He uses only the mouse, so it is invulnerable to that method, actually. You need to capture the mouse actions and the screen simultaneously. This is something not easily done in separate hardware.
A tired but true argument: Doesn't that go for most weapons? I'd think basically any incendiary weapon would make for an agonizing death.
Besides, this particular device sounds much more along the lines of shaped-charge munitions which have been used for a long time; it just throws some electromagnetics into the mix.
Exactly: Not exactly something to freak out about, is it? Here's a little anecdote, with some background:
My girlfriend is from Poland, and she and her parents laugh at the things we Americans do. She thinks we're paranoid and prudish about sexuality and race, and should really just lighten up and stop taking ourselves so seriously.
It's funny, because actually Poland is more conservative in many ways. It's not as though the sexual revolution hit Poland and ever since it's been a rollicking orgy!! But people just don't freak out if a little kid is naked. For them, being naked is just something that little kids often do, and nothing particularly to worry about.
My favorite example is this: When her parents first moved here, they were a little surprised to see that it was considered indecent for a little girl -- say, 5 yrs old -- to be at the swimming pool without a swimsuit top. You see, from their point of view, it was nonsense: "What are you covering up? There's nothing to hide yet!" In fact, I wonder if requiring little girls to wear clothes that imply breasts only sexualizes them more, and American prudishness is just counterproductive in this respect. Of course, her parents are very laid back people, so for them the answer was just to laugh, shake their heads, and go buy an American-style swimsuit.
I think they've got the right attitude. We shouldn't be so paranoid.
You'd still be moving the document, of course -- just, in this case, as a bitmap -- possibly JPEG-compressed. And if you're X-forwarding, then the text is actually available, in fact.
The problem is basic to the technology, but I think it's much more fundamental.
Analog electronics had a problem: Data was degraded as it was processed. Digital electronics solved the problem -- by copying the data in order to restore it at each step. Copying is inherent to the nature of digital technology. The minute I give you a D-flip-flop, you can copy a bit.
Adding layers of abstraction -- like "we'll send bitmap snapshots of the document to you instead of an ASCII stream" -- complicates matters a little, and might foil casual copying, but it doesn't fundamentally change the situation.
That said, I think you may be on to something. Crypto took a giant leap forward when people said, "let's use a crypto function that everybody understands, but which is simply hard to invert." Likewise, perhaps data management problems can be solved by saying, "We acknowledge that you can copy whatever bits we give you; we're just going to choose a horrendously inefficient coding scheme (like representing text documents as bitmaps) so that there are too many bits for you to copy using the available time and bandwidth without someone noticing." In fact, this even has some mathematical foundation: There is currently no way to determine the Kolmogorov complexity of a sequence. (Although we have some good compression algorithms, they still make assumptions which can be easily broken: Try dumping the output of a million calls to rand() to a file and zipping it. It'll be huge, even though you can represent the sequence with just the definition of your rand() function and the seed you used). I'm thinking this basic idea is a large part of the MPAA's motivation the move to higher and higher HD, for instance; in the extreme, they could give up on encryption, and replace it with a known nontrivial problem: Downsampling and recoding video. It's not quite the same magnitude as factoring products of large primes, but it's still a computational pain in the butt when you're talking about a 50GB Blu-Ray disc.
Another question, though, would be this: Is copying documents the main security issue companies face?
I'd say that La Marseillaise would give it a run for its money, but that one is just a bit too over-the-top to be taken seriously:
Arise, children of the Fatherland,
The day of glory has arrived!
Against us, tyranny's
Bloody banner is raised,
Bloody banner is raised.
Do you hear in the countryside
The braying of these ferocious soldiers?
They are coming into our midst
To cut the throats of our sons, our wives!
To arms, citizens!
Form your battalions!
March, march!
May impure blood,
Water our fields!
Yep. By comparison, the Soviet anthem sounded a lot nobler.
I've seen both. I did my undergrad at an Ivy, and am currently doing grad work at a top-4 engineering school.
Undergraduate education -- including in technical subjects -- was better at the Ivy.
I'm currently a TA at the Big Tech where I teach an electronics course, and although the kids are generally bright, it's also clear that they haven't been prepared very well by their previous courses. They seem too worn down by the grind, and the focus on grades and performance almost distracts from the actual learning. You know the premed stereotype: That they memorize everything and understand nothing -- because when the stakes are as high as getting into med school, thinking is a liability? There's some truth to it, and I've seen a similar thing happening here. The whole experience has been funny in a way, as there'd always been a part of me that had suspected that I would somehow have known more if I'd gone to a Tech: That suspicion has completely gone away. I have a lot of confidence now in my undergraduate education.
So I'm going to say this: Think about the incentives that professors have at each school. At the Big Tech, research comes first. (That's why you go there for grad school!.) At the small liberal arts school, undergrads will be the focus.
You want to be the focus, so go to the liberal arts school.
Semicolons do roughly the same thing as periods; they're just not a full stop. They're wonderful things because they allow you to imply logical connections between ideas without getting bogged down in really pedantic conjunctions. Just take two clauses and put a semicolon in-between; you've just implied a relationship between them!
Embrace the semicolon! Don't overdo it of course, but that goes for most things in life!
The idealized Stirling cycle can achieve the Carnot efficiency, which is the maximum theoretical efficiency of any heat engine.
There are idealized thermodynamic cycles that describe many heat engines. The Otto Cycle describes what an idealized gasoline engine does; there are also idealized Diesel, Brayton (gas-turbine), and of course Stirling cycles. They're great for getting a high-level first-principles understanding of what's going on, but they all neglect important effects.
So, there is an element of the ideal Stirling cycle which is not achievable in real life: Isothermal expansion and compression. This requires that the working fluid remain at the same temperature as either the hot or cold reservoir as it expands or is compressed, and this is only achievable in the limit as you perform the expansion or compression infinitely slowly.
Screens are measured along the diagonal. What most people look at is this number, so it sets the price.
Now, it's a simple exercise in Calc 101 to show that, for a given diagonal length, the rectangle that maximizes area is a square.
The longer and skinnier the screen is, the more (area)/(diagonal) shrinks. In the limit, you're left with a line segment with zero area.;-)
But on the plus side, 16:9 is closer to the golden ratio than 4:3.
(My understanding is that the real reason widescreen displays are cheap is that the screens are "rolled out", and it's easy to "roll out" a longer strip than to build a wider roller. Someone who knows more about the industry may like to comment on that.)
The UAC API is a horrible piece of junk. Here's what happened one day when I tried writing a Vista sudo for Cygwin, once upon a time...
Backstory first:
I was used to running Cygwin on XP, which I like very much (and think is a great combination for getting stuff done). When I got a laptop with Vista, I found that a lot of the GNU tools on Cygwin simply wouldn't work if UAC was on; they simply returned an error, something like, "Permission denied." I wouldn't have minded if the programs had triggered a UAC elevation; I'd have seen that as akin to sudo. But instead, they just flat-out failed.
It seems that programs on Vista do not automatically raise UAC when they attempt to do something that requires elevated privileges. So I asked, "Can I make a program -- I'll call it 'sudo' -- that triggers UAC and then runs another program with the elevated privileges?"
It turns out that the answer is "not really." (I know scripts exist that people call 'sudo for Vista,' but they don't do quite what I wanted; I'll get to that in a second). (EDIT: it may actually be possible, through a somewhat convoluted process involving a number of different EXEs and DLLs with appropriate manifests. I'll get to that at the end. But it's certainly not something provided in any sane way by the API.)
The best way to explain my goals for a Cygwin 'sudo' is with a simple example:
cd/cygdrive/c/Program\ Files/ # Some protected directory
sudo mv a.txt b.txt
#***Vista UAC Prompt pops up; I click OK.***
# (file has been successfully moved)
This seems useful, no? It would be a way to keep UAC, yet also use the commandline tools it currently cripples.
Now, as I mentioned before, there do exist various scripts calling themselves 'sudo for Vista,' but none of them really achieve the above. Here's why: Rather than running mv in the same terminal, they pop up another terminal on top running mv. This sort of does what you want, but not quite -- and subtly breaks a lot of things: For a simple-if-not-compelling example, it's impossible with this scheme to run one program with "sudo" and pipe its output to an un-elevated program (one run without sudo).
At the heart of the problem is the fact that, at the end of the day, there is only one nice way to get UAC out of Vista, and it is a most inflexible one: The ShellExecuteEx() function -- essentially, this is what gets called when you double-click on something in Explorer.
That's a slight oversimplification: There are some other obnoxious hoops you can jump through to get UAC [changing manifests (What's up with that? Tell me how to do that with gcc.), some COM garbage, or simply -- and this is a little silly -- including the word 'setup' or 'install' in your executable's filename], but as far as I could tell they all take you to roughly the same place.
(EDIT: It turns out that there might be just enough wiggle-room to get slightly different results from these different approaches.)
Eventually, frustrated, I gave up.
My conclusion was that the Vista UAC API was a horrible piece of garbage, as this sort of thing is not terribly difficult to achieve on Linux.
EDIT: It seems that, since that day, someone else may have succeeded where I failed. I'll need to try out his solution myself before I can be completely sure that it's what I want, but what I see looks very good. If so, then the author -- Thomas Hruska -- deserves kudos for figuring out a very clever workaround. But I think the very fact that such a workaround is necessary at all merely reinforces my original point that the UAC API is a steaming mess.
Mmm, indeed. What better way to increase U.S. test scores than with concentration-enhancing uppers?
The funny thing is that I was actually being serious about gcc and Firefox! Learning to program changes the way you think and helps to motivate much of math. Growing up it had always been recognized that my chief skills were verbal, not quantitative (I still have trouble with arithmetic), yet now I'm embarking on a PhD in control theory -- a very mathematical discipline! Teaching myself to program literally changed my mind and with it my life.
Give a kid a compiler, and show him John Carmack for a hero -- screw famous athletes -- and see what happens.
In fact, I'll be even more specific: Give a kid,
1 - A simple programming language. I started with QBasic. I hear Python might be a good modern-day language for beginners.
2 - Graphics libraries. The simpler the better. Save the OpenGL for later; in the beginning, give kids a fast inline putpixel. That sure was enough to capture my imagination. What better way to learn about parametric equations -- and be excited about them -- than to start out wanting to draw circles, and arrive naturally at the intuitive x=r*cos(t); y=r*sin(t) answer?
Video games are inherently mathematical objects, and graphics programming is the way to get kids excited about math. Forget this sugar-coated "educational software;" it just bores kids. Introduce them to programming, and from there CS; you'll capture many more imaginations that way than you ever can with Mavis Beacon Teaches Typing.
The point about efficiency in my previous post (parent) is incorrect!
Correction: Pebble-bed reactors have good thermal efficiencies.
At first I read this paper in which they seem to be working with pebbles at 900 C, and I had thought "that's barely lukewarm by powerplant standards."
The problem is I'd been thinking in Fahrenheit, American that I am. We're talking about 1600 F, which is pretty darn hot. Hotter, in fact, than combustion plants, which typically run at about 1000 to 1200 F.
So pebble-bed reactors have higher thermal efficiencies!
That only leaves the waste-management issue, which pebble-beds exacerbate. I'll need to think about this a bit more now before I form an opinion.
But I'm not convinced about the pebble-bed design. It strikes me as wasteful, for two big reasons:
1 - You produce a whole lot of radioactive ceramic coatings (the pebble casings) that you don't produce in a conventional design.
2 - (The bigger reason): The operating temperature is low by comparison even to combustion power plants, which means that the thermal efficiency is poor. Think about Carnot efficiency, and remember that nuclear plants are just heat engines, in the end!
One additional observation: I agree that nuclear gives us a solution to the problem -- but only if we use breeder reactors. (This is mostly orthogonal to the pebble bed discussion, because a quick google turns up that it's quite possible to build pebble-bed breeders.) Conventional nuclear will use up our U235 within a small number of human generations; but there's tons of U238 (and yet more Thorium).
My personal vote would go, I think, to "normal" breeders. But there's clearly more work to be done in this field, as liquid sodium is nasty stuff, and the alternatives (lead, water) have their own problems.
Indeed. In Photoshop, I've done this with photos of my own face. It looks like CG, or an alien's attempt at producing a human face. This could, admittedly, be less to do with symmetry and more to do with unnatural lighting (if one side is lit, you'd expect the other to be slightly shadowed.) Nevertheless, I think it's interesting, so you might want to try it sometime with your favorite photo editor!
Is it really less efficient? It's true that you have more data points than you "really need" in the sensor. But since you combine them by averaging, you would expect this to increase your signal-to-noise ratio, and give you more effective bits of precision.
In short, I think one needs to sit down and do a thorough information-theoretic analysis of this scheme, because it's not obvious (to me) that it's actually less (or more) efficient.
The point is to actually capture the red data, the green data, and the blue data from the scene. Sure, I can take a grayscale image of a scene and then artificially tint it, but that doesn't actually tell me anything about how much red, or blue, or green there really is. The original poster isn't trying to tint the image; he's trying to capture the red, green, and blue data from it. And he recognizes that he can do this at higher resolution using three shots with a monochrome sensor and solid-color filters than he can with a single shot using a Bayer pattern filter.
Indeed. What's the date on that project? I didn't see one on the website. Me, I had posted a similar idea on the Halfbakery years ago: Here it is. I guess this is an idea lots of people come up with?
The novel part here, I'd say, is micromachining the thing on the die.
The article says that there were 1,485,280 Apache defacements and 815,119 IIS defacements. This implies a total of 2,300,399 samples, of which 64.6% were Linux. For comparison, other posters here have cited a Google survey reporting that 60% of webservers run Apache. That would seem to imply that, if you pick an IIS server at random or an Apache server at random, each is about as likely to be successfully attacked as the other.
Conclusion: IIS is just as good as Apache (contrary to popular Slashdot opinion). Of course, there's a flip side: Apache is just as good as IIS -- and it's free.
[Take all this modulo the fact that 370% of statistics are, if not made up on the spot, at least full of so much noise as to be meaningless. (Sometimes the Law of Large Numbers really does require large numbers!]
Slashdot Mirror
He uses only the mouse, so it is invulnerable to that method, actually. You need to capture the mouse actions and the screen simultaneously. This is something not easily done in separate hardware.
A tired but true argument: Doesn't that go for most weapons? I'd think basically any incendiary weapon would make for an agonizing death.
Besides, this particular device sounds much more along the lines of shaped-charge munitions which have been used for a long time; it just throws some electromagnetics into the mix.
Brilliant!!!
You need to be careful though: Do you know that hot "18-year-old" is really legal?
You should fill your desktop with nothing but grannies, just to be safe.
Exactly: Not exactly something to freak out about, is it? Here's a little anecdote, with some background:
My girlfriend is from Poland, and she and her parents laugh at the things we Americans do. She thinks we're paranoid and prudish about sexuality and race, and should really just lighten up and stop taking ourselves so seriously.
It's funny, because actually Poland is more conservative in many ways. It's not as though the sexual revolution hit Poland and ever since it's been a rollicking orgy!! But people just don't freak out if a little kid is naked. For them, being naked is just something that little kids often do, and nothing particularly to worry about.
My favorite example is this: When her parents first moved here, they were a little surprised to see that it was considered indecent for a little girl -- say, 5 yrs old -- to be at the swimming pool without a swimsuit top. You see, from their point of view, it was nonsense: "What are you covering up? There's nothing to hide yet!" In fact, I wonder if requiring little girls to wear clothes that imply breasts only sexualizes them more, and American prudishness is just counterproductive in this respect. Of course, her parents are very laid back people, so for them the answer was just to laugh, shake their heads, and go buy an American-style swimsuit.
I think they've got the right attitude. We shouldn't be so paranoid.
You'd still be moving the document, of course -- just, in this case, as a bitmap -- possibly JPEG-compressed. And if you're X-forwarding, then the text is actually available, in fact.
The problem is basic to the technology, but I think it's much more fundamental.
Analog electronics had a problem: Data was degraded as it was processed. Digital electronics solved the problem -- by copying the data in order to restore it at each step. Copying is inherent to the nature of digital technology. The minute I give you a D-flip-flop, you can copy a bit.
Adding layers of abstraction -- like "we'll send bitmap snapshots of the document to you instead of an ASCII stream" -- complicates matters a little, and might foil casual copying, but it doesn't fundamentally change the situation.
That said, I think you may be on to something. Crypto took a giant leap forward when people said, "let's use a crypto function that everybody understands, but which is simply hard to invert." Likewise, perhaps data management problems can be solved by saying, "We acknowledge that you can copy whatever bits we give you; we're just going to choose a horrendously inefficient coding scheme (like representing text documents as bitmaps) so that there are too many bits for you to copy using the available time and bandwidth without someone noticing." In fact, this even has some mathematical foundation: There is currently no way to determine the Kolmogorov complexity of a sequence. (Although we have some good compression algorithms, they still make assumptions which can be easily broken: Try dumping the output of a million calls to rand() to a file and zipping it. It'll be huge, even though you can represent the sequence with just the definition of your rand() function and the seed you used). I'm thinking this basic idea is a large part of the MPAA's motivation the move to higher and higher HD, for instance; in the extreme, they could give up on encryption, and replace it with a known nontrivial problem: Downsampling and recoding video. It's not quite the same magnitude as factoring products of large primes, but it's still a computational pain in the butt when you're talking about a 50GB Blu-Ray disc.
Another question, though, would be this: Is copying documents the main security issue companies face?
What are those?
I got, "so, here's the laser."
(Now I'm in a different lab where there is nothing more dangerous than a few small robots which present a minor tripping hazard.)
That's the best national anthem ever.
I'd say that La Marseillaise would give it a run for its money, but that one is just a bit too over-the-top to be taken seriously:
Arise, children of the Fatherland,The day of glory has arrived!
Against us, tyranny's
Bloody banner is raised,
Bloody banner is raised.
Do you hear in the countryside
The braying of these ferocious soldiers?
They are coming into our midst
To cut the throats of our sons, our wives!
To arms, citizens!
Form your battalions!
March, march!
May impure blood,
Water our fields!
Yep. By comparison, the Soviet anthem sounded a lot nobler.
Y'know, there's even a documentary on the subject...
Ah! I'd read your correction and thought it was the original. My apologies; you are correct.
GOOD: "I am unintimidated by litigation; I sometimes rather miss it." (as you wrote)
LESS GOOD: "Not only am I unintimidated by litigation; I sometimes rather miss it."
That said, I don't think it's a tremendous sin. It's a use of the semicolon I could forgive.
I've seen both. I did my undergrad at an Ivy, and am currently doing grad work at a top-4 engineering school.
Undergraduate education -- including in technical subjects -- was better at the Ivy.
I'm currently a TA at the Big Tech where I teach an electronics course, and although the kids are generally bright, it's also clear that they haven't been prepared very well by their previous courses. They seem too worn down by the grind, and the focus on grades and performance almost distracts from the actual learning. You know the premed stereotype: That they memorize everything and understand nothing -- because when the stakes are as high as getting into med school, thinking is a liability? There's some truth to it, and I've seen a similar thing happening here. The whole experience has been funny in a way, as there'd always been a part of me that had suspected that I would somehow have known more if I'd gone to a Tech: That suspicion has completely gone away. I have a lot of confidence now in my undergraduate education.
So I'm going to say this: Think about the incentives that professors have at each school. At the Big Tech, research comes first. (That's why you go there for grad school!.) At the small liberal arts school, undergrads will be the focus.
You want to be the focus, so go to the liberal arts school.
I hate to point out the obvious, but that's half the point. The more he makes Monsters' lawyers work, the less they want to bother harassing him.
My understanding is that this sort of thing is common practice.
*sigh*
It is perfectly good English.
Semicolons do roughly the same thing as periods; they're just not a full stop. They're wonderful things because they allow you to imply logical connections between ideas without getting bogged down in really pedantic conjunctions. Just take two clauses and put a semicolon in-between; you've just implied a relationship between them!
Embrace the semicolon! Don't overdo it of course, but that goes for most things in life!
The idealized Stirling cycle can achieve the Carnot efficiency, which is the maximum theoretical efficiency of any heat engine.
There are idealized thermodynamic cycles that describe many heat engines. The Otto Cycle describes what an idealized gasoline engine does; there are also idealized Diesel, Brayton (gas-turbine), and of course Stirling cycles. They're great for getting a high-level first-principles understanding of what's going on, but they all neglect important effects.
So, there is an element of the ideal Stirling cycle which is not achievable in real life: Isothermal expansion and compression. This requires that the working fluid remain at the same temperature as either the hot or cold reservoir as it expands or is compressed, and this is only achievable in the limit as you perform the expansion or compression infinitely slowly.
Hope that helps.
Screens are measured along the diagonal. What most people look at is this number, so it sets the price.
Now, it's a simple exercise in Calc 101 to show that, for a given diagonal length, the rectangle that maximizes area is a square.
The longer and skinnier the screen is, the more (area)/(diagonal) shrinks. In the limit, you're left with a line segment with zero area. ;-)
But on the plus side, 16:9 is closer to the golden ratio than 4:3.
(My understanding is that the real reason widescreen displays are cheap is that the screens are "rolled out", and it's easy to "roll out" a longer strip than to build a wider roller. Someone who knows more about the industry may like to comment on that.)
The UAC API is a horrible piece of junk. Here's what happened one day when I tried writing a Vista sudo for Cygwin, once upon a time...
Backstory first:
I was used to running Cygwin on XP, which I like very much (and think is a great combination for getting stuff done). When I got a laptop with Vista, I found that a lot of the GNU tools on Cygwin simply wouldn't work if UAC was on; they simply returned an error, something like, "Permission denied." I wouldn't have minded if the programs had triggered a UAC elevation; I'd have seen that as akin to sudo. But instead, they just flat-out failed.
It seems that programs on Vista do not automatically raise UAC when they attempt to do something that requires elevated privileges. So I asked, "Can I make a program -- I'll call it 'sudo' -- that triggers UAC and then runs another program with the elevated privileges?"
It turns out that the answer is "not really." (I know scripts exist that people call 'sudo for Vista,' but they don't do quite what I wanted; I'll get to that in a second). (EDIT: it may actually be possible, through a somewhat convoluted process involving a number of different EXEs and DLLs with appropriate manifests. I'll get to that at the end. But it's certainly not something provided in any sane way by the API.)
The best way to explain my goals for a Cygwin 'sudo' is with a simple example:
Suppose I attempted the following:
cd /cygdrive/c/Program\ Files/ # Some protected directory
mv a.txt b.txt
Error: Permission Denied
I would want to be able to instead do,
cd /cygdrive/c/Program\ Files/ # Some protected directory
sudo mv a.txt b.txt
#***Vista UAC Prompt pops up; I click OK.***
# (file has been successfully moved)
This seems useful, no? It would be a way to keep UAC, yet also use the commandline tools it currently cripples.
Now, as I mentioned before, there do exist various scripts calling themselves 'sudo for Vista,' but none of them really achieve the above. Here's why: Rather than running mv in the same terminal, they pop up another terminal on top running mv. This sort of does what you want, but not quite -- and subtly breaks a lot of things: For a simple-if-not-compelling example, it's impossible with this scheme to run one program with "sudo" and pipe its output to an un-elevated program (one run without sudo).
At the heart of the problem is the fact that, at the end of the day, there is only one nice way to get UAC out of Vista, and it is a most inflexible one: The ShellExecuteEx() function -- essentially, this is what gets called when you double-click on something in Explorer.
That's a slight oversimplification: There are some other obnoxious hoops you can jump through to get UAC [changing manifests (What's up with that? Tell me how to do that with gcc.), some COM garbage, or simply -- and this is a little silly -- including the word 'setup' or 'install' in your executable's filename], but as far as I could tell they all take you to roughly the same place.
(EDIT: It turns out that there might be just enough wiggle-room to get slightly different results from these different approaches.)
Eventually, frustrated, I gave up.
My conclusion was that the Vista UAC API was a horrible piece of garbage, as this sort of thing is not terribly difficult to achieve on Linux.
EDIT: It seems that, since that day, someone else may have succeeded where I failed. I'll need to try out his solution myself before I can be completely sure that it's what I want, but what I see looks very good. If so, then the author -- Thomas Hruska -- deserves kudos for figuring out a very clever workaround. But I think the very fact that such a workaround is necessary at all merely reinforces my original point that the UAC API is a steaming mess.
Mmm, indeed. What better way to increase U.S. test scores than with concentration-enhancing uppers?
The funny thing is that I was actually being serious about gcc and Firefox! Learning to program changes the way you think and helps to motivate much of math. Growing up it had always been recognized that my chief skills were verbal, not quantitative (I still have trouble with arithmetic), yet now I'm embarking on a PhD in control theory -- a very mathematical discipline! Teaching myself to program literally changed my mind and with it my life.
Give a kid a compiler, and show him John Carmack for a hero -- screw famous athletes -- and see what happens.
In fact, I'll be even more specific: Give a kid,
1 - A simple programming language. I started with QBasic. I hear Python might be a good modern-day language for beginners.
2 - Graphics libraries. The simpler the better. Save the OpenGL for later; in the beginning, give kids a fast inline putpixel. That sure was enough to capture my imagination. What better way to learn about parametric equations -- and be excited about them -- than to start out wanting to draw circles, and arrive naturally at the intuitive x=r*cos(t); y=r*sin(t) answer?
Video games are inherently mathematical objects, and graphics programming is the way to get kids excited about math. Forget this sugar-coated "educational software;" it just bores kids. Introduce them to programming, and from there CS; you'll capture many more imaginations that way than you ever can with Mavis Beacon Teaches Typing.
1 - gcc
2 - Firefox (w. Google & Wikipedia)
When their powers combine, you can build anything. It's like Legos on crack. And who doesn't like Legos?
The point about efficiency in my previous post (parent) is incorrect!
Correction: Pebble-bed reactors have good thermal efficiencies.
At first I read this paper in which they seem to be working with pebbles at 900 C, and I had thought "that's barely lukewarm by powerplant standards."
The problem is I'd been thinking in Fahrenheit, American that I am. We're talking about 1600 F, which is pretty darn hot. Hotter, in fact, than combustion plants, which typically run at about 1000 to 1200 F.
So pebble-bed reactors have higher thermal efficiencies!
That only leaves the waste-management issue, which pebble-beds exacerbate. I'll need to think about this a bit more now before I form an opinion.
I agree with you for the most part.
But I'm not convinced about the pebble-bed design. It strikes me as wasteful, for two big reasons:
1 - You produce a whole lot of radioactive ceramic coatings (the pebble casings) that you don't produce in a conventional design.
2 - (The bigger reason): The operating temperature is low by comparison even to combustion power plants, which means that the thermal efficiency is poor. Think about Carnot efficiency, and remember that nuclear plants are just heat engines, in the end!
One additional observation: I agree that nuclear gives us a solution to the problem -- but only if we use breeder reactors. (This is mostly orthogonal to the pebble bed discussion, because a quick google turns up that it's quite possible to build pebble-bed breeders.) Conventional nuclear will use up our U235 within a small number of human generations; but there's tons of U238 (and yet more Thorium).
My personal vote would go, I think, to "normal" breeders. But there's clearly more work to be done in this field, as liquid sodium is nasty stuff, and the alternatives (lead, water) have their own problems.
There are stats on how many companies are led by engineers, and those statistics are pretty impressive.
But, not being a detail person, I won't bother to look up the list. ;-)
Indeed. In Photoshop, I've done this with photos of my own face. It looks like CG, or an alien's attempt at producing a human face. This could, admittedly, be less to do with symmetry and more to do with unnatural lighting (if one side is lit, you'd expect the other to be slightly shadowed.) Nevertheless, I think it's interesting, so you might want to try it sometime with your favorite photo editor!
Is it really less efficient? It's true that you have more data points than you "really need" in the sensor. But since you combine them by averaging, you would expect this to increase your signal-to-noise ratio, and give you more effective bits of precision.
In short, I think one needs to sit down and do a thorough information-theoretic analysis of this scheme, because it's not obvious (to me) that it's actually less (or more) efficient.
The point is to actually capture the red data, the green data, and the blue data from the scene. Sure, I can take a grayscale image of a scene and then artificially tint it, but that doesn't actually tell me anything about how much red, or blue, or green there really is. The original poster isn't trying to tint the image; he's trying to capture the red, green, and blue data from it. And he recognizes that he can do this at higher resolution using three shots with a monochrome sensor and solid-color filters than he can with a single shot using a Bayer pattern filter.
Indeed. What's the date on that project? I didn't see one on the website. Me, I had posted a similar idea on the Halfbakery years ago: Here it is. I guess this is an idea lots of people come up with?
The novel part here, I'd say, is micromachining the thing on the die.
The article says that there were 1,485,280 Apache defacements and 815,119 IIS defacements. This implies a total of 2,300,399 samples, of which 64.6% were Linux. For comparison, other posters here have cited a Google survey reporting that 60% of webservers run Apache. That would seem to imply that, if you pick an IIS server at random or an Apache server at random, each is about as likely to be successfully attacked as the other.
Conclusion: IIS is just as good as Apache (contrary to popular Slashdot opinion). Of course, there's a flip side: Apache is just as good as IIS -- and it's free.
[Take all this modulo the fact that 370% of statistics are, if not made up on the spot, at least full of so much noise as to be meaningless. (Sometimes the Law of Large Numbers really does require large numbers!]