the NSA, FBI, or your company's Chinese competitors don't have an interest in it. Or a random Materials Science or Electrical Engineering student, they might be able to hack something together too.;-)
Probably not. Their current Civic Hybrid doesn't do any better than the Prius.
That's because the Civic is a different type of Hybrid than the Prius. The Prius is a full hybrid. It can run off either the engine, the motors, or both. The Civic does not have nearly as powerful of electric motors, so the gasoline engine is essential. A full hybrid design is much more expensive, but also yields better results. Personally, I bet they can beat the current Prius. Rumor has it, however, that the next major Prius revision (probably for the 2010 model year) will be a significant improvement.
Slashdot Mirror
It varies by state. Where I live (Wisconsin), you can get a learner's permit at 15.5 years old (meaning you can drive only with an adult), and a license at 16. However, this requires having a sponsor (usually a parent) and taking a drivers' education class - at 18, those requirements are removed. Also, after I got my license, they started a "graduated licensing" system, where a new teen driver doesn't have complete driving privileges: specifically, they can't drive between midnight and 5 AM (except between home, school, and work), and they can only have one passenger in the car (excluding family or responsible adults).
It seems to work, for the most part - and you can bet there are a lot of relieved parents who no longer have to worry about driving their kids to school activities. My mom was thankful when I got my license, because my dad no longer had the opportunity to work on Saturdays while I was playing in a youth orchestra. My dad was not so happy, as it meant that he had to do yardwork and similar chores.
...allows us to walk in on voting day. Not only that, but if I'm registered, I can say "Yeah, he's with me!" and my friend gets registered on the spot too - no proof of address needed. It's like going to a popular nightclub. Except all the employees are over 55, and you're in an elementary school gym.
People do not vote on the laws directly. Rather, they elect a body of people who then vote on the laws. It's fine to like a different system, but you have to understand that the US is a Republic and is working as intended.
For a few years, I've thought it would be interesting to augment the system to allow individual voters to have input. For example, the representatives elected from your district will cast their votes as they do today. However, if you disagree with one of your representatives' votes, you can cast your own in a different manner, effecting a minute change in the vote totals (ie a fractional vote - a politician representing 10,000 people would then have 0.9999 of his vote in one category, and the other 0.0001 would be representing your dissenting voice). So, if something unpopular comes along (like telecom immunity), people might actually be able to make a difference (though likely not when the spread is more than just a couple votes).
However, my thought of this ends at "interesting." I can think of approximately ten thousand reasons why it would never work. For example, a readily available system for performing this voting would be needed (eg the Internet), however if we can't trust a Slashdot poll, how could we ever trust real voting? There's the usual uneducated voter argument (though maybe you could do something like divide the politician's vote according to the number of votes in the last election among all candidates - if you don't vote, you can't participate later in the process, either). And it's not likely that it would be all that effective - who's going to take the time to read the legislation if there's not a good chance of being able to make a difference? And without enough people doing so, there's no way that change can come about.
My mom likes the Quakers' idea; every decision must be unanimous. People will quickly realize that to get anything done, you have to be fair and compromise.
My plan: buy a small island, secede from whatever country, and start my own. Now accepting applications from those interested in funding my vision!
Gotta put in the requisite "why is this news?" reply...
So some website registers the most hits on its web pages of netbooks, and suddenly Slashdot declares that "Netbooks move up in notebook rankings"? I'm sure it's a valid headline, but at least back it up with a real ranking (eg sales numbers), and not the number of hits on a website. By that metric, Rick Astley should be this year's top artist.
...and a project (also from the UW) involving several guys I know, called Visual Taste does that as well. There are pictures and videos, if the average slashdot reader can be troubled to follow the link...
No, it's not. The certificate has a small tear taken out of it. The other IDs didn't look to have fared quite as well, but also don't show any obvious burn marks - they're torn, tattered, and dirty. A year of exposure to the elements probably caused all the damage you see to those papers and the bills.
More plausible: He was able to get the plane on the ground, probably destroyed it in the process, suffered significant injuries, and was unable to get help. He may have left some items behind, dropped them, or perhaps his body was disturbed by wildlife. I expect they'll find more evidence upon searching more thoroughly.
It's 3 partial cycles, but in terms of battery lifetime, it's roughly equivalent to a full discharge-charge cycle. For example, see the second from last paragraph on Apple's battery site.
Others have given more technical details, but for a real-world example; the iPod Classic charges to 80% in 2 hours, and 100% in 4. The other iPods are the same, except the nano charges to 80% in 1.5 hours and 100% in 3. Any lithium ion battery should be charged in a similar manner.
You can find more information about charging LiIon batteries here.
s/cellphone/MP3 player/g ...etc.
s/iPhone/iPod/g
Sure, Apple's decisions aren't always popular, but that doesn't mean that the iPhone is going to fade away into the sunset. They have an amazing amount of brand power (pronounced: "fanboys"). And they have a pretty good product from the point of view of the consumer - and that's where the money comes from.
The iPod is an even more closed platform (developers by invitation only!), yet it has been wildly popular. And this is in the face of a similar market - a late entry, a more expensive product, and countless others trying to imitate or improve upon it.
That's not to say that the iPhone is a guaranteed success, either. But so far, it has done pretty well - I doubt any single new phone will outdo it any time soon. On the other hand, an extension to an existing line (like from RIM) or an avalanche of Android phones might bury it. For the short term, I see no reason to discount the iPhone.
Having tons of liquified CO2 sitting around does not sound like a long term solution.
So maybe we should plan to reverse these reactions we've been relying on for energy for the last few centuries? All sorts of fuel exist, and don't really seem to have a big effect on the environment - so let's take our CO2 and a bunch of energy, and produce fuel. It can be safely stored away where we got it from, until somebody forgets why we did that in the first place, and they go burn it all again.
The disadvantage - this would require us to make not just all of our current energy consumption green, but also a significant percentage above what we consume in order to reverse the damage done. So this might be a long term solution, but it's a long time away from being feasible.
But you know damned well that 99% of the population is going to say: "Ok, great! We've got these CO2 filter thingies now, so the problem is solved. Now where's the keys to my Hummer?"
And the better solution is never going to come along.
On the other hand, their kids will see this, and many of them might just realize that there's a better way (especially if we teach them that there is!). Breaking habits is very difficult. You won't get most of the population to change their ways any time soon - the real change will happen over generations.
I wouldn't write that technique off completely... There might be times when it could be quite effective. Like if function A calls one of 10 different functions (call them 1 through 10), and each one of those might call function B, possibly multiple times. Now, there might be a case where function B wants to return all the way to function A. Do you really want to add if statements to each of functions 1 through 10, possibly in multiple locations of each, to check B's return value, and possibly then return to A? I could envision using an exception here instead, as long as performance is not critical, or the event is uncommon.
There are other ways one might implement this; say implement functions 1 through 10 as a single function, and the parts that are different could be implemented, say, as case statements. That significantly reduces the number of places you'd have to check the return value, but also gives you one very large function. I'm sure there are many more ways one could reorganize the code, perhaps avoiding the need to pop up to function A at all, depending on the specific algorithm.
But, the point is, don't dismiss using an exception just because it doesn't "seem" exception-like (eg, an I/O error). According to Sun:
An exception is an event, which occurs during the execution of a program, that disrupts the normal flow of the program's instructions.
OK, so that leaves some more questions - like what's the "normal flow" of a program? In any case, there's nothing here to say that you can't use an exception to break out of deep nesting. If it works, use it. If there's a better way (in terms of maintenance, speed, readability, code size, coding ease, etc.), use that instead. Often, you'll find that absent a specific guideline ("speed is more important than anything else"), several options might work well. Don't discount a technique just because it's not the "usual" thing you're used to seeing.
On the other hand, if you're not used to seeing it, that's a mark against it for readability and maintenance - so it had better have some benefits in the other categories.
I actually prefer to read the discussion completely unnested, and from left to right. It requires a really tall monitor, but I enjoy the challenge.
You see, it's a lot like drifting...
I agree with the sibling post that recumbents are fine on hills - at least, anecdotally. I've only tried them briefly, but I biked across the US about 11 years ago (with an organized group), and one of the faster riders rode a recumbent. This included more than a few hills, such as the Rockies (which, incidentally, aren't too bad, since they're not nearly as steep as shorter hills tend to be). Of course, two of the slower riders also rode recumbents, and did have a lot of trouble on hills, so it does take some conditioning, as the sibling post also notes.
In any case, I never stand while pedaling up a hill - I've always felt that it's horribly inefficient. It might be OK for a quick boost of power, but it seems to waste a lot of energy in lateral and other motion that doesn't go into the pedals. Plus, it usually seems to me like there's a dead spot at the bottom of the pedal motion - after falling with the pedal, it takes a tiny bit of time to shift your weight to the other pedal and push on it instead. When just on the seat, I can push much more consistently on the pedals, because I can use the seat to balance myself, even if I'm putting a lot of my weight into the pedals anyway. I've found that I can almost always beat somebody standing on the pedals up a hill by either downshifting, putting a little more power into it, or both. I imagine that, on a recumbent, you'd need to be able to press into the back of the seat to get as much power, since you don't have gravity acting on your weight to help. But you should still be able to get up a hill just fine.
I was recently cleaning out some really old stuff, and came across a physics lab from high school. We measured our power output by timing how long it took to run or walk up a flight of stairs; my peak was running 2 steps at a time, when I reached 865 Watts! Granted, this was only sustained for 3.5 seconds, but just imagine if the human body could sustain that amount of output...
I was thinking the exact same thing. I don't know how many people work in their headquarters, but it's quite possible that all the technology (PCs, projectors, printers, etc.) OUTSIDE the datacenter was worth more than the servers, routers, etc. inside (plus the infrastructure, which is also quite valuable).
Of course, in either case, the real estate is probably worth the most. And, real estate won't depreciate as quickly as the data center will - in a few years, the servers and routers will be too slow and not have enough capacity; the cooling capabilities will probably not be enough to match the power density of new servers; and the power distribution network might not be up to the task of feeding all the servers that could fill that data center.
Please remember the quantum demonstration that whether a quantum packet is a particle or a wave depends on how you measure it.
No, it doesn't. Sometimes, matter behaves like we expect a wave to. Sometimes, it behaves like we expect a particle to. It's just convenient to think of its behavior that way, because waves and particles are things we understand.
There was a good quantum explanation linked from Slashdot (I think) a while ago. I think he has finished the series of explanations at this point - they're quite good, but quantum can be very difficult to understand.
And, now that we've reached about -10 offtopic, I think I'll stop writing.
Okay, I'm thinking of a number between 0000 0000 0000 0000 and 9999 9999 9999 9999. Once a psychic gets that right, we can move on to the expiration date.
This type of test is fairly commonplace in certain engineering fields, and should be. Specific technical knowledge, trivial to test for, are much more important and much simpler to test for in engineers and technicians than they are in professional workers.
Having gone through a good number of engineering interviews, I second this. My various interviews (some lasting up to a full 8 hours) have all included questions that I'd expect from some of the early undergrad-level courses I took (even after getting a Master's degree... good student does not equal good employee). Often, they ask a couple more-advanced questions, but not always. In any case, they could easily put these questions on paper if they wanted - and I encountered one company that did, as a quick screening before they talk to you further.
Really, the companies may be cheating themselves if they give you a written test; some things are definite and can be written down, but many of the questions I've been asked require a dialog. Maybe the interviewer doesn't give you enough information to solve the problem, and they want to make sure you ask questions, and ask the right questions. Or perhaps there are several possible solutions, so the interviewer might encourage you to find another after your first solution. In any case, these are probably far better than a test, though harder to score.
Bottom line: stop whining, answer a few questions, and try to avoid making a written test if you ever have to interview somebody. It's far better to find out how they think than whether they have all the answers stored in their brain. Answers can be found with Google, so an employee that can think about the problems (and has internet access!) will do better than one that just has the answers.
So if you have code that isn't SIMD-able you are really only using 1/32 available threads per unit of branching code.
In addition to what's already been said, there are other techniques that can be used when your code does in fact need to branch. For example, you can take BOTH paths, and then later pick the result from the path you want. This is common when you have lots of parallel hardware, whether made for you in a GPU, or in hardware you're designing yourself, like an ASIC or FPGA. So if you have
if( A ) {
Z = B + C;
} else {
Z = B - C;
}
then you have instructions (or hardware) that perform B+C, separate instructions (or additional hardware) that perform B-C, and then a special instruction (in a GPU) or a multiplexer (for an FPGA or ASIC) that selects one of those two results (based on A) and puts it in Z. (I just wanted to add another set of parentheses here because I didn't think there were enough, so you can ignore this side remark.)
In this way, you do most of your computation in parallel (assuming your computation is a little bit more complex than just an addition/subtraction), although you end up doing twice as much work. Obviously, there's a limit to how many different paths you can compute this way before you lose the advantage.
Actually, *both* parties pay in the US, at least for any plan I'm familiar with. In my case, I pay 20 cents to send or receive a text. If I send a text to somebody else with a pay-per-use text plan, then *they* pay 20 cents, too. So, AT&T has just gouged its two customers 40 cents for 140 bytes of data. That's $2,995.93/MB, and it's very nearly pure profit.
Sure, there's some overhead too that this doesn't account for, but it's entirely a profit center for the cell phone providers, much like land lines are for the more old-school telecom companies. Basically, they want everybody to pay $5 more per month, because that recurring revenue stream is a lot more valuable to them.
I applaud Kohl - as a Wisconsin resident, I'm proud of both him and our other Senator, Russ Feingold. They almost convince me that politicians don't have to be sleazy...
...but since carbon dating is all wrong, obviously this forest existed only 150 million years ago.
Wait, no, I've just been told that this fossilized forest was created for us to find, around 6,000 years ago. Just like carbon dating.
Yes, if you have access to the challenge/response pairs, then you can program that information into some RFID chip without their fancy PUF thingy, and it could masquerade as any of the RFID chips for which it has that data. It's dependent on the reader only asking one of the known challenges, of course. For that matter, if you have access to only the challenges (or if a programmer always seeds the random number generator with the same number, or something stupid like that), and can also get access to the RFID tag, you can query it with all of those challenges and build your own database.
The difficulty in directly cloning this chip is that, if you don't know what the challenges will be, it's nearly impossible. The fastest RFID rate I found (which is from a 2008 academic paper, so probably not actually implemented) was 4 Mb/s; typical rates seem to be hundreds of kilobits/sec. Even at the faster rate, you can only read 65536 64-bit responses per second (assuming you had full-duplex communications to send challenges at the same time, could generate responses as fast as they could be sent, and basically every other unreasonable assumption possible). At that rate, it would take almost 9 million years to gather every possible challenge/response pair for a single RFID tag.
So, if you're given one of these RFID chips and want it to be (fairly) secure, generate your own challenges (and gather the responses inside a Faraday cage, so nobody else can listen in). And make sure they're RANDOM. Heck, when verifying the chip, challenge it several times, and make it get them all right. And, most importantly, keep that challenge/response database secure. If the database is compromised, the only way to regain security is to take each of your RFID tags, verify that it is accurate (by a method other than RFID), and generate new challenge/response pairs, and delete the old ones.
More details can be found for the geekily-minded in their academic paper (PDF warning!).
Basically, it's a series of multiplexers. The challenge selects exactly what pair of paths through the multiplexers are taken, and the output is a 0 or 1 depending on which path is faster. Presumably, this then gets replicated or reused several times to make a multi-bit response. They show an LFSR in their diagram, but don't explicitly say what they use it for - my guess would be they initialize it with the challenge, then use it to generate the programming bits to select a path through the multiplexers.
So yeah, it's pretty difficult to manufacture a circuit that exactly matches it. And it would probably take too long to exhaustively try all challenges to discover what the responses are. However, I still see several possible weaknesses.
First, the challenge/response pairs that are stored (which are outside the RFID chip, used to verify that it is valid) must be selected randomly. If an attacker can reduce the number of possible challenges from 2^64 down to a much smaller number, it's no longer secure: he can interrogate the RFID chip for its responses to those challenges, and then program those into a new chip. It's not completely cloned, but as far as anybody can tell from the stored challenge/response pairs, it is identical.
Second, the paper shows that about 11 bits out of every 128 are different each time you use the *same* challenge with the *same* chip. To catch most false negatives with the fewest false positives (ie highest security possible), the threshold would have to be probably only 104 correct bits out of 128. (The same challenge with different chips is close to the ideal of 64 changed bits out of 128 total). Presumably, these numbers are approximately halved when using 64-bit challenges and responses. This makes the chip weaker than something that really has 2^64 combinations; you don't have to get all 64 bits right, you just have to get maybe 52 of them right. In the paper, they suggest a threshold of 96 correct bits - or presumably 48-bits with the 64-bit implementation. That effectively knocks a good 5 orders of magnitude off the number of possible responses.
Third, what's to stop somebody from figuring out the timing parameters of a particular RFID, and emulating the circuit? They say in the paper that they "scramble its output to thwart such 'model building' attacks." OK, how? Is this why the LFSR is in the design? Obviously, they're trying to prevent their competitors from copying their work, but are they also trying to get security through obscurity? We all know how well *that* works.
Fourth, the challenge/response pairs have to be stored securely. If an attacker can get them, it's game over. Considering most companies still haven't figured out how to secure their customers' credit card numbers, the only thing keeping an attacker at bay is a lack of motivation. Make the payoff good enough, and this is probably the weak point in the system that would be hacked first.
Fifth, if I'm a malicious supplier of RFID chips, I might be able to find two similar chips. I sell one to somebody else, and keep the second for my own malicious purposes. Since it doesn't have to be exactly identical (within a few bits is fine), and I can use the principles of the birthday attack, this shouldn't be a terribly difficult thing to do. Now, if I did my math right, a malicious supplier would have to buy around 83 million RFID chips to have a 50% chance of getting one pair that are considered to be matches, *if* the threshold is set at the most secure level possible. I'd bet a typical threshold would drop that by another order of magnitude or so. That's a lot of RFID tags, but given RFID's target (low-cost, high-volume), it's not so unreasonable.
The paper, like many involving an actual company, lacks a lot o
OK, so the reward is now $500, they can take the drive apart, there's no deposit, and it's been publicized on Slashdot.
The reward is still waaaaay too small. A normal fee would be well over $1,000. And that's only if your drive is broken. This probably requires specialized equipment beyond what most recovery companies have. So yeah, zeroing your drive is probably usually safe, as long as the NSA, FBI, or your company's Chinese competitors don't have an interest in it. Or a random Materials Science or Electrical Engineering student, they might be able to hack something together too. ;-)
Probably not. Their current Civic Hybrid doesn't do any better than the Prius.
That's because the Civic is a different type of Hybrid than the Prius. The Prius is a full hybrid. It can run off either the engine, the motors, or both. The Civic does not have nearly as powerful of electric motors, so the gasoline engine is essential. A full hybrid design is much more expensive, but also yields better results.
Personally, I bet they can beat the current Prius. Rumor has it, however, that the next major Prius revision (probably for the 2010 model year) will be a significant improvement.