Mod parent up, it is cheap per-student (one computer, one large screen), it can sort-of be used without power (you can type, you just can't tell if you are typing correctly), and it could do double duty as a video player. No doubt a Linux guy here on Slashdot can suggest some way to make 20 keyboards input into a single application?
I think I understand their approach to national security too: it helps increase the power of the federal government, it pleases the lobbyists that want the government to purchase billions of dollars worth of equipment from a particular manufacturer, and it distracts people from more important issues that the politicians would rather not discuss. When looking for motives, ask: who benefits?
I wish people were not so foolish as to think that terrorist attacks can be stopped via airport security. Obviously the terrorists know that airport security exists, so they will attack trains, malls, concerts or busses instead. I also wish people could get it through their heads that terrorist attacks are too rare to worry about. And that the best medicine is prevention (don't breed terrorists through foolish foreign policy).
If a simple pat-down "induced PTSD and sexual abuse trauma", it is more likely to suggest a problem with the passenger rather than the TSA. Even so, America really can't afford billions of dollars in unnecessary equipment and personnel just to provide security theatre, especially since this particular theatre is not the slightest bit entertaining when it happens to you.
And when you can get away with ignoring a court order, isn't that a symptom of a larger problem?
You may be right, at least when it comes to the kind of companies that already squeeze your wallet today.
The reason I really want an easy micropayments system is for the sake of the little guys, the independents, individuals and small businesses, who can't easily convince people to put their credit-card details into their little-known website for a 99-cent payment or donation, and can barely afford the infrastructure to handle logins, payments, etc. And being under the thumb of Paypal isn't ideal.
You don't have to give your money to those price-gougers (assuming no telecom monopoly in your area) and double-dippers who offer high prices and advertising at the same time. Just like I have Netflix instead of cable. Vote with your feet.
I would certainly not advocate automatic micropayments, except if the user explicitly whitelists a page (even then, payments should only be possible in response to user clicks, not auto-refreshes).
The point I wanted to make is that micropayments could be a feasible and fair alternative to advertising, if done right, even something that could fund higher-quality content than advertising can fund, and it's unfortunate that no such system exists. Maybe it's because there are a lot of details to work out, such as how to keep transaction costs far below one cent, but maybe the big boys in the financial industry just don't see profit potential in transactions of less than a penny.
If you're opposed to paying a cent or less per page, you could still go to the many websites that would continue providing service for free.
Besides, micropayments don't have to block access to page content. A page could show its content and still request a one-click 5-cent tip (the one-click, don't-leave-the-page nature of it is the key to convince more people to give tips). The way I envision it, the micropayment button would be part of the web browser, outside the page content, where it is impossible to trick a user into clicking a hidden pay button, or to claim that a different amount will be paid than will actually be paid. The feedback about payments having occurred would also be part of the web browser, making it impossible to hide the fact that there has been a transaction.
I presume at the level of 1-cent transactions, it might be too costly to deal with chargebacks/refunds, so clarity of the user interface would be very important.
But, another way micropayments could work would be that rather than being actual payments, they "buffer up" so that actual money only transfers in minimum amounts of 25 cents or so, an amount large enough to cover the costs of managing refund requests and interfacing with traditional financial systems. In that case, you could essentially visit 24 one-cent pages of a web site "for free" with a real transaction occurring after 25 pages or more. A micropayment standard, then, could just be a cheap buffering system that exists to minimize the large transaction costs of traditional financial systems.
The fine details, such as who would provide the buffering functionality (it can't be the web site itself or the web browser itself, since neither is guaranteed trustworthy) is something I leave to the security/crypto/trust theorists.
What we really need is a micropayment system that makes it feasible for consumers to spend one cent per page or less, to avoid advertising without "subscribing" to a website, without the inconvenience of getting out their credit card, without having to share any private information with websites, without age restrictions (did you have a credit card at 13?). A system in which websites do not have to implement complicated paywall, billing, or log-in infrastructure, do not have to subject themselves to capricious decisions by Paypal, etc.
Until we have such a system, advertising will have to be the main source of revenue in general.
A hot summer does not prove AGW, nor does a cold winter disprove it. In fact, any number of hot summers does not prove AGW; at most they only prove that warming is occurring.
To prove that the Global Warming is Anthropomorphic requires a lot of additional evidence, which has been gathered. There is now a strong concensus among scientists that "man-made" is the only explanation that fits.
There's no indication here about hardware, drivers, or any number of external factors here. This is purely Valve having another dig at Microsoft in the press
I think Valve's main point was that they started with 6 FPS on Linux and are now over 300. I really think their main purpose was to show that their porting effort to Linux/OpenGL is a complete success, there's no need to discuss benchmarking hardware just to tell us that their work is going well. The "dig at Microsoft" you speak of is just editorializing by Slashdot ("MrSeb").
I didn't buy any mainstream music for the past few years because Amazon won't sell music in Canada and I'm boycotting iTunes entirely.
But then a friend referred me to a company willing to take my money, 7digital. Oddly, they don't show up in the first page of Google results for "Buy music in Canada".
From the web site: "Opa is the most advanced application framework for JavaScript. All aspects are directly written in Opa: Frontend code, backend code, database queries and configuration. And everything is strongly statically typed."
So you write code in one language, and your functions are automatically translated to Javascript as needed. The same exact code can run on client and server, but if it runs on the server then it can be optimized better because Opa (unlike Javascript) is statically typed.
Opa appears to be a functional language like Haskell, OCaml, etc., which means that you can write most of the code without specifying data types, as types will be inferred automatically. It also means that data is immutable (write-once) by default. The syntax has a slightly more "conventional" appearance than some other functional languages, so it doesn't look quite as foreign to those who are new to functional programming. Opa is not object-oriented, but it does at least offer "modules" that support dot-notation.
Personally, lack of OO features always makes me uncomfortable because I'm not sure how to use the design patterns I am used to using. IMO the Opa manual should really have a chapter like "Opa for OO dummies" where it explains what the functional equivalent of each of the myriad OO design patterns. And this should be preceded by a "Opa for procedural dummies" chapter, that explains how to replace your efficient procedural code based on hashtables or arrays with Opa code.
Baby steps. The copyright lobby got 95 year copyright terms in the U.S. Even 50 years would be a major improvement, and most all of us slashdot types could agree to reduce copyright to 30 years.
10 years probably suffices for industry to turn a profit, but when you're fighting powerful companies and ideologues who want perpetual copyright, who fear competition from 50-year-old works, who would claim it's an injustice that every single person that reads Mark Twain should pay their great-great-grandkids...
30-year copyright would be an enormous victory. We would have the right to freely play and remix numerous early NES games, to distribute the original Star Wars (the one with the less honorable Han Solo), to have free collections of oldies, disco, Elvis and the Beatles bundled with every new iPod, and so forth.
First of all, reducing terms from 20 to 5 years would reduce the effect of software patents by 75% all by itself.
And they *are* lobbying for more than shorter durations. The EFF's 7 suggested points, taken together, would remove most of the remaining effect of software patents. Like most of us software developers, I expect the EFF would support complete elimination of software patents, but given how patent-friendly Washington is, perhaps they thought it would be a more effective strategy to lobby for weaker patents instead.
Well, let's say there is a kilogram of antimatter floating through space and it hits Earth.
The antimatter is annihilated in an explosion of 180,000 Terajoules of energy. Oh, and some of Earth too.
There can't be much antimatter in the universe because it explodes on contact with any matter it touches. Given e=mc^2, one kilogram of antimatter plus one kilogram of matter equals 2c^2 = 18e16 joules of energy = 180,000 TJ.
You may as well ask programmers to never make a mistake, or people in general, for that matter.
I remember reading the report on the disaster. It wasn't just that a 16-bit variable overflowed. It overflowed in a noncritical system, which led to the shutdown of the main Inertial Reference System and the backup, leaving nothing to fly the rocket.
I don't have the report handy, but it was roughly four problems in combination that brought down the rocket: the bug itself, lack of testing for the bug with altered telemetry (the test telemetry used in simulation was significantly different from the real-life trajectory), failure to handle the exception, and the assumption that an unhandled exception indicated hardware failure (causing the main and backup computer to both shut down).
Software always has bugs. One must be sure to do enough testing to find them, and (if failure is obscenely costly) to plan out some sane ways to handle unexpected bugs in the field.
What are these "channels" you speak of? I use Netflix. I just watch whatever I feel like, whenever I feel like it. I only need one channel. The Netflix channel.
Anyway, I'd hardly say all shows are "niche" that are not reality shows. There is one reality show I watch--it's about Gordon Ramsay rehabilitating restaurants. How niche is that?
No. Just no. This guy stole from Target, not Lego. Why would Lego lower their prices just because there's a shoplifter at Target? Anyway, Lego's patents have expired, and other companies are free to compete.
It seems to me that another important use case is 911 in remote areas. Your reception may not be good enough for a voice call, but it may still be good enough for a text.
That's not really on-topic. Yes, WPF makes life difficult, but Mono for Android doesn't even support WPF. On Android, you'll be using the same widgets in C# as Java developers do.
Not exactly. The bulk of the construction costs are at the last mile. Wires into homes are built regardless of whether a homeowner uses them, and neighborhood-level equipment must be built and operated (at roughly constant cost per house) regardless of how much downloading or uploading occurs. So it doesn't automatically make sense to count local communication toward a bandwidth cap. Instead, the standard monthly fee seems like a fairer way to cover those costs (and the fee is certainly high enough to do so!)
Long-distance links, while also a fixed cost to build, are used at capacity or near capacity; therefore, they tend to be built according to demand and so their cost is dominated by the amount of bandwidth needed instead of by the number of potential customers.
It does, therefore, make some sense for Comcast to "forgive" local traffic since it usually doesn't go though very many saturated internet links. To evaluate fairness, my question would be, is this a special exemption for Xfinity, or is all local traffic (e.g. file sharing, Skype) forgiven, within, say, city limits?
Of course, if the long-distance costs are very small, as you imply, then that just means the long-distance bandwidth cap ought be large, and overage fees small.
Python is a good choice because it offers an interactive interpreter prompt. Or better yet, see if you can acquire the (Javascript) interactive programming tools by Bret Victor, featured in his CUSEC 2012 talk.
An 11 year old can learn from books, too--that's how I learned when I was 11 (most of the books in the library were about BASIC, I thought "integer" was pronounced "inte-geer", and I didn't know that other languages existed for about a year.). But I'm not sure it's necessary in the internet age.
Some have suggested C... give me a break. I started using C when I was 13 and hated it compared to BASIC. I mean, there's no function to read a single key without reading a whole line? No function to check if a key has been pressed? No function for changing the text color or drawing stuff? What's this pointer nonsense? Why do I have to wait for it to "compile" and "link"? I use C++ routinely now, but it's not for beginners.
Despite your probability error, come to think of it you have a better case if you consider not just WC but all the various diseases against which parents are not vaccinating for fear of autism. The combined risk of death from all of those unvaccinated illnesses may be similar to the rate of autism.
But what really matters, of course, is not the rate of autism in the general population, but rather how much the risk of autism drops by not having vaccinations. We know this is roughly 0.0%, but even if you believed in this link, the "safety" gained by not vaccinating must surely be less than the total rate of autism (which is 0.1-0.2% according to Wikipedia and 0.6% for Autism Spectrum Disorder).
Nope, that's an elementary probability mistake. The rate of autism applies to everyone while the rate of death from whooping cough only applies to the people who contract whooping cough, which is probably a small fraction of those who are vaccinated (partly because most people are vaccinated). You have to multiply P(dying from WC) times P(getting WP) before you can compare with the autism rate.
Slashdot Mirror
Mod parent up, it is cheap per-student (one computer, one large screen), it can sort-of be used without power (you can type, you just can't tell if you are typing correctly), and it could do double duty as a video player. No doubt a Linux guy here on Slashdot can suggest some way to make 20 keyboards input into a single application?
That's a link to OLPC - the specific thing he mentioned was too expensive for his needs. (Plus, they don't sell small quantities to individuals.)
I think I understand their approach to national security too: it helps increase the power of the federal government, it pleases the lobbyists that want the government to purchase billions of dollars worth of equipment from a particular manufacturer, and it distracts people from more important issues that the politicians would rather not discuss. When looking for motives, ask: who benefits? I wish people were not so foolish as to think that terrorist attacks can be stopped via airport security. Obviously the terrorists know that airport security exists, so they will attack trains, malls, concerts or busses instead. I also wish people could get it through their heads that terrorist attacks are too rare to worry about. And that the best medicine is prevention (don't breed terrorists through foolish foreign policy).
If a simple pat-down "induced PTSD and sexual abuse trauma", it is more likely to suggest a problem with the passenger rather than the TSA. Even so, America really can't afford billions of dollars in unnecessary equipment and personnel just to provide security theatre, especially since this particular theatre is not the slightest bit entertaining when it happens to you.
And when you can get away with ignoring a court order, isn't that a symptom of a larger problem?
You may be right, at least when it comes to the kind of companies that already squeeze your wallet today.
The reason I really want an easy micropayments system is for the sake of the little guys, the independents, individuals and small businesses, who can't easily convince people to put their credit-card details into their little-known website for a 99-cent payment or donation, and can barely afford the infrastructure to handle logins, payments, etc. And being under the thumb of Paypal isn't ideal.
You don't have to give your money to those price-gougers (assuming no telecom monopoly in your area) and double-dippers who offer high prices and advertising at the same time. Just like I have Netflix instead of cable. Vote with your feet.
I would certainly not advocate automatic micropayments, except if the user explicitly whitelists a page (even then, payments should only be possible in response to user clicks, not auto-refreshes).
The point I wanted to make is that micropayments could be a feasible and fair alternative to advertising, if done right, even something that could fund higher-quality content than advertising can fund, and it's unfortunate that no such system exists. Maybe it's because there are a lot of details to work out, such as how to keep transaction costs far below one cent, but maybe the big boys in the financial industry just don't see profit potential in transactions of less than a penny.
If you're opposed to paying a cent or less per page, you could still go to the many websites that would continue providing service for free.
Besides, micropayments don't have to block access to page content. A page could show its content and still request a one-click 5-cent tip (the one-click, don't-leave-the-page nature of it is the key to convince more people to give tips). The way I envision it, the micropayment button would be part of the web browser, outside the page content, where it is impossible to trick a user into clicking a hidden pay button, or to claim that a different amount will be paid than will actually be paid. The feedback about payments having occurred would also be part of the web browser, making it impossible to hide the fact that there has been a transaction.
I presume at the level of 1-cent transactions, it might be too costly to deal with chargebacks/refunds, so clarity of the user interface would be very important.
But, another way micropayments could work would be that rather than being actual payments, they "buffer up" so that actual money only transfers in minimum amounts of 25 cents or so, an amount large enough to cover the costs of managing refund requests and interfacing with traditional financial systems. In that case, you could essentially visit 24 one-cent pages of a web site "for free" with a real transaction occurring after 25 pages or more. A micropayment standard, then, could just be a cheap buffering system that exists to minimize the large transaction costs of traditional financial systems.
The fine details, such as who would provide the buffering functionality (it can't be the web site itself or the web browser itself, since neither is guaranteed trustworthy) is something I leave to the security/crypto/trust theorists.
What we really need is a micropayment system that makes it feasible for consumers to spend one cent per page or less, to avoid advertising without "subscribing" to a website, without the inconvenience of getting out their credit card, without having to share any private information with websites, without age restrictions (did you have a credit card at 13?). A system in which websites do not have to implement complicated paywall, billing, or log-in infrastructure, do not have to subject themselves to capricious decisions by Paypal, etc.
Until we have such a system, advertising will have to be the main source of revenue in general.
Ideas anyone?
A hot summer does not prove AGW, nor does a cold winter disprove it. In fact, any number of hot summers does not prove AGW; at most they only prove that warming is occurring.
To prove that the Global Warming is Anthropomorphic requires a lot of additional evidence, which has been gathered. There is now a strong concensus among scientists that "man-made" is the only explanation that fits.
I think Valve's main point was that they started with 6 FPS on Linux and are now over 300. I really think their main purpose was to show that their porting effort to Linux/OpenGL is a complete success, there's no need to discuss benchmarking hardware just to tell us that their work is going well. The "dig at Microsoft" you speak of is just editorializing by Slashdot ("MrSeb").
I didn't buy any mainstream music for the past few years because Amazon won't sell music in Canada and I'm boycotting iTunes entirely.
But then a friend referred me to a company willing to take my money, 7digital. Oddly, they don't show up in the first page of Google results for "Buy music in Canada".
From the web site: "Opa is the most advanced application framework for JavaScript. All aspects are directly written in Opa: Frontend code, backend code, database queries and configuration. And everything is strongly statically typed."
So you write code in one language, and your functions are automatically translated to Javascript as needed. The same exact code can run on client and server, but if it runs on the server then it can be optimized better because Opa (unlike Javascript) is statically typed.
Opa appears to be a functional language like Haskell, OCaml, etc., which means that you can write most of the code without specifying data types, as types will be inferred automatically. It also means that data is immutable (write-once) by default. The syntax has a slightly more "conventional" appearance than some other functional languages, so it doesn't look quite as foreign to those who are new to functional programming. Opa is not object-oriented, but it does at least offer "modules" that support dot-notation.
Personally, lack of OO features always makes me uncomfortable because I'm not sure how to use the design patterns I am used to using. IMO the Opa manual should really have a chapter like "Opa for OO dummies" where it explains what the functional equivalent of each of the myriad OO design patterns. And this should be preceded by a "Opa for procedural dummies" chapter, that explains how to replace your efficient procedural code based on hashtables or arrays with Opa code.
Baby steps. The copyright lobby got 95 year copyright terms in the U.S. Even 50 years would be a major improvement, and most all of us slashdot types could agree to reduce copyright to 30 years.
10 years probably suffices for industry to turn a profit, but when you're fighting powerful companies and ideologues who want perpetual copyright, who fear competition from 50-year-old works, who would claim it's an injustice that every single person that reads Mark Twain should pay their great-great-grandkids...
30-year copyright would be an enormous victory. We would have the right to freely play and remix numerous early NES games, to distribute the original Star Wars (the one with the less honorable Han Solo), to have free collections of oldies, disco, Elvis and the Beatles bundled with every new iPod, and so forth.
First of all, reducing terms from 20 to 5 years would reduce the effect of software patents by 75% all by itself.
And they *are* lobbying for more than shorter durations. The EFF's 7 suggested points, taken together, would remove most of the remaining effect of software patents. Like most of us software developers, I expect the EFF would support complete elimination of software patents, but given how patent-friendly Washington is, perhaps they thought it would be a more effective strategy to lobby for weaker patents instead.
Well, let's say there is a kilogram of antimatter floating through space and it hits Earth.
The antimatter is annihilated in an explosion of 180,000 Terajoules of energy. Oh, and some of Earth too.
There can't be much antimatter in the universe because it explodes on contact with any matter it touches. Given e=mc^2, one kilogram of antimatter plus one kilogram of matter equals 2c^2 = 18e16 joules of energy = 180,000 TJ.
IANAP (I am not a physicist, grain of salt etc.)
You may as well ask programmers to never make a mistake, or people in general, for that matter. I remember reading the report on the disaster. It wasn't just that a 16-bit variable overflowed. It overflowed in a noncritical system, which led to the shutdown of the main Inertial Reference System and the backup, leaving nothing to fly the rocket. I don't have the report handy, but it was roughly four problems in combination that brought down the rocket: the bug itself, lack of testing for the bug with altered telemetry (the test telemetry used in simulation was significantly different from the real-life trajectory), failure to handle the exception, and the assumption that an unhandled exception indicated hardware failure (causing the main and backup computer to both shut down). Software always has bugs. One must be sure to do enough testing to find them, and (if failure is obscenely costly) to plan out some sane ways to handle unexpected bugs in the field.
What are these "channels" you speak of? I use Netflix. I just watch whatever I feel like, whenever I feel like it. I only need one channel. The Netflix channel. Anyway, I'd hardly say all shows are "niche" that are not reality shows. There is one reality show I watch--it's about Gordon Ramsay rehabilitating restaurants. How niche is that?
Huh? Metric is based on multiples of 10. So 100 years is a metric century. Maybe you're thinking of 36 years, the imperial yard-century.
No. Just no. This guy stole from Target, not Lego. Why would Lego lower their prices just because there's a shoplifter at Target? Anyway, Lego's patents have expired, and other companies are free to compete.
Something far more basic would work. Ban software patents.
Okay, not every tech lawsuit is about software. Or even about patents. But it would do a lot to reduce these risks and it's an easy change.
P.S. Treble damages is a real legal term, believe it or not. Because "triple" would sound too ordinary?
It seems to me that another important use case is 911 in remote areas. Your reception may not be good enough for a voice call, but it may still be good enough for a text.
That's not really on-topic. Yes, WPF makes life difficult, but Mono for Android doesn't even support WPF. On Android, you'll be using the same widgets in C# as Java developers do.
Not exactly. The bulk of the construction costs are at the last mile. Wires into homes are built regardless of whether a homeowner uses them, and neighborhood-level equipment must be built and operated (at roughly constant cost per house) regardless of how much downloading or uploading occurs. So it doesn't automatically make sense to count local communication toward a bandwidth cap. Instead, the standard monthly fee seems like a fairer way to cover those costs (and the fee is certainly high enough to do so!)
Long-distance links, while also a fixed cost to build, are used at capacity or near capacity; therefore, they tend to be built according to demand and so their cost is dominated by the amount of bandwidth needed instead of by the number of potential customers.
It does, therefore, make some sense for Comcast to "forgive" local traffic since it usually doesn't go though very many saturated internet links. To evaluate fairness, my question would be, is this a special exemption for Xfinity, or is all local traffic (e.g. file sharing, Skype) forgiven, within, say, city limits?
Of course, if the long-distance costs are very small, as you imply, then that just means the long-distance bandwidth cap ought be large, and overage fees small.
Python is a good choice because it offers an interactive interpreter prompt. Or better yet, see if you can acquire the (Javascript) interactive programming tools by Bret Victor, featured in his CUSEC 2012 talk.
An 11 year old can learn from books, too--that's how I learned when I was 11 (most of the books in the library were about BASIC, I thought "integer" was pronounced "inte-geer", and I didn't know that other languages existed for about a year.). But I'm not sure it's necessary in the internet age.
Some have suggested C... give me a break. I started using C when I was 13 and hated it compared to BASIC. I mean, there's no function to read a single key without reading a whole line? No function to check if a key has been pressed? No function for changing the text color or drawing stuff? What's this pointer nonsense? Why do I have to wait for it to "compile" and "link"? I use C++ routinely now, but it's not for beginners.
Despite your probability error, come to think of it you have a better case if you consider not just WC but all the various diseases against which parents are not vaccinating for fear of autism. The combined risk of death from all of those unvaccinated illnesses may be similar to the rate of autism.
But what really matters, of course, is not the rate of autism in the general population, but rather how much the risk of autism drops by not having vaccinations. We know this is roughly 0.0%, but even if you believed in this link, the "safety" gained by not vaccinating must surely be less than the total rate of autism (which is 0.1-0.2% according to Wikipedia and 0.6% for Autism Spectrum Disorder).
Nope, that's an elementary probability mistake. The rate of autism applies to everyone while the rate of death from whooping cough only applies to the people who contract whooping cough, which is probably a small fraction of those who are vaccinated (partly because most people are vaccinated). You have to multiply P(dying from WC) times P(getting WP) before you can compare with the autism rate.