It sure sounds to me like any interpreted language would infringe, or be prior art. Lisp had an interpreter all the way back in 1958. A bit more recently, the Bourne Shell did this. I can't find a patent number anywhere, but I doubt this predates Lisp and the Bourne Shell. Time for us to get even more pissed off than we did when Eolas beat Microsoft in the trial court. This really needs to be dealt with in Congress. Go write your Representatives and Senators. I am.
It's quite obvious that Condorcet Voting and Instant Runoff Voting are designed to help third parties. Since they eliminate the danger of helping your opposition by making a third party vote (the criticism of IRV in the linked article applies just as well to Condorcet voting, so it's really moot) we can expect LOTS of 3rd parties on the ballot. It's hard enough getting people to find out about 2 candidates. It's impossible and unreasonable to expect people to find out about 20. Both schemes can easily handle people casting single-shot votes, so they do not disenfranchise anyone as long as incomplete ballots are acceptable.
First, the inherent complexities of the ballot:
Since a voter could vote for 0 to n candidates in any order, an IRV ballot has SUM(i=0..n) Perm(n, i) = e * n! = THETA(n!) possible assignments.
Since a voter has n(n-1)/2 choices that can be 0,1, or don't care, there are 3^(n(n-1)/2) = THETA (sqrt(3^(n^2))) possible assignments. Since this is polyexponential, it dominates the THETA (n!) complexity of IRV.
Second, complexity of representing an individual ballot:
IRV has n identifiers that must be of size lg(n), so it can take THETA(n lg(n)) space to represent a full ballot.
Condorcet ballots can use implicit ordering, so the size of representing each decision is constant, making the cost of representing n(n-1)/2 decisions THETA(n^2).
Third, the decision complexity required to cast a single-shot ballot
IRV allows a voter to vote their preferred candidate first and then vote for no one else. While this decision itself is more complicated than a binary decision, a single-shot voter can go into the booth and make one simple action and not be disenfranchised.
Condorcet voting requires that a voter make (n-1) comparisons just to make their single-shot vote. Furthermore, they must make an additional (n-2) comparisons if there's a guy they REALLY want to lose. As of right now, there are a whole lot of "Anyone but Bush" people who will be voting for Kerry (possibly holding their noses) because it's a very simple way to do the best they can to make sure Bush is defeated. Condorcet voting requires these people to do two separate linear complexity actions.
Fourth, complexity of completely filling a ballot:
IRV is essentially a comparison sort. We know this to be of complexity O(n lg(n)). We can expect most people's process of completing the ballot to be more like selection sort though, which is complexity O(n^2).
Condorcet voting is trivially in THETA(n^2).
Fifth, average case complexity:
We can expect most voters to care about some subset of the candidates. My guess would be that it's upper-bounded by the "Seven, plus or minus two" rule for how many things people can keep in active memory at once.
IRV requires them to sort the 9 or fewer candidates they care about. While selection sort is quadratic in complexity, this is upper bounded by a constant, so the maximum amount of work a voter will have to do to make their voice heard is a constant.
Condorcet requires them to vote for all the people they like and against all the people they dislike for their opinion to be fully heard, so this complexity is linear in the number of candidates, regardless of how many they actually have opinions about.
Sixth, counting complexity:
At worst, IRV requires n-1 passes through v votes, with each vote costing O(n) to process on each pass. This can be optimized of course, but this is the algorithm that would be used for a hand recount, and we want to be able to do a hand recount. This complexity is O(n^2 * v).
Condorcet voting requires THETA(n^2 * v) operations to fill the matrix, with complexity of resolving ambiguities probably varying by method.
Seventh, certificate complexity:
In order for a vote to matter, its results must be certified. Election officials must be able to say to the people why one candidate won.
Approval voting rewards negative campaigning. You end up much better off if the voters can't stand anyone but you. Instant runoff voting rewards positive campaigning, because negative campaigning only helps at all if you successfully discredit every last one of your opponents, and the cost of bashing all of them generally exceeds the cost of actually telling the voters why they should vote for you. Almost everyone agrees that more positive campaigning and less negative campaigning is good for democracy, and is thus good. Electronic voting systems can make instant run-off voting really easy to implement. The convenient thing about it is that if people have a hard time grasping the concept, they can just vote in the first round. Most people vote for only one of two parties, and third party supporters would be especially likely to be aware of how to support their candidate without withdrawing support for the person most likely to beat out the guy they really hate. The complexity isn't really such a big deal, and the benefits are tremendous.
Minnesota code seems to think that your landlord only had three weeks. I'd give you a URL, but the site I found had session-specific URLs. You should be able to find it yourself searching at http://www.revisor.leg.state.mn.us:8181/SEARCH/BAS IS/mnstat/public/www/SF
You may want to consult a lawyer, but I hope that helps.
Many states require that a landlord enumerate all deductions from the security deposit by the end of some period of time that begins after you vacate. If they don't do this, you get it all back, no matter how badly the place is messed up. See your state law.
A lot of people have 200+ kbps uplinks that are artificially capped in the realm of modem speeds by their ISPs. I wonder how many of these have been counted in this survey?
OpenSSH's privilege escalation vulnerability was due to a double free bug. Thus, the only root exploit in the default install to ever have been found in OpenBSD was due to a double free. The zlib vulnerability, which affects a whole bunch of programs that link with zlib, was also a double free bug. It's not something that typically gets taught in undergrad CS courses, like buffer overrun, but it's not unheard of for it to be exploited.
I'm dead serious. Keep them in a safe place. I used to work academic tech support, and we got a call from a guy who'd been getting abuse emails, even though his computer had been stolen a week prior. It's a good thing whoever stole it set it up as a warez server on our network, or he would never have gotten it back. We had implemented campus-wide MAC address registration that fall, so we were able to match it up. Of course, he never would have known to ask, but you, good slashdotter, do know to ask your campus IT if your ethernet card has hit their DHCP server anywhere, or if your wireless card has hit any of their APs. On a heavily urban campus, this is somewhat less useful, but if you're in a college town, there's a decent chance that your stolen gear will get used on campus.
When doing BW conversions from RGB, all you do is multiply R by.29, G by.60, and B by.11 This roughly corresponds to the sensitivity of the human eye. 16-bit color is pretty much always 5-6-5. 8-bit color is sometimes 3-3-2, but having a 256-entry lookup table isn't really that expensive. Of course, this can lead to some pretty nasty side effects, like the color palette issues in some implementations of X, as the active window gets to set the palette for everything on the screen. It's really annoying when you want to get work done, but it's also kind of fun in the same way that pressing F9 over and over with Expose on a Mac can be.
In many countries, this study would be meaningless or impossible. I'm starting to believe that the US two party system has profoundly impaired the way Americans think about politics. Like the tribe that was recently studied that only had words for "one", "two", and "many" that could barely keep track of numbers as high as four, we, as a culture that lacks the nuances of a system such as a several-party parliamentary democracy, are doomed to think of politics as a neverending battle between two ends of a one-dimensional scale. Granted, we know at a conscious level that our own system actually is a bit more nuanced than that, just like that tribe knows that numbers higher than two are distinct from each other, but this is our default way of thinking, and deviating from it requires more effort than we're typically inclined to put ourselves through in our everyday life.
This is why we had 70% of the population believing that the 9/11 highjackers were Iraqi, rather than being mostly Saudi. All they knew was that they looked Arab and the Iraqis are mostly Arab and they're the bad guys. (Don't get me started on "bad guys") This is why Bush's resignation wasn't immediately demanded by the American people when he said "You're either with us or you're against us." This is why Michael Moore can put together a movie that shows that a lot of the people who are trying to attack us are connected to people that our government supports, and people react as though this is an insightful or inflamatory accusation, when foreign news has been reporting on this rather dispassionately for years.
The only reason we don't have mini-PBXs built into our home wireless routers already to handle this with ultra-paranoid encryption and key control is that we need backwards compatibility with POTS, or so we believe. The best way to solve this seems to be by usage creep. It's taken a long time, but for many people email has just about entirely supplanted postal mail, and the only exceptions to this are generally financial documents whose physical delivery is mandated by law. It's done so more by convenience benefits than cost benefits. When people realize that they can have conference calls and the like using VoIP, they'll really pick up on it. What we really need are VoIP services that are capable of handling either internet addresses or POTS exchanges as endpoints. If user@host rises to equal status in people's minds as (###)###-#### then we'll start seeing people dropping off the POTS network completely, and then we can really have decentralized phone service. When that happens, they can wiretap my phone calls when they pry my soldering iron from my cold, dead hands.
If I had to guess where this trend was going to start, I'd say college campuses. Large companies have their own internal phone networks anyway, but they need to be reachable to the outside world. I ditched land telephones altogether as a result of college living, and I'm hoping to never go back, unless maybe for VoIP. College students are already using Xbox games for free long distance to their high school friends who have gone off to other institutions. They get to share all the gossip they normally would over their high-bandwidth, low-latency connection, except it's free, and if your buddy confesses that he hooked up with your old flame, you can shove a rocket down his throat.
The great thing about the similarity matching algorithms is that they read with noise filtering the same way that humans do. They also allow for like-character matching without any added computational overhead. This means that you can make a table of unicode characters that are similar to certain ascii characters that gets incorporated into the similarity matrix. By the power of these properties combined, your spam filter can recognize that c;al_is is intended to look like cialis, without a lot of expensive extra computations.
Now that we've neutralized that form of message garbling, we're left to dealing with bayes filter poisoning. This is something that entropy-based filtering deals with quite well.
All spam filtering techniques have weaknesses, but if you use a few different methods in concert, preferably within the same package to spare the poor user from having to set up a whole lot, you can get just about all of it.
Even using a few of these different methods together, I still get a few ads from companies I've done business with that have screwed up my communication preferences. This sucks, but most of these companies are clueless rather than malicious. Threatening to take my business elsewhere has never failed to correct these problems.
This certainly isn't the case with the 6820. The button is there, but you gotta jam your thumbnail down on it for a few seconds to get it to respond. I'd say they overcorrected, but I guess it's better this way than with what you endured.
My real problem with it is that when I lose signal for extended periods of time, it sometimes doesn't re-establish signal when I get back to a place where it's perfectly clear, occasionally blaming the problem on a SIM card error, usually just not reconnecting. I have to power cycle it to get back on. It seems like a simple software bug, so I'm holding out hope that I might be able to get an update.
I would say that solar vehicles are generally both more likely to get into crashes and better at protecting their operators in crashes than motorcycles are. One of the critical differences is that you've got a lead and chase vehicle to shield you from most hazards, and to warn you of things that are difficult to see from 2 feet off the ground. While this is certainly a huge safety gain, it's out of the driver's immediate control. Air travel is much safer than ground travel, but even people who are well aware of this have a visceral fear of it for this reason.
"The long run economic profit of any perfectly competitive business is zero."
There. Now that we've reminded ourselves of one of the many reasons that economics is called "the dismal science", the fact that they're making $1 on each unit seems like a good idea for everyone involved.
This is an experimental solar racing car, not designed for road. If the steering and weight are not adequate for road safety, then why not just put it on a trailer and float it between events?
You never expect your steering system to fail on the road. It just happens a little more often with an experimental device. There doesn't need to be something fundamentally wrong for this to happen, just less resistance to natural entropic effects. As for inertia, see motorcycles. Certainly more dangerous than cars, but still street legal.
According to another post here, U of T won a safety at on of their recent events. Obviously the car met criteria for racing. Why did they need to take it on the road?
Awards like that are generally given out for excellent design. Unfortunately, there's a difference between robots programmed to weld one joint to certain precise tolerances, and a human with a wrench adjusting things to make room for some change in the experimental system. It's not a shortcoming of the human so much as an advantage of the machine. Things are a little more likely to go wrong.
Anyway, many solar car races are on the road. The American Solar Challenge and the World Solar Challenge are both road races. You want to get road experience when your driver and support team are not under the stress of the race, which often involves getting 4 hours of sleep for more than a week straight. If you don't get some road experience when you're not racing, you're probably increasing your risk of accident, not decreasing it. Anyway, track racing is much simpler than road racing, and just doesn't stress the technology as much. Less is learned from it. That's why we take them on the road.
Bull crap! If the cars are safe for the road, then drive them ther, and if no then keep them off. If this was not caused by a mechanical failure, then it was because the car should not have been on the road. It is a terrible shame that someone had to die, but hopefully from now on people will think twice about taking their experimental cars on public streets. The U of T team (and any one else who has been taking this risk) should revisit their policies regarding transportation between events.
Solar car racing has been going on for a very long time. This is the first fatality. Driving a solar car is safer than having cosmetic surgery. It's safer than taking a ride on the space shuttle. It's still not as safe as driving a Volvo. You take many safety measures, like making cockpits easily escapable, having lead and chase vehicles, having a scout vehicle to warn of road hazards, etc. After all this, it's still a risk. I wouldn't call it an irrational one, but it's certainly one greater than some people have the nerve for.
Solar Cars, due to the extremely low amount of energy available to them, have to be extraordinarily light. When I was in the 2001 American Solar Challenge, there were cars that only weighed 2 1/2 times their driver. This is with metal roll cages on the inside. I know people are experimenting with full-chassis composite construction, which will make the cars even lighter. While it's true that F1 cars all have composite roll bars because of their strength, the problem is simply a matter of inertia. When a 1000 kg car hits a 2000 kg truck head-on, it's bad for the car, but when a 160 kg car with an 80 kg driver hits a 2000 kg truck head on, it's absolutely devastating, no matter how strong the material is holding it together.
I'm curious to see how this will affect solar racing rules. It's not like they're going to require crash testing of your half million dollar prototype that you bring to the race. Personally, I think there's probably a lot more room to be stricter with accident avoidance stuff, like making sure your steering and suspension is REALLY secure. My team nearly lost its car to a suspension failure, while going 65 on an interstate down a hill towards a bridge over a very deep chasm. The driver kept it kinda under control, but we got lucky. Turns out there was nothing inherently wrong with our design, aside from the fact that it wasn't sufficiently redundant to resist the force of miniscule human error in construction, followed by 1000 miles of road wear. Point is, wheels just don't fall off of modern production automobiles, but things like that happen with experimental prototypes.
On a personal note, driving a solar car that I built myself was one of the greatest thrills of my life. I was too big to drive our team's car with the top on, but even taking it around the parking lot on battery power was a great thrill. I can't imagine how taking that out on the road feels, but I imagine it compensates somewhat for the very real danger that exists whenever people strap themselves into unorthodox moving objects for the sake of enhancing the body of human knowledge. Whether it's a solar car developed and built by college students or a multi-billion dollar space shuttle designed by one of the largest engineering teams ever assembled, there is no substitute for experience, as NASA has tragically learned twice.
If anyone who knew Andrew is reading this, I hope you realize that he took a risk in pursuit of something greater than himself, for the benefit of everything on Earth.
Actually, you can get a grocery card without actually handing them the information sheet. The easy way to do it is to go up to the register, look at your keychain, say "crap, it looks like my card fell off". They'll get out a new one, swipe it for you, and tell you to turn in the form after they ring you up. Pay for your groceries and leave. Not at all hard. This insurance thing is a lot creepier. I'm not so much worried about Progressive, since they let you review the data before you send it, but the next step on the slippery slope won't be so friendly.
He's been there for 17 years. Non-competes weren't trendy back then, and if he was under one, it would have been mentioned. It also wouldn't be newsworthy. He is, however, under NDA. The argument that Seagate is trying to make is that it is impossible for him to honor the NDA if he takes the job. Now that we actually do have companies requiring employees to sign non-competes, the stark comparison is going to make Seagate look really weak here. In order to be valid, a contract must represent a "meeting of the minds". This means that both sides have to understand what they're getting into, or if they don't, it's because someone screwed up in the reading of it, rather than someone deliberately writing a confusing contract. It's pretty easy to argue that Pete Goglia could not reasonably have been expected to understand that the NDA was implicitly a non-compete. The bigger danger is really on Western Digital's end, as this opens them up to potential litigation for misappropriation of trade secrets for anything they come out with in the near future, regardless of whether or not any trade secrets were in fact misappropriated.
I don't know what was like for him there, but I'm going to take a stab in the dark and guess that all of this probably could have been avoided if Seagate had given him a raise, better benefits, and perhaps more respect around the office. Courts have a tendency of drawing the same conclusion, and don't really like attempts like this to keep employees from taking jobs with competitors. They'll uphold an explicit and unambiguous non-compete, but I'd be absolutely astonished if they go for this legal theory.
Actually, we do have non-volatile memory with unlimited write cycles. It's just a lot more expensive, so it doesn't get used much, particularly in very high volume. It sounds like this technology is sufficiently flexible that they could put whatever sort of memory they see fit on the inside and it would work the same way. You put in the expensive stuff, and you've got a replacement for the general purpose hard drive. You put in the cheap stuff and you've just replaced a media storage disk.
These systems are absolutely wonderful. Of course, access to the card itself is an issue, that's why they use two-factor authentication.
Requiring users to frequently change gibberish passwords tends to be much less secure than either frequently changed non-gibberish passwords or long-term gibberish passwords, because the users forget them, and either spend a lot of time on the phone with IT (social engineering waiting to happen, many IT shops are happy to authenticate users over the phone by DOB and SSN, which are easy to come by,) or write it down in obvious places if IT is annoying about password change requests (which it should be).
The effect of this is to make your skills non-portable. You won't want to leave your job because your experience is so highly specialized that you'd basically be an entry-level programmer wherever you end up. Not only will you be entry-level, but your MS will be only marginally more useful than the guy who took a couple community college courses in the cube next to you.
This whole idea is centered around getting more code written, cheaper. While it may in the short term improve quality due to specialization, in the long term it serves to replace software engineers with codemonkeys.
Ideas like this make Stallman look like an optimist.
I accidentally modded you down. Not quite sure how that happened. Anyway, replying here ought to undo that. It would be nice if we had a way to undo our own mods without requiring a post, wouldn't it.
Slashdot Mirror
It sure sounds to me like any interpreted language would infringe, or be prior art. Lisp had an interpreter all the way back in 1958. A bit more recently, the Bourne Shell did this. I can't find a patent number anywhere, but I doubt this predates Lisp and the Bourne Shell. Time for us to get even more pissed off than we did when Eolas beat Microsoft in the trial court. This really needs to be dealt with in Congress. Go write your Representatives and Senators. I am.
It's quite obvious that Condorcet Voting and Instant Runoff Voting are designed to help third parties. Since they eliminate the danger of helping your opposition by making a third party vote (the criticism of IRV in the linked article applies just as well to Condorcet voting, so it's really moot) we can expect LOTS of 3rd parties on the ballot. It's hard enough getting people to find out about 2 candidates. It's impossible and unreasonable to expect people to find out about 20. Both schemes can easily handle people casting single-shot votes, so they do not disenfranchise anyone as long as incomplete ballots are acceptable.
First, the inherent complexities of the ballot:
Since a voter could vote for 0 to n candidates in any order, an IRV ballot has SUM(i=0..n) Perm(n, i) = e * n! = THETA(n!) possible assignments.
Since a voter has n(n-1)/2 choices that can be 0,1, or don't care, there are 3^(n(n-1)/2) = THETA (sqrt(3^(n^2))) possible assignments. Since this is polyexponential, it dominates the THETA (n!) complexity of IRV.
Second, complexity of representing an individual ballot:
IRV has n identifiers that must be of size lg(n), so it can take THETA(n lg(n)) space to represent a full ballot.
Condorcet ballots can use implicit ordering, so the size of representing each decision is constant, making the cost of representing n(n-1)/2 decisions THETA(n^2).
Third, the decision complexity required to cast a single-shot ballot
IRV allows a voter to vote their preferred candidate first and then vote for no one else. While this decision itself is more complicated than a binary decision, a single-shot voter can go into the booth and make one simple action and not be disenfranchised.
Condorcet voting requires that a voter make (n-1) comparisons just to make their single-shot vote. Furthermore, they must make an additional (n-2) comparisons if there's a guy they REALLY want to lose. As of right now, there are a whole lot of "Anyone but Bush" people who will be voting for Kerry (possibly holding their noses) because it's a very simple way to do the best they can to make sure Bush is defeated. Condorcet voting requires these people to do two separate linear complexity actions.
Fourth, complexity of completely filling a ballot:
IRV is essentially a comparison sort. We know this to be of complexity O(n lg(n)). We can expect most people's process of completing the ballot to be more like selection sort though, which is complexity O(n^2).
Condorcet voting is trivially in THETA(n^2).
Fifth, average case complexity:
We can expect most voters to care about some subset of the candidates. My guess would be that it's upper-bounded by the "Seven, plus or minus two" rule for how many things people can keep in active memory at once.
IRV requires them to sort the 9 or fewer candidates they care about. While selection sort is quadratic in complexity, this is upper bounded by a constant, so the maximum amount of work a voter will have to do to make their voice heard is a constant.
Condorcet requires them to vote for all the people they like and against all the people they dislike for their opinion to be fully heard, so this complexity is linear in the number of candidates, regardless of how many they actually have opinions about.
Sixth, counting complexity:
At worst, IRV requires n-1 passes through v votes, with each vote costing O(n) to process on each pass. This can be optimized of course, but this is the algorithm that would be used for a hand recount, and we want to be able to do a hand recount. This complexity is O(n^2 * v).
Condorcet voting requires THETA(n^2 * v) operations to fill the matrix, with complexity of resolving ambiguities probably varying by method.
Seventh, certificate complexity:
In order for a vote to matter, its results must be certified. Election officials must be able to say to the people why one candidate won.
IRV requires showing vote tot
Approval voting rewards negative campaigning. You end up much better off if the voters can't stand anyone but you. Instant runoff voting rewards positive campaigning, because negative campaigning only helps at all if you successfully discredit every last one of your opponents, and the cost of bashing all of them generally exceeds the cost of actually telling the voters why they should vote for you. Almost everyone agrees that more positive campaigning and less negative campaigning is good for democracy, and is thus good. Electronic voting systems can make instant run-off voting really easy to implement. The convenient thing about it is that if people have a hard time grasping the concept, they can just vote in the first round. Most people vote for only one of two parties, and third party supporters would be especially likely to be aware of how to support their candidate without withdrawing support for the person most likely to beat out the guy they really hate. The complexity isn't really such a big deal, and the benefits are tremendous.
Minnesota code seems to think that your landlord only had three weeks. I'd give you a URL, but the site I found had session-specific URLs. You should be able to find it yourself searching at http://www.revisor.leg.state.mn.us:8181/SEARCH/BAS IS/mnstat/public/www/SF
You may want to consult a lawyer, but I hope that helps.
Many states require that a landlord enumerate all deductions from the security deposit by the end of some period of time that begins after you vacate. If they don't do this, you get it all back, no matter how badly the place is messed up. See your state law.
A lot of people have 200+ kbps uplinks that are artificially capped in the realm of modem speeds by their ISPs. I wonder how many of these have been counted in this survey?
I saw "BMI" and thought "BMG". Thanks for catching me.
No, the artists are still starving. BMI is doing well though.
OpenSSH's privilege escalation vulnerability was due to a double free bug. Thus, the only root exploit in the default install to ever have been found in OpenBSD was due to a double free. The zlib vulnerability, which affects a whole bunch of programs that link with zlib, was also a double free bug. It's not something that typically gets taught in undergrad CS courses, like buffer overrun, but it's not unheard of for it to be exploited.
I'm dead serious. Keep them in a safe place. I used to work academic tech support, and we got a call from a guy who'd been getting abuse emails, even though his computer had been stolen a week prior. It's a good thing whoever stole it set it up as a warez server on our network, or he would never have gotten it back. We had implemented campus-wide MAC address registration that fall, so we were able to match it up. Of course, he never would have known to ask, but you, good slashdotter, do know to ask your campus IT if your ethernet card has hit their DHCP server anywhere, or if your wireless card has hit any of their APs. On a heavily urban campus, this is somewhat less useful, but if you're in a college town, there's a decent chance that your stolen gear will get used on campus.
When doing BW conversions from RGB, all you do is multiply R by .29, G by .60, and B by .11 This roughly corresponds to the sensitivity of the human eye. 16-bit color is pretty much always 5-6-5. 8-bit color is sometimes 3-3-2, but having a 256-entry lookup table isn't really that expensive. Of course, this can lead to some pretty nasty side effects, like the color palette issues in some implementations of X, as the active window gets to set the palette for everything on the screen. It's really annoying when you want to get work done, but it's also kind of fun in the same way that pressing F9 over and over with Expose on a Mac can be.
In many countries, this study would be meaningless or impossible. I'm starting to believe that the US two party system has profoundly impaired the way Americans think about politics. Like the tribe that was recently studied that only had words for "one", "two", and "many" that could barely keep track of numbers as high as four, we, as a culture that lacks the nuances of a system such as a several-party parliamentary democracy, are doomed to think of politics as a neverending battle between two ends of a one-dimensional scale. Granted, we know at a conscious level that our own system actually is a bit more nuanced than that, just like that tribe knows that numbers higher than two are distinct from each other, but this is our default way of thinking, and deviating from it requires more effort than we're typically inclined to put ourselves through in our everyday life.
This is why we had 70% of the population believing that the 9/11 highjackers were Iraqi, rather than being mostly Saudi. All they knew was that they looked Arab and the Iraqis are mostly Arab and they're the bad guys. (Don't get me started on "bad guys") This is why Bush's resignation wasn't immediately demanded by the American people when he said "You're either with us or you're against us." This is why Michael Moore can put together a movie that shows that a lot of the people who are trying to attack us are connected to people that our government supports, and people react as though this is an insightful or inflamatory accusation, when foreign news has been reporting on this rather dispassionately for years.
The only reason we don't have mini-PBXs built into our home wireless routers already to handle this with ultra-paranoid encryption and key control is that we need backwards compatibility with POTS, or so we believe. The best way to solve this seems to be by usage creep. It's taken a long time, but for many people email has just about entirely supplanted postal mail, and the only exceptions to this are generally financial documents whose physical delivery is mandated by law. It's done so more by convenience benefits than cost benefits. When people realize that they can have conference calls and the like using VoIP, they'll really pick up on it. What we really need are VoIP services that are capable of handling either internet addresses or POTS exchanges as endpoints. If user@host rises to equal status in people's minds as (###)###-#### then we'll start seeing people dropping off the POTS network completely, and then we can really have decentralized phone service. When that happens, they can wiretap my phone calls when they pry my soldering iron from my cold, dead hands.
If I had to guess where this trend was going to start, I'd say college campuses. Large companies have their own internal phone networks anyway, but they need to be reachable to the outside world. I ditched land telephones altogether as a result of college living, and I'm hoping to never go back, unless maybe for VoIP. College students are already using Xbox games for free long distance to their high school friends who have gone off to other institutions. They get to share all the gossip they normally would over their high-bandwidth, low-latency connection, except it's free, and if your buddy confesses that he hooked up with your old flame, you can shove a rocket down his throat.
The great thing about the similarity matching algorithms is that they read with noise filtering the same way that humans do. They also allow for like-character matching without any added computational overhead. This means that you can make a table of unicode characters that are similar to certain ascii characters that gets incorporated into the similarity matrix. By the power of these properties combined, your spam filter can recognize that c;al_is is intended to look like cialis, without a lot of expensive extra computations.
Now that we've neutralized that form of message garbling, we're left to dealing with bayes filter poisoning. This is something that entropy-based filtering deals with quite well.
All spam filtering techniques have weaknesses, but if you use a few different methods in concert, preferably within the same package to spare the poor user from having to set up a whole lot, you can get just about all of it.
Even using a few of these different methods together, I still get a few ads from companies I've done business with that have screwed up my communication preferences. This sucks, but most of these companies are clueless rather than malicious. Threatening to take my business elsewhere has never failed to correct these problems.
This certainly isn't the case with the 6820. The button is there, but you gotta jam your thumbnail down on it for a few seconds to get it to respond. I'd say they overcorrected, but I guess it's better this way than with what you endured.
My real problem with it is that when I lose signal for extended periods of time, it sometimes doesn't re-establish signal when I get back to a place where it's perfectly clear, occasionally blaming the problem on a SIM card error, usually just not reconnecting. I have to power cycle it to get back on. It seems like a simple software bug, so I'm holding out hope that I might be able to get an update.
I would say that solar vehicles are generally both more likely to get into crashes and better at protecting their operators in crashes than motorcycles are. One of the critical differences is that you've got a lead and chase vehicle to shield you from most hazards, and to warn you of things that are difficult to see from 2 feet off the ground. While this is certainly a huge safety gain, it's out of the driver's immediate control. Air travel is much safer than ground travel, but even people who are well aware of this have a visceral fear of it for this reason.
Repeat after me:
"The long run economic profit of any perfectly competitive business is zero."
There. Now that we've reminded ourselves of one of the many reasons that economics is called "the dismal science", the fact that they're making $1 on each unit seems like a good idea for everyone involved.
This is an experimental solar racing car, not designed for road. If the steering and weight are not adequate for road safety, then why not just put it on a trailer and float it between events?
You never expect your steering system to fail on the road. It just happens a little more often with an experimental device. There doesn't need to be something fundamentally wrong for this to happen, just less resistance to natural entropic effects. As for inertia, see motorcycles. Certainly more dangerous than cars, but still street legal.
According to another post here, U of T won a safety at on of their recent events. Obviously the car met criteria for racing. Why did they need to take it on the road?
Awards like that are generally given out for excellent design. Unfortunately, there's a difference between robots programmed to weld one joint to certain precise tolerances, and a human with a wrench adjusting things to make room for some change in the experimental system. It's not a shortcoming of the human so much as an advantage of the machine. Things are a little more likely to go wrong.
Anyway, many solar car races are on the road. The American Solar Challenge and the World Solar Challenge are both road races. You want to get road experience when your driver and support team are not under the stress of the race, which often involves getting 4 hours of sleep for more than a week straight. If you don't get some road experience when you're not racing, you're probably increasing your risk of accident, not decreasing it. Anyway, track racing is much simpler than road racing, and just doesn't stress the technology as much. Less is learned from it. That's why we take them on the road.
Bull crap! If the cars are safe for the road, then drive them ther, and if no then keep them off. If this was not caused by a mechanical failure, then it was because the car should not have been on the road. It is a terrible shame that someone had to die, but hopefully from now on people will think twice about taking their experimental cars on public streets. The U of T team (and any one else who has been taking this risk) should revisit their policies regarding transportation between events.
Solar car racing has been going on for a very long time. This is the first fatality. Driving a solar car is safer than having cosmetic surgery. It's safer than taking a ride on the space shuttle. It's still not as safe as driving a Volvo. You take many safety measures, like making cockpits easily escapable, having lead and chase vehicles, having a scout vehicle to warn of road hazards, etc. After all this, it's still a risk. I wouldn't call it an irrational one, but it's certainly one greater than some people have the nerve for.
Solar Cars, due to the extremely low amount of energy available to them, have to be extraordinarily light. When I was in the 2001 American Solar Challenge, there were cars that only weighed 2 1/2 times their driver. This is with metal roll cages on the inside. I know people are experimenting with full-chassis composite construction, which will make the cars even lighter. While it's true that F1 cars all have composite roll bars because of their strength, the problem is simply a matter of inertia. When a 1000 kg car hits a 2000 kg truck head-on, it's bad for the car, but when a 160 kg car with an 80 kg driver hits a 2000 kg truck head on, it's absolutely devastating, no matter how strong the material is holding it together.
I'm curious to see how this will affect solar racing rules. It's not like they're going to require crash testing of your half million dollar prototype that you bring to the race. Personally, I think there's probably a lot more room to be stricter with accident avoidance stuff, like making sure your steering and suspension is REALLY secure. My team nearly lost its car to a suspension failure, while going 65 on an interstate down a hill towards a bridge over a very deep chasm. The driver kept it kinda under control, but we got lucky. Turns out there was nothing inherently wrong with our design, aside from the fact that it wasn't sufficiently redundant to resist the force of miniscule human error in construction, followed by 1000 miles of road wear. Point is, wheels just don't fall off of modern production automobiles, but things like that happen with experimental prototypes.
On a personal note, driving a solar car that I built myself was one of the greatest thrills of my life. I was too big to drive our team's car with the top on, but even taking it around the parking lot on battery power was a great thrill. I can't imagine how taking that out on the road feels, but I imagine it compensates somewhat for the very real danger that exists whenever people strap themselves into unorthodox moving objects for the sake of enhancing the body of human knowledge. Whether it's a solar car developed and built by college students or a multi-billion dollar space shuttle designed by one of the largest engineering teams ever assembled, there is no substitute for experience, as NASA has tragically learned twice.
If anyone who knew Andrew is reading this, I hope you realize that he took a risk in pursuit of something greater than himself, for the benefit of everything on Earth.
Actually, you can get a grocery card without actually handing them the information sheet. The easy way to do it is to go up to the register, look at your keychain, say "crap, it looks like my card fell off". They'll get out a new one, swipe it for you, and tell you to turn in the form after they ring you up. Pay for your groceries and leave. Not at all hard. This insurance thing is a lot creepier. I'm not so much worried about Progressive, since they let you review the data before you send it, but the next step on the slippery slope won't be so friendly.
He's been there for 17 years. Non-competes weren't trendy back then, and if he was under one, it would have been mentioned. It also wouldn't be newsworthy. He is, however, under NDA. The argument that Seagate is trying to make is that it is impossible for him to honor the NDA if he takes the job. Now that we actually do have companies requiring employees to sign non-competes, the stark comparison is going to make Seagate look really weak here. In order to be valid, a contract must represent a "meeting of the minds". This means that both sides have to understand what they're getting into, or if they don't, it's because someone screwed up in the reading of it, rather than someone deliberately writing a confusing contract. It's pretty easy to argue that Pete Goglia could not reasonably have been expected to understand that the NDA was implicitly a non-compete. The bigger danger is really on Western Digital's end, as this opens them up to potential litigation for misappropriation of trade secrets for anything they come out with in the near future, regardless of whether or not any trade secrets were in fact misappropriated.
I don't know what was like for him there, but I'm going to take a stab in the dark and guess that all of this probably could have been avoided if Seagate had given him a raise, better benefits, and perhaps more respect around the office. Courts have a tendency of drawing the same conclusion, and don't really like attempts like this to keep employees from taking jobs with competitors. They'll uphold an explicit and unambiguous non-compete, but I'd be absolutely astonished if they go for this legal theory.
Actually, we do have non-volatile memory with unlimited write cycles. It's just a lot more expensive, so it doesn't get used much, particularly in very high volume. It sounds like this technology is sufficiently flexible that they could put whatever sort of memory they see fit on the inside and it would work the same way. You put in the expensive stuff, and you've got a replacement for the general purpose hard drive. You put in the cheap stuff and you've just replaced a media storage disk.
These systems are absolutely wonderful. Of course, access to the card itself is an issue, that's why they use two-factor authentication.
Requiring users to frequently change gibberish passwords tends to be much less secure than either frequently changed non-gibberish passwords or long-term gibberish passwords, because the users forget them, and either spend a lot of time on the phone with IT (social engineering waiting to happen, many IT shops are happy to authenticate users over the phone by DOB and SSN, which are easy to come by,) or write it down in obvious places if IT is annoying about password change requests (which it should be).
The effect of this is to make your skills non-portable. You won't want to leave your job because your experience is so highly specialized that you'd basically be an entry-level programmer wherever you end up. Not only will you be entry-level, but your MS will be only marginally more useful than the guy who took a couple community college courses in the cube next to you.
This whole idea is centered around getting more code written, cheaper. While it may in the short term improve quality due to specialization, in the long term it serves to replace software engineers with codemonkeys.
Ideas like this make Stallman look like an optimist.
I accidentally modded you down. Not quite sure how that happened. Anyway, replying here ought to undo that. It would be nice if we had a way to undo our own mods without requiring a post, wouldn't it.