Something is "cool" if, on some level, it makes you think "I want one." I could easily imagine commuting to work in one of these (naturally I can't imagine *everyone* doing that, which of course is part of the charm).
Yeah, but in many cases the Western companies' management doesn't care, because they're focused on returns over the short to mid term.
And why not? Why not give away the company's crown jewels, if you don't see the downside for another three or four years? You might even reduce costs over that timeframe by taking a Chinese partner. If you're the kind of investor who holds stocks for less than a year, why would you care? If you are the kind of investor who rebalances his portfolio every year or so, you might well come out ahead. It's only the people who buy stocks and stick them in the vault for decades, or the people who work for the company that are screwed.
Some of the most innovative, creative papers have been rejected before peer review.
You say that like it's a bad thing. You see, the thing about the Gordian Knot is that Alexander cheated. Scientific consensus is a Gordian knot. You're expected to unravel it a bit at a time, not slice it apart.
Suppose you decide to turn over some piece of scientific consensus, say the opinion that chimpanzees are more closely related to humans than gorillas. The creative, innovative approach is to blow the scientific consensus out of the water by definitively showing that gorillas are more closely related to to humans than chimps.. The problem is that even where the reasoning and data looks pretty good on such a paper, it's almost certainly wrong. Just because the reviewers don't see the huge flaw doesn't mean it's not there.
So what you do is break your paper up into pieces, working your way up from demurely questioning the accuracy of some assumptions in the chimp hypothesis to shoveling the dirt in its grave. It's good for the CV, because then you end up with a brilliant *career*, not just a brilliant paper. Assuming you are right.
True, but it's hard to judge the impact of a disclosure from the nature of the information. That's a major bug in our conception of privacy, particularly in the US. It's not *what* the information is, but how it is used that matters.
I'm reasonably expert when it comes to information privacy issues, but even I don't feel like I fully understand the consequences of granting each permission. I sometimes contact an app developer when an app requests permissions that don't seem right. Usually it has something to do with advertising revenue, but that really just shifts the uncertainty elsewhere.
What the user ought to be asked is not assent to fine grained permissions like "read phone state and identity", but rather usage scenarios like "transmit my identity and application usage to an advertiser for calculating reimbursement to the app developer" that implies a package of fine grained permissions. Furthermore, any party in the transaction should be contractually limited to those uses. A developer who collects or provides usage data beyond what is needed to calculated reimbursement should be liable. An advertiser who sells the data to credit reporting agencies should be liable.
Let's say it costs you ten dollars to make a widget, including paying yourself. The marginal value of the next widget to your customer base is $9.99. You can't sell another widget. Now you have a competitor who can build an equivalent widget for $5.00. Is he going to sell the next widget he makes for $5.00? Of course not. There's a customer out there who values that widget at $9.99 and is willing to pay that, so he sells his widget at that price and pockets the $4.99 as profit.
This is the motivation for nearly all investment in technology development: to be able to produce something for less but sell it at the same price you would if it were more costly.
The way pricing works is that buyers have no reason to care about how much it costs the producer to make a widget, because the producer is going to price his widget at market rates. The producer, on the other hand, doesn't really have any reason to worry about how much a consumer would be willing to pay for a specific widget in a different market. The current market has priced that widget and that's his target. Producers elaborate their products in hopes of snagging customers who want some feature the other widgets don't have, but at some point consumers give up and say, "it's just a widget".
So as a producer you have an optimization problem: how do I price my goods to maximize my returns? In this case, you have more than one place to set prices (the device and the content), but that only means it becomes a more complicated optimization problem. No matter what you do, some customers will walk away from your offer and choose your competitor. The opposite is true too: your customers walked away from the exact deal other potential customers chose. That's economic life.
If somebody can make money selling ebooks for less money and charging more for the device, good for them. The reason it doesn't happen is that the content owners want higher margins on ebooks than they get on paper books. Why? Because they can. If the ebook store can sell a book for $9.99 and it only costs $5.00 to produce, why should the ebook store pocket the extra $4.99 of profit and not the content owner? Also the content owner really doesn't want to see their product selling so cheap in the ebook medium becuase they're still dependent on paper sales. Consumers aren't perfectly rational, after all. They might be willing to pay $20 for a hardbound book, but not if the book is available as an ebook for $5.00, even though the customer in question doesn't have an ebook reader. That's because the customer values the content, and if he realizes the content is actually less than $5.00 of the cost of the book he might change his valuation. That's irrational, but it's also economic life.
without seeing the architecture of the application.
Anybody can write an quiz game that beats you every time. What is (might be) impressive is how you are beaten. If I understand this, the impressive part is supposed to be that the system processes the questions as natural language, generates a number of matches, then chooses the one that it "thinks" matches the sense of the question most closely.
That's a pretty tall order, but still as a demo the app isn't impressive, because the app designer chooses the question or constructs the question in a certain way. This is the kind of demo I, as a programmer, would be delighted to be able to create as a proof of concept, but it's a far cry from being able to answer questions that originate outside the development team.
It's like the old security conundrum: any idiot can create security measures he himself is unable to break.
Well, jails are built for economic reasons too. They are built to optimize the returns for politicians and their clients.
What jails are not necessarily built to do is be part of a rational, effective response to crime.
As an engineer, how would you optimize the use of a limited budget to reduce crime? You'd start with some kind of model. What kinds of crimes and criminals are there? What are the situations in which criminals commit crimes. Then you'd work out interventions to take in each case. One of the things he'd probably do is work out is segregate the convicted criminals who are pose a high risk to people, and he'd make sure they get the longest possible prison time.
Since that leaves a lot fewer prison beds, and building and maintaining more prison space is expensive, he'd probably move the least dangerous criminals out to parole earlier. If that involved monitoring, he'd work out the most likely failure modes to GPS monitoring, and have a response for that worked out and ready.
What he wouldn't do is put somebody who poses a risk to the public out on a GPS device, then have no system for responding to the failure of those devices. Putting GPS devices on those guys turns the monitoring procedure into a man-rated system.
I'm not one of those engineers who think the world would be better if it were run by engineers. But I do think it would be a hell of a lot better if more people could learn to think like one when the need arose.
Actually, SCOTUS took a case where the cops used infrared imaging to get a warrant to search a house where pot was being grown. They caught them by the heat of the grow lamps. SCOTUS threw out the warrant, but it was by a narrow margin. Interestingly, Scalia sided with the defendant. This makes it pretty clear that the cops can't use IR imaging to figure out what's going on behind your walls, but not very clear *why*, or how far the cops can go with sensory enhancement before they cross over the line.
I'd say the law is currently dancing around the fact that it doesn't really have a consistent, working concept of privacy. The infrared through the walls business falls afoul of a simplistic concept of privacy as it applies to information that applies to you but that you are protecting. Information about your movements in public places isn't considered private, although now we can extract patterns of behavior from those "public" facts that perhaps should be considered private.
Actually, legally speaking, it is. At least in the US.
Privacy has undergone a technology driven transformation much as copyright has. What copyright is supposed to do is provide financial incentive for creating things. It does this by exploiting the former fact that copying things is a difficult, awkward process that ordinary users of information seldom want to perform.
Here in the US, privacy was protected by the former fact that there didn't used to be things like biometrics, data networks and databases. It's perfectly fine to put a security camera up watching a public place, because the police or private company has a right to be in a place where they can observe things happening there (it gets complicated and unclear when we talk about sensory enhancements like infrared). So, the cops can put up cameras all over the place, tie them into biometric software that helps them detect you, network those cameras together and put the output in the database, because our legal thinking is rooted in an era when those things did not exist.
The government could potentially track your every move and compile a complete history, without the bother of actually shadowing you. And they won't need a warrant to do it. They won't ever need to say "papers please" either.
The key point: the ability to automate suspicion changes its privacy significance. In US law, at leas, suspicion is not considered a likely intrusion on an individual's rights because it used to be expensive and impractical to act on unless there was a good reason. When systems can do your suspecting for you, things change. The problems people have with TSA are an example of this. It's easier and cheaper to widen the scope of suspicion by sweeping false positives onto the watch list than it is to narrow the scope of suspicion by pruning that list.
The 5% confidence rate as we all know is an arbitrary benchmark, and it has been increasingly under criticism from statisticians in recent years. That said, even if we accept the 5% significance level as objectively correct, the way you're using statistical significance levels is wrong in nearly every respect.
A failure to reach statistical significance represents a failure to prove the null hypothesis under conditions very favorable to the null hypothesis. It does not represent disproof of the hypothesis itself.
Suppose you were planning a sample size of 20, but after 5 samples you run your statistical algorithm and discover (horrors!) that you've only reached a statistical significance level of p = 0.1. Do you throw your hands up and say, "well that disproves that so I may as well give up?" Of course not. You are heartened because reaching p = 0.1 under these circumstances constitutes strong evidence you are on the right track. It's just not enough evidence to publish an affirmative finding yet, but it's plenty of encouragement to do the rest of the work.
Non-significant findings are reported all the time as evidence in support of a hypothesis, although it is not enough to claim a finding. You see this all the time: "The data supported hypothesis 1 significantly (p=0.03), and hypothesis 2 at near significance( p=0.1)." The implication is that although we can't claim this as proof of hypothesis 2 yet, we have good reason to put h2 to more stringent examination.
When dealing with data from the wild, you especially pay attention to results that aren't quite significant. A result that occurs with p=0.1 that just happens is in some ways more interesting than a result that occurs with p=0.05 in a carefully designed experiment that controls confounding factors. If you shuffle a deck of cards under laboratory conditions and deal out a poker hand that has p 0.1, that is nothing remarkable. When a poker dealer deals himself such a hand, it's reasonable grounds for suspicion.
No, that's not how it works. You can't equate the probabilities of system failure (including failure to prevent tampering) on both sides of the outcome, because the a priori evidence for each outcome is different.
Or if you prefer frequentist language, a 10% false positive tampering rate when candidate A wins doesn't translate to a 10% false positive tampering rate when candidate B wins, because the two rates are calculated on disjoint outcome sets.
What's that old saying?...Never ascribe to malice that which is adequately explained by incompetence.
Yep. It's widely promoted by malicious people. If you think about it, it's a really, really stupid thing to say. What never? Then all a malicious person has to do is to make sure there's at least one plausible, innocent explanation for anything they do.
Let me suggest an alternative saying: "Plausibility is not proof."
If you are a conspiracy theorist, the fact you think it's plausible the gummint is beaming thought control rays from satellites doesn't make that true.
If you are a voting machine apologist, the fact that it's plausible that a random software glitch or plain old public stupidity generated an unexpected electoral result doesn't mean that's the explanation.
Moderate skepticism is the healthiest stance to take. If an issue is important, you look at all the plausible and even some reasonably implausible explanations skeptically, and then back the one that stands up best after a thorough examination.
Reminds me of the early days of Scientology. I was attending MIT, and there used to be recruiters who'd harass you when you went through Central Square, asking, "Do you want to take a free personality test?"
A friend of mine came up with the perfect response. He'd say, "Did you pass?" Without exception, the recruiters would respond (with a straight face), "Oh, no, it's not *that* kind of a test."
It was freaky, like they'd lost their capacity to recognize irony along with their body thetans.
Folks don't remember the story of the tortoise and the hare.
After years in business, one word I'm allergic to is "vision". I can't say how many worthless "leader" after another I've heard babble on about "vision", as if what they meant by it was something only a very few human beings could do. In fact, anybody can daydream. What very few people can do is plan, then execute the plan.
If "vision" means anything, it should mean being able to see the end, and a sequence of milestones along the way that validate that end and reward the efforts so far. I can easily imagine world peace. The Arab-Israeli conflict? I can easily imagine that solved. And that's an important first step to planning a solution. Unfortunately the next step is to imagine a sequence of milestones that will show us we're making progress, but won't become unfeasible after the one or two small successes.
I call this the Switzerland/Nigeria dichotomy: would you rather be a citizen of Switzerland, whose wealth is in the labor of its people, or Nigeria, whose wealth comes out of the ground?
It isn't that we're going to end up staying there indefinitely. It's that when we leave there will be a government that values the people little more than it would a spade for digging stuff out of the ground, and that will suit us fine.
Well, you know the suspicion everyone has that the people running things are rich idiots who purchase influence so they can hang the cost of their mistakes on other people?
It tells us that that suspicion is true.
But if we had the capacity to learn from events like this, they wouldn't keep happening.
I didn't even try that, although its an obvious test.
Lenovo's face recognition failed for me because it slows down the login process. Even where it worked right off the bat (which it didn't always) it has to load the software, take the picture, scan it, then analyze it. If your face isn't optimally positioned, you have to stop what you're doing and orient yourself correctly to give the software a chance.
The result was far *slower* than typing a password in, so what was the point? If it were as instantaneous, flexible and reliable as human face recognition, that would be a different matter.
Well, details matter, don't they? We've had these debates before, but we were debating different things.
Radio and TV had huge impacts on culture, but nothing like the web's impact on what is like to be an intellectual. Ultimately the web's impact on culture is going to be much larger than TV or radio -- probably more like the invention of the printing press.
What bothers me about these debates is the assumption that change is necessarily entirely good or entirely bad. When things change, humanity adjusts to a new reality that is sometimes much better than what went before, but never entirely satisfactory.
Take books. We often take the invention of books a baseline case that has no drawbacks. Books give us access to the words and thoughts of people we've never met or who are long dead. Furthermore that access is much more direct an unmediated than the hearsay of oral tradition. But does that mean we only gained, and lost nothing?
I think we probably lost some things with the invention of books. For one thing you can't have anything like orthodoxy without books. Books means you can be educated by the state or religious authorities and then then examined to ensure conformity. The invention of the printing press really destroyed oral tradition, replacing it with commercial popular culture. Now instead of retelling local legends and myths, we recount episodes of "Lost". We may have lost a kind of psychological richness by moving from stories that moved through hundreds or even thousands of storytellers to stories that are created in fixed form. "King Arthur" was the "Harry Potter" of the Middle Ages, but the idea that story must not be touched to remain authentic, that it is even the property of the author means it's unlikely we'll be reading "Harry Potter" hundreds of years from now.
Does this mean we should give up on books or the printing press? Obviously not. It's just that no change comes for free, and great gains aren't made without some corresponding loss.
Well, one thing I don't like about Oracle is its rapaciously sales-oriented corporate culture. Of course business is all about making sales, but Oracle really doesn't care of you bought the wrong thing or spent 10x what you needed to. They don't care if you buy the wrong class of license and have to throw it away because it doesn't work for you. I've seen cases where organizations have paid huge amounts of money to get Oracle Enterprise where they only intended to use features in Oracle Standard at a tiny fraction of the price. I've seen customers buy per seat licenses, not being informed that that requires a CAL for every computer or device (e.g. phone) that uses Oracle hosted data, then have to pay through the nose for CALs or throw out their CALs and buy a per processor license.
Oracle really doesn't care if you buy something from them that will make you really really mad at them a few months down the line, so long as they have your signature on the dotted line and your money in the bank.
That's not the way customers should be treated. Not if you want them to be loyal.
I'm talking rather of Oracle's technical merits, which are considerable, and in many cases unique. But if you can get away with not using a product's unique abilities, your way better off if you have to port away from it in the future. The same goes for MySQL, which has unique aspects that are not features, per se, but idiosyncrasies. Both Oracle and MySQL are not easy products to port away from, unless you're using really, really common denominator features through some kind of abstraction layer like Hibernate.
Well, let's start by confining ourselves to people who make their living at being environmentalists, and whom we have actual knowledge of not mediated through some other source. Yes, there are people who are basically chaos mongers, but they are very, very few and far between, although grossly overrepresented in the media.
You have to take this into account: the "news media" is now entertainment, not information. News is selected to provoke an emotional reaction, not thought. An environmentalist respected in the field saying something thought provoking and nuanced isn't going to get on the nightly news or be quoted in blogs. But get one news hog who saying we should send letter bombs to businesses, he'll be repeated all over the place and identified as an "environmental movement leader", even if his organization consists of himself and a couple of cronies. The fact that you encounter more information about those people doesn't make them common. They are news precisely because they are un-representative.
What you should do is look at the writings of respetected thinkers in the movement, such as Paul Hawken ("Natural Capitalism" and "The Ecology of Commerce"), David Orr ("The Nature of Design: Ecology, Culture, and Human Intention"), Karl-Henrik Robèrt ("The Natural Step Story: Seeding a Quiet Revolution").
In short, two equally smart database systems should perform about the same.
Bingo. Ergo, something is wrong with your Oracle installations.
I've done the same, supported an application that is available for MS SQL, Oracle and Sybase Sql Anywhere. MS SQL and Oracle blow SQL Anywhere out of the water in performance, but in most situations you couldn't tell the difference.
From a developer's POV, SQL Anywhere was in most cases a pleasure to work with, Oracle was acceptable in most cases and outstanding in a few, and MS SQL was a horrible PITA. MS SQL doesn't even provide you with a utility to get a human readable dump of the transaction log, much less any way to use the transaction log in a complex recovery (unless you fancy working with page addresses). When a customer does something really stupid, and calls you up saying, "please, please make it like that never happened," it's as easy as rolling off a log in Oracle, practical in SQL Anywhere and not worth doing in MS SQL.
Slashdot Mirror
You fail to understand the concept of "cool".
Something is "cool" if, on some level, it makes you think "I want one." I could easily imagine commuting to work in one of these (naturally I can't imagine *everyone* doing that, which of course is part of the charm).
Yeah, but in many cases the Western companies' management doesn't care, because they're focused on returns over the short to mid term.
And why not? Why not give away the company's crown jewels, if you don't see the downside for another three or four years? You might even reduce costs over that timeframe by taking a Chinese partner. If you're the kind of investor who holds stocks for less than a year, why would you care? If you are the kind of investor who rebalances his portfolio every year or so, you might well come out ahead. It's only the people who buy stocks and stick them in the vault for decades, or the people who work for the company that are screwed.
Some of the most innovative, creative papers have been rejected before peer review.
You say that like it's a bad thing. You see, the thing about the Gordian Knot is that Alexander cheated. Scientific consensus is a Gordian knot. You're expected to unravel it a bit at a time, not slice it apart.
Suppose you decide to turn over some piece of scientific consensus, say the opinion that chimpanzees are more closely related to humans than gorillas. The creative, innovative approach is to blow the scientific consensus out of the water by definitively showing that gorillas are more closely related to to humans than chimps.. The problem is that even where the reasoning and data looks pretty good on such a paper, it's almost certainly wrong. Just because the reviewers don't see the huge flaw doesn't mean it's not there.
So what you do is break your paper up into pieces, working your way up from demurely questioning the accuracy of some assumptions in the chimp hypothesis to shoveling the dirt in its grave. It's good for the CV, because then you end up with a brilliant *career*, not just a brilliant paper. Assuming you are right.
True, but it's hard to judge the impact of a disclosure from the nature of the information. That's a major bug in our conception of privacy, particularly in the US. It's not *what* the information is, but how it is used that matters.
I'm reasonably expert when it comes to information privacy issues, but even I don't feel like I fully understand the consequences of granting each permission. I sometimes contact an app developer when an app requests permissions that don't seem right. Usually it has something to do with advertising revenue, but that really just shifts the uncertainty elsewhere.
What the user ought to be asked is not assent to fine grained permissions like "read phone state and identity", but rather usage scenarios like "transmit my identity and application usage to an advertiser for calculating reimbursement to the app developer" that implies a package of fine grained permissions. Furthermore, any party in the transaction should be contractually limited to those uses. A developer who collects or provides usage data beyond what is needed to calculated reimbursement should be liable. An advertiser who sells the data to credit reporting agencies should be liable.
That's not how economics works.
Let's say it costs you ten dollars to make a widget, including paying yourself. The marginal value of the next widget to your customer base is $9.99. You can't sell another widget. Now you have a competitor who can build an equivalent widget for $5.00. Is he going to sell the next widget he makes for $5.00? Of course not. There's a customer out there who values that widget at $9.99 and is willing to pay that, so he sells his widget at that price and pockets the $4.99 as profit.
This is the motivation for nearly all investment in technology development: to be able to produce something for less but sell it at the same price you would if it were more costly.
The way pricing works is that buyers have no reason to care about how much it costs the producer to make a widget, because the producer is going to price his widget at market rates. The producer, on the other hand, doesn't really have any reason to worry about how much a consumer would be willing to pay for a specific widget in a different market. The current market has priced that widget and that's his target. Producers elaborate their products in hopes of snagging customers who want some feature the other widgets don't have, but at some point consumers give up and say, "it's just a widget".
So as a producer you have an optimization problem: how do I price my goods to maximize my returns? In this case, you have more than one place to set prices (the device and the content), but that only means it becomes a more complicated optimization problem. No matter what you do, some customers will walk away from your offer and choose your competitor. The opposite is true too: your customers walked away from the exact deal other potential customers chose. That's economic life.
If somebody can make money selling ebooks for less money and charging more for the device, good for them. The reason it doesn't happen is that the content owners want higher margins on ebooks than they get on paper books. Why? Because they can. If the ebook store can sell a book for $9.99 and it only costs $5.00 to produce, why should the ebook store pocket the extra $4.99 of profit and not the content owner? Also the content owner really doesn't want to see their product selling so cheap in the ebook medium becuase they're still dependent on paper sales. Consumers aren't perfectly rational, after all. They might be willing to pay $20 for a hardbound book, but not if the book is available as an ebook for $5.00, even though the customer in question doesn't have an ebook reader. That's because the customer values the content, and if he realizes the content is actually less than $5.00 of the cost of the book he might change his valuation. That's irrational, but it's also economic life.
without seeing the architecture of the application.
Anybody can write an quiz game that beats you every time. What is (might be) impressive is how you are beaten. If I understand this, the impressive part is supposed to be that the system processes the questions as natural language, generates a number of matches, then chooses the one that it "thinks" matches the sense of the question most closely.
That's a pretty tall order, but still as a demo the app isn't impressive, because the app designer chooses the question or constructs the question in a certain way. This is the kind of demo I, as a programmer, would be delighted to be able to create as a proof of concept, but it's a far cry from being able to answer questions that originate outside the development team.
It's like the old security conundrum: any idiot can create security measures he himself is unable to break.
Harry lied to you, Norman. Everything Harry says is a lie. Remember that, Norman. *Everything* he says is a lie...
Well, jails are built for economic reasons too. They are built to optimize the returns for politicians and their clients.
What jails are not necessarily built to do is be part of a rational, effective response to crime.
As an engineer, how would you optimize the use of a limited budget to reduce crime? You'd start with some kind of model. What kinds of crimes and criminals are there? What are the situations in which criminals commit crimes. Then you'd work out interventions to take in each case. One of the things he'd probably do is work out is segregate the convicted criminals who are pose a high risk to people, and he'd make sure they get the longest possible prison time.
Since that leaves a lot fewer prison beds, and building and maintaining more prison space is expensive, he'd probably move the least dangerous criminals out to parole earlier. If that involved monitoring, he'd work out the most likely failure modes to GPS monitoring, and have a response for that worked out and ready.
What he wouldn't do is put somebody who poses a risk to the public out on a GPS device, then have no system for responding to the failure of those devices. Putting GPS devices on those guys turns the monitoring procedure into a man-rated system.
I'm not one of those engineers who think the world would be better if it were run by engineers. But I do think it would be a hell of a lot better if more people could learn to think like one when the need arose.
Actually, SCOTUS took a case where the cops used infrared imaging to get a warrant to search a house where pot was being grown. They caught them by the heat of the grow lamps. SCOTUS threw out the warrant, but it was by a narrow margin. Interestingly, Scalia sided with the defendant. This makes it pretty clear that the cops can't use IR imaging to figure out what's going on behind your walls, but not very clear *why*, or how far the cops can go with sensory enhancement before they cross over the line.
I'd say the law is currently dancing around the fact that it doesn't really have a consistent, working concept of privacy. The infrared through the walls business falls afoul of a simplistic concept of privacy as it applies to information that applies to you but that you are protecting. Information about your movements in public places isn't considered private, although now we can extract patterns of behavior from those "public" facts that perhaps should be considered private.
Actually, legally speaking, it is. At least in the US.
Privacy has undergone a technology driven transformation much as copyright has. What copyright is supposed to do is provide financial incentive for creating things. It does this by exploiting the former fact that copying things is a difficult, awkward process that ordinary users of information seldom want to perform.
Here in the US, privacy was protected by the former fact that there didn't used to be things like biometrics, data networks and databases. It's perfectly fine to put a security camera up watching a public place, because the police or private company has a right to be in a place where they can observe things happening there (it gets complicated and unclear when we talk about sensory enhancements like infrared). So, the cops can put up cameras all over the place, tie them into biometric software that helps them detect you, network those cameras together and put the output in the database, because our legal thinking is rooted in an era when those things did not exist.
The government could potentially track your every move and compile a complete history, without the bother of actually shadowing you. And they won't need a warrant to do it. They won't ever need to say "papers please" either.
The key point: the ability to automate suspicion changes its privacy significance. In US law, at leas, suspicion is not considered a likely intrusion on an individual's rights because it used to be expensive and impractical to act on unless there was a good reason. When systems can do your suspecting for you, things change. The problems people have with TSA are an example of this. It's easier and cheaper to widen the scope of suspicion by sweeping false positives onto the watch list than it is to narrow the scope of suspicion by pruning that list.
The 5% confidence rate as we all know is an arbitrary benchmark, and it has been increasingly under criticism from statisticians in recent years. That said, even if we accept the 5% significance level as objectively correct, the way you're using statistical significance levels is wrong in nearly every respect.
A failure to reach statistical significance represents a failure to prove the null hypothesis under conditions very favorable to the null hypothesis. It does not represent disproof of the hypothesis itself.
Suppose you were planning a sample size of 20, but after 5 samples you run your statistical algorithm and discover (horrors!) that you've only reached a statistical significance level of p = 0.1. Do you throw your hands up and say, "well that disproves that so I may as well give up?" Of course not. You are heartened because reaching p = 0.1 under these circumstances constitutes strong evidence you are on the right track. It's just not enough evidence to publish an affirmative finding yet, but it's plenty of encouragement to do the rest of the work.
Non-significant findings are reported all the time as evidence in support of a hypothesis, although it is not enough to claim a finding. You see this all the time: "The data supported hypothesis 1 significantly (p=0.03), and hypothesis 2 at near significance( p=0.1)." The implication is that although we can't claim this as proof of hypothesis 2 yet, we have good reason to put h2 to more stringent examination.
When dealing with data from the wild, you especially pay attention to results that aren't quite significant. A result that occurs with p=0.1 that just happens is in some ways more interesting than a result that occurs with p=0.05 in a carefully designed experiment that controls confounding factors. If you shuffle a deck of cards under laboratory conditions and deal out a poker hand that has p 0.1, that is nothing remarkable. When a poker dealer deals himself such a hand, it's reasonable grounds for suspicion.
No, that's not how it works. You can't equate the probabilities of system failure (including failure to prevent tampering) on both sides of the outcome, because the a priori evidence for each outcome is different.
Or if you prefer frequentist language, a 10% false positive tampering rate when candidate A wins doesn't translate to a 10% false positive tampering rate when candidate B wins, because the two rates are calculated on disjoint outcome sets.
What's that old saying?...Never ascribe to malice that which is adequately explained by incompetence.
Yep. It's widely promoted by malicious people. If you think about it, it's a really, really stupid thing to say. What never? Then all a malicious person has to do is to make sure there's at least one plausible, innocent explanation for anything they do.
Let me suggest an alternative saying: "Plausibility is not proof."
If you are a conspiracy theorist, the fact you think it's plausible the gummint is beaming thought control rays from satellites doesn't make that true.
If you are a voting machine apologist, the fact that it's plausible that a random software glitch or plain old public stupidity generated an unexpected electoral result doesn't mean that's the explanation.
Moderate skepticism is the healthiest stance to take. If an issue is important, you look at all the plausible and even some reasonably implausible explanations skeptically, and then back the one that stands up best after a thorough examination.
Reminds me of the early days of Scientology. I was attending MIT, and there used to be recruiters who'd harass you when you went through Central Square, asking, "Do you want to take a free personality test?"
A friend of mine came up with the perfect response. He'd say, "Did you pass?" Without exception, the recruiters would respond (with a straight face), "Oh, no, it's not *that* kind of a test."
It was freaky, like they'd lost their capacity to recognize irony along with their body thetans.
Folks don't remember the story of the tortoise and the hare.
After years in business, one word I'm allergic to is "vision". I can't say how many worthless "leader" after another I've heard babble on about "vision", as if what they meant by it was something only a very few human beings could do. In fact, anybody can daydream. What very few people can do is plan, then execute the plan.
If "vision" means anything, it should mean being able to see the end, and a sequence of milestones along the way that validate that end and reward the efforts so far. I can easily imagine world peace. The Arab-Israeli conflict? I can easily imagine that solved. And that's an important first step to planning a solution. Unfortunately the next step is to imagine a sequence of milestones that will show us we're making progress, but won't become unfeasible after the one or two small successes.
I call this the Switzerland/Nigeria dichotomy: would you rather be a citizen of Switzerland, whose wealth is in the labor of its people, or Nigeria, whose wealth comes out of the ground?
It isn't that we're going to end up staying there indefinitely. It's that when we leave there will be a government that values the people little more than it would a spade for digging stuff out of the ground, and that will suit us fine.
Well, you know the suspicion everyone has that the people running things are rich idiots who purchase influence so they can hang the cost of their mistakes on other people?
It tells us that that suspicion is true.
But if we had the capacity to learn from events like this, they wouldn't keep happening.
Hypothetically, would it make any difference to how you felt about that post if I told you it was a mash-up of stuff I found on Google?
I didn't even try that, although its an obvious test.
Lenovo's face recognition failed for me because it slows down the login process. Even where it worked right off the bat (which it didn't always) it has to load the software, take the picture, scan it, then analyze it. If your face isn't optimally positioned, you have to stop what you're doing and orient yourself correctly to give the software a chance.
The result was far *slower* than typing a password in, so what was the point? If it were as instantaneous, flexible and reliable as human face recognition, that would be a different matter.
Well, details matter, don't they? We've had these debates before, but we were debating different things.
Radio and TV had huge impacts on culture, but nothing like the web's impact on what is like to be an intellectual. Ultimately the web's impact on culture is going to be much larger than TV or radio -- probably more like the invention of the printing press.
What bothers me about these debates is the assumption that change is necessarily entirely good or entirely bad. When things change, humanity adjusts to a new reality that is sometimes much better than what went before, but never entirely satisfactory.
Take books. We often take the invention of books a baseline case that has no drawbacks. Books give us access to the words and thoughts of people we've never met or who are long dead. Furthermore that access is much more direct an unmediated than the hearsay of oral tradition. But does that mean we only gained, and lost nothing?
I think we probably lost some things with the invention of books. For one thing you can't have anything like orthodoxy without books. Books means you can be educated by the state or religious authorities and then then examined to ensure conformity. The invention of the printing press really destroyed oral tradition, replacing it with commercial popular culture. Now instead of retelling local legends and myths, we recount episodes of "Lost". We may have lost a kind of psychological richness by moving from stories that moved through hundreds or even thousands of storytellers to stories that are created in fixed form. "King Arthur" was the "Harry Potter" of the Middle Ages, but the idea that story must not be touched to remain authentic, that it is even the property of the author means it's unlikely we'll be reading "Harry Potter" hundreds of years from now.
Does this mean we should give up on books or the printing press? Obviously not. It's just that no change comes for free, and great gains aren't made without some corresponding loss.
into "distracted, shallow thinkers", but it sure as heck makes life more interesting for us.
You ain't kiddin' about Oracle installs being a PITA. Not for somebody who does them all the time, but if you support apps you get reminded.
Most RDBMS EULAs forbid publishing benchmarks unless they've been improved by the vendor, so who knows? Your experience pretty much agrees with mine.
Well, one thing I don't like about Oracle is its rapaciously sales-oriented corporate culture. Of course business is all about making sales, but Oracle really doesn't care of you bought the wrong thing or spent 10x what you needed to. They don't care if you buy the wrong class of license and have to throw it away because it doesn't work for you. I've seen cases where organizations have paid huge amounts of money to get Oracle Enterprise where they only intended to use features in Oracle Standard at a tiny fraction of the price. I've seen customers buy per seat licenses, not being informed that that requires a CAL for every computer or device (e.g. phone) that uses Oracle hosted data, then have to pay through the nose for CALs or throw out their CALs and buy a per processor license.
Oracle really doesn't care if you buy something from them that will make you really really mad at them a few months down the line, so long as they have your signature on the dotted line and your money in the bank.
That's not the way customers should be treated. Not if you want them to be loyal.
I'm talking rather of Oracle's technical merits, which are considerable, and in many cases unique. But if you can get away with not using a product's unique abilities, your way better off if you have to port away from it in the future. The same goes for MySQL, which has unique aspects that are not features, per se, but idiosyncrasies. Both Oracle and MySQL are not easy products to port away from, unless you're using really, really common denominator features through some kind of abstraction layer like Hibernate.
Well, let's start by confining ourselves to people who make their living at being environmentalists, and whom we have actual knowledge of not mediated through some other source. Yes, there are people who are basically chaos mongers, but they are very, very few and far between, although grossly overrepresented in the media.
You have to take this into account: the "news media" is now entertainment, not information. News is selected to provoke an emotional reaction, not thought. An environmentalist respected in the field saying something thought provoking and nuanced isn't going to get on the nightly news or be quoted in blogs. But get one news hog who saying we should send letter bombs to businesses, he'll be repeated all over the place and identified as an "environmental movement leader", even if his organization consists of himself and a couple of cronies. The fact that you encounter more information about those people doesn't make them common. They are news precisely because they are un-representative.
What you should do is look at the writings of respetected thinkers in the movement, such as Paul Hawken ("Natural Capitalism" and "The Ecology of Commerce"), David Orr ("The Nature of Design: Ecology, Culture, and Human Intention"), Karl-Henrik Robèrt ("The Natural Step Story: Seeding a Quiet Revolution").
In short, two equally smart database systems should perform about the same.
Bingo. Ergo, something is wrong with your Oracle installations.
I've done the same, supported an application that is available for MS SQL, Oracle and Sybase Sql Anywhere. MS SQL and Oracle blow SQL Anywhere out of the water in performance, but in most situations you couldn't tell the difference.
From a developer's POV, SQL Anywhere was in most cases a pleasure to work with, Oracle was acceptable in most cases and outstanding in a few, and MS SQL was a horrible PITA. MS SQL doesn't even provide you with a utility to get a human readable dump of the transaction log, much less any way to use the transaction log in a complex recovery (unless you fancy working with page addresses). When a customer does something really stupid, and calls you up saying, "please, please make it like that never happened," it's as easy as rolling off a log in Oracle, practical in SQL Anywhere and not worth doing in MS SQL.