Yeah, that's one way of going about it. The other way to look at it is that if all it takes is one employee with an infected device to fry your network, your network is in a pretty sorry state.
I work in medical research. My previous lab was on a hospital network. One day someone, somewhere in the hospital brought in a notebook with a virus. Most of the machines in the hospital went down, including one of the MR scanner consoles. It was a huge crisis. Our lab barely noticed -- we were running Macs. Our Windows terminal server was properly patched and firewalled.
Hospital IT responded by cracking down on outside devices but NOT really tightening up security on individual machines. Of course, if someone, either with malicious intent or by mistake, plugged an infected laptop into the network, they would be right back at square one.
Perhaps I shouldn't have said "universal" because you might be tempted to believe that means every person, everywhere, for all time, agrees. There are lots of morals that are universal in terms of societies. For example, all human societies have injunctions against killing. They do differ in the details, but the basic idea seems to be always present. Most individuals also carry this concept around in their heads, and those who don't are usually judged to be mentally ill.
There is no such thing as absolute morals in the way you seem to be implying. Even if a god did dictate them, someone, somewhere would disagree.
Yes, obviously some of our second line morals are cultural. I didn't say all moral beliefs everywhere can be determined solely by evolutionary arguments. Even so, many of those cultural beliefs are rooted in more fundamental morals that can be scientifically justified. Let's use your example of stealing. A culture that has no concept of possession obviously can't have a concept of stealing, but they may well have a moral injunction against grabbing something from your neighbour while he's using it. Most of us probably learned that in the communal environment of kindergarten. Thou shalt not steal and thou shalt not be grabby are just reflections of the same underlying moral, modified slightly by particular environmental factors. What about the explanation for the underlying moral? Well, anyone who's ever grabbed/stole food from a hungry animal while it is eating is likely to be familiar with the real and immediate physical consequences.
There is no reason what we call morals shouldn't have a rational, scientific explanation. Just in case you're only being pedantic, note that I didn't say evolutionary arguments are morals, but that they explain our morals. And the study of ethics can very much be (and should be) based on science. That's actually one of the basic principles of secular humanism.
Take your example. The moral: you ought not kill (your own kind, except in certain circumstances). Why not? Because a book says so? Why does that book say so? More importantly, why do ALL our ethical systems (written in books or otherwise) tend to generally agree on this point? Even many species of animal generally adhere to this principle, although the details differ. Why?
The answer is fairly obvious from a careful consideration of environmental pressures. Killing, except to eat, protect yourself, your kin, or certain other exceptions, is usually a genetic disadvantage. You take a large risk for very little gain.
Secular humanism promotes scientific testing of ethical or moral assertions and should presumably welcome scientific explanations of why particular moral beliefs are advantageous. However, many religious people, and many so-called secular humanists seem to detest the idea that our morals have such inglorious origins as instinct and natural selection.
While that may be so, the letter certainly appears to me to be a general statement of policy, perhaps in response to a particular case, but not limited to it.
Anyway, the thing to do is to put him on trial and figure out exactly who did what. Maybe he's innocent. Let's find out.
When the abuse was happening what do you suppose the pope (this one) was doing at the time? There are documents that indicate he (not as a local bishop, but as a cardinal giving those local bishops orders) was actively participating in the cover up.
So yeah, let's follow your advice and come down hard on everybody responsible. If you're a Catholic you should be screaming for his head.
Absolutely. But when the organization gets caught, you don't reason it away. You come down hard on them, and clean house. You certainly don't promote one of the worst offenders to chief (pope) and let him stay there.
You don't need a god to dictate fixed morals anyway. There are ample evolutionary arguments for why we universally consider certain behaviour bad. The pope seems to have the worst case of moral relativism anyway - stealing crackers is worse than genocide and child abuse isn't all THAT bad... at least if a priest is doing it.
You're right, I was careless when I said "aperture." In photography that word usually means "relative aperture," or f-stop, which is defined as N=f/D (N is the f-stop, f is the focal length and D is the diameter of the light permitting pupil). I really meant D, the diameter of the light permitting pupil.
Take an example. Suppose I have a 100 mm lens that is perfectly paired with a 24x36 mm (standard 35 mm full frame sensor). If this lens is set at f8 the diameter of the light permitting pupil will be:
N = f/D D = f/N D = 100 / 8 = 12.5 mm
This combination will have a field of view (I'm just going to calculate the small dimension, but you can feel free to do the other if you wish) of:
where d is the size of the sensor in the chosen direction.
Now, take the case of a Canon 1.6x crop factor sensor. We're going to manufacture a lens that is just like our other lens in terms of light permitting pupil size and field of view (and thus magnification) but forms an image that is perfectly paired to the smaller sensor. Of course, this means the focal length will have to change:
f = d / (2tan(alpha/2)) = 62.5 mm
Note that 62.5 mm * 1.6 is 100 mm, as we would expect. So what f-stop are we using on the new lens?
N = 62.5 / 12.5
= 5
As you point out, this means that the light intensity on our smaller sensor is greater than on our bigger sensor. If you care to work it through you'll find that, if you divide the smaller sensor into the same number of pixels as the larger, the total light incident on any given pixel-sized area is the same.
This can be shown more easily by considering total incident light. The total number of photons allowed into the camera will be proportional to the area of the light permitting pupil. In our two examples this is:
Full frame case:
A = pi*(D/2)**2 = pi * (12.5 / 2)**2 = 122.7 mm^2
Note that the small sensor case is the same because the pupil is the same diameter. Call the amount of light entering the camera through this pupil 1 unit.
Now, in the full frame case this 1 unit of light is spread over an area of 24x36 mm = 864 mm^2. The incident light on this sensor has an intensity of 1/864 units/mm^2. If we divide the sensor into 10 megapixels (it's a perfect sensor with no wastage), each will have an area of 864 / 10^6 mm and every pixel will get 1 / 10^6 units of light.
Okay, what about the small frame sensor? We have the same amount of light coming into the box, but this time it's spread over a 22.5 mm × 15 mm sensor (area = 337.5 mm^2). The incident light on the sensor has an intensity of 1/337.5 units/mm^2. Note that it is more intense, as we expect from our f-stops. Again, divide the smaller sensor into 10 megapixels, each with an area of 337.5 / 10^6. Again, each pixel will get 1 / 10^6 units of light.
That is, the larger and smaller sensors, with properly designed lenses and neglecting wasted space on the sensor, get equal amounts of light per pixel. So what changes? The focal length and the f-stop do. But the physical dimensions of the lens (except for the thickness) do not change. So for a given light gathering area and a properly designed lens, neglecting wasted space on the sensor and non-dominant sources of noise that do not scale with sensor size, noise performance isn't affected by the size of the sensor.
But wait! The apertures aren't the same! Well, aperture in digital photography doesn't mean what it did in film photography. Given a piece of film, whether 35 mm size or large format, for the same film design and chemistry a 1/100 s @ f8 shot will give the same exposure no matter the format because the film grains (the "pixels") are all the same size. The situation is NOT the same in digital photography. Since the pixels change size in digital sensors depending on the size of the sensor, f8 for one sensor may not be the same as f8 for another,
It does take a certain amount of time. However, you can e-mail your photos to the office and someone can do it on a computer there, or you can do it on your own laptop on your way back from the shoot, or in the hotel room that night. Anyone reasonably familiar with the process can take a RAW and spit out something at least as good as a first-pass developing/printing operation in ten or fifteen seconds.
Compare that to film processing. It requires chemicals (that have a limited shelf life, particularly for colour), a special room, other special equipment, parts of it have to be done in complete darkness by feel (or automatically by a very expensive machine), the reactions take several minutes and once you're done you have to wait for everything to dry. Then you have to scan the negatives and you're right back with your RAW image file.
All that is if you even have a darkroom (never mind a medium format capable darkroom). If not you have to send everything away and wait for the mail.
I very, very rarely enter the menu system on my 30D. Certainly not to adjust shutter speed or aperture. There's one dial for shutter speed, another for aperture. Pretty simple. Are you sure you've used a digital SLR?
If you really want to use your old lenses, go ahead. Depending on the age and make of your lens you can either a) just stick it right on a modern SLR from the same manufacturer b) buy a simple adapter ring and go to (a) or c) file off a particular bit and go to (a).
"Secondly as I understand it (i've done a little bit on optics but i'm not an expert) depth of field is related to the ratio between aperture and sensor size."
More or less, but the relationship is the opposite of what you've said. A bigger aperture and/or a bigger sensor means less depth of field. Small sensor cameras can be desirable for things like macro photography where you're normally starved for depth of field, or landscape where you want everything in focus. They're undesirable for things like portraits, where you want to make sure the background is well blurred by achieving a narrow depth of field.
One of the reasons compact and cell phone camera shots look amateurish is because the depth of field is very wide, so everything is in focus. More professional looking photos generally use narrower depths of field.
A bigger sensor doesn't do you the least bit of good without a bigger lens, generating a bigger image, to go with it.
Neglecting certain inefficiencies, a small sensor with a big lens will give you just as good an image (in terms of noise) as a big sensor with a big lens. The aperture determines how much light is gathered and focused on the sensor.
Who lost billions? Do you suppose the airlines gave back all the money they were paid for tickets? Even if they did, do you suppose all the people who were stranded decided to stay where they were and not fly home when flights started up again?
It's even possible the airlines made money. This time of year many flights might well go half full, but you can bet everything is stuffed to capacity catching up on the backlog.
Doing a highly simplified analysis, the one data point we have is four engines, all failing, one permanently. So in that case there was a 100% chance of (dangerous) temporary failure, and a 25% chance of permanent failure.
Suppose only the permanent failure of all your engines kills you. That 4 engine 747 then had a 0.25*0.25*0.25*0.25 = 0.4% chance of all it's engines failing completely. A two engine aircraft would have had a 6% chance. The two engine aircraft would have been 16 times more at risk.
Perhaps you could link to some evidence for this? Richard Branson is a businessman and adventurer who has partnered with some excellent engineers including Burt Rutan, who does design and build spaceships. I didn't find any evidence that he's an engineer himself, nor a pilot, although it seems fairly likely he has a private pilot's license. That's a far cry from being an experienced airline pilot with the lives of a few hundred passengers directly in your hands. Wikipedia says Branson was a fairly poor student who suffered from dyslexia and got ahead by connecting with people, not by being technically educated.
Branson is a businessman. He seems to be one of the better ones, but he is probably not qualified to assess the danger to an aircraft himself, and he certainly has a strong conflict of interest in this case. Not to mention he's piping up after the fact - it's easy to criticize someone else's decisions after the danger is past.
It's highly unlikely Agnilux is building processors from scratch. It's much more probable that whatever they are doing starts with... an ARM core.
So no, it doesn't make sense. Google may very well want to start building low power ARM-based servers but it's highly unlikely they bought Agnilux so they can start putting non-ARM processors in their phone.
Slashdot Mirror
Yeah, that's one way of going about it. The other way to look at it is that if all it takes is one employee with an infected device to fry your network, your network is in a pretty sorry state.
I work in medical research. My previous lab was on a hospital network. One day someone, somewhere in the hospital brought in a notebook with a virus. Most of the machines in the hospital went down, including one of the MR scanner consoles. It was a huge crisis. Our lab barely noticed -- we were running Macs. Our Windows terminal server was properly patched and firewalled.
Hospital IT responded by cracking down on outside devices but NOT really tightening up security on individual machines. Of course, if someone, either with malicious intent or by mistake, plugged an infected laptop into the network, they would be right back at square one.
I bet I can make an iPhone app with embedded web browser in objective c faster than you can make one in Flash.
The issue is that the ISP is redirecting your malformed URL before Firefox can, right? And yet the ISP is sleazy and Firefox is a victim?
How about everybody not fiddling with DNS responses, at least not without asking permission first?
Even a call going through a geostationary satellite is more than bearable. After a minute or two you don't even really notice.
The original iPhone jailbreak was via visiting a particular webpage in mobile safari.
Perhaps I shouldn't have said "universal" because you might be tempted to believe that means every person, everywhere, for all time, agrees. There are lots of morals that are universal in terms of societies. For example, all human societies have injunctions against killing. They do differ in the details, but the basic idea seems to be always present. Most individuals also carry this concept around in their heads, and those who don't are usually judged to be mentally ill.
There is no such thing as absolute morals in the way you seem to be implying. Even if a god did dictate them, someone, somewhere would disagree.
Yes, obviously some of our second line morals are cultural. I didn't say all moral beliefs everywhere can be determined solely by evolutionary arguments. Even so, many of those cultural beliefs are rooted in more fundamental morals that can be scientifically justified. Let's use your example of stealing. A culture that has no concept of possession obviously can't have a concept of stealing, but they may well have a moral injunction against grabbing something from your neighbour while he's using it. Most of us probably learned that in the communal environment of kindergarten. Thou shalt not steal and thou shalt not be grabby are just reflections of the same underlying moral, modified slightly by particular environmental factors. What about the explanation for the underlying moral? Well, anyone who's ever grabbed/stole food from a hungry animal while it is eating is likely to be familiar with the real and immediate physical consequences.
There is no reason what we call morals shouldn't have a rational, scientific explanation. Just in case you're only being pedantic, note that I didn't say evolutionary arguments are morals, but that they explain our morals. And the study of ethics can very much be (and should be) based on science. That's actually one of the basic principles of secular humanism.
Take your example. The moral: you ought not kill (your own kind, except in certain circumstances). Why not? Because a book says so? Why does that book say so? More importantly, why do ALL our ethical systems (written in books or otherwise) tend to generally agree on this point? Even many species of animal generally adhere to this principle, although the details differ. Why?
The answer is fairly obvious from a careful consideration of environmental pressures. Killing, except to eat, protect yourself, your kin, or certain other exceptions, is usually a genetic disadvantage. You take a large risk for very little gain.
Secular humanism promotes scientific testing of ethical or moral assertions and should presumably welcome scientific explanations of why particular moral beliefs are advantageous. However, many religious people, and many so-called secular humanists seem to detest the idea that our morals have such inglorious origins as instinct and natural selection.
While that may be so, the letter certainly appears to me to be a general statement of policy, perhaps in response to a particular case, but not limited to it.
Anyway, the thing to do is to put him on trial and figure out exactly who did what. Maybe he's innocent. Let's find out.
When the abuse was happening what do you suppose the pope (this one) was doing at the time? There are documents that indicate he (not as a local bishop, but as a cardinal giving those local bishops orders) was actively participating in the cover up.
So yeah, let's follow your advice and come down hard on everybody responsible. If you're a Catholic you should be screaming for his head.
Absolutely. But when the organization gets caught, you don't reason it away. You come down hard on them, and clean house. You certainly don't promote one of the worst offenders to chief (pope) and let him stay there.
You don't need a god to dictate fixed morals anyway. There are ample evolutionary arguments for why we universally consider certain behaviour bad. The pope seems to have the worst case of moral relativism anyway - stealing crackers is worse than genocide and child abuse isn't all THAT bad... at least if a priest is doing it.
You're right, I was careless when I said "aperture." In photography that word usually means "relative aperture," or f-stop, which is defined as N=f/D (N is the f-stop, f is the focal length and D is the diameter of the light permitting pupil). I really meant D, the diameter of the light permitting pupil.
Take an example. Suppose I have a 100 mm lens that is perfectly paired with a 24x36 mm (standard 35 mm full frame sensor). If this lens is set at f8 the diameter of the light permitting pupil will be:
N = f/D
D = f/N
D = 100 / 8
= 12.5 mm
This combination will have a field of view (I'm just going to calculate the small dimension, but you can feel free to do the other if you wish) of:
alpha = 2*atan(d/(2f))
= 2 * atan(24 / (2*100))
= 0.24
where d is the size of the sensor in the chosen direction.
Now, take the case of a Canon 1.6x crop factor sensor. We're going to manufacture a lens that is just like our other lens in terms of light permitting pupil size and field of view (and thus magnification) but forms an image that is perfectly paired to the smaller sensor. Of course, this means the focal length will have to change:
f = d / (2tan(alpha/2))
= 62.5 mm
Note that 62.5 mm * 1.6 is 100 mm, as we would expect. So what f-stop are we using on the new lens?
N = 62.5 / 12.5
= 5
As you point out, this means that the light intensity on our smaller sensor is greater than on our bigger sensor. If you care to work it through you'll find that, if you divide the smaller sensor into the same number of pixels as the larger, the total light incident on any given pixel-sized area is the same.
This can be shown more easily by considering total incident light. The total number of photons allowed into the camera will be proportional to the area of the light permitting pupil. In our two examples this is:
Full frame case:
A = pi*(D/2)**2
= pi * (12.5 / 2)**2
= 122.7 mm^2
Note that the small sensor case is the same because the pupil is the same diameter. Call the amount of light entering the camera through this pupil 1 unit.
Now, in the full frame case this 1 unit of light is spread over an area of 24x36 mm = 864 mm^2. The incident light on this sensor has an intensity of 1/864 units/mm^2. If we divide the sensor into 10 megapixels (it's a perfect sensor with no wastage), each will have an area of 864 / 10^6 mm and every pixel will get 1 / 10^6 units of light.
Okay, what about the small frame sensor? We have the same amount of light coming into the box, but this time it's spread over a 22.5 mm × 15 mm sensor (area = 337.5 mm^2). The incident light on the sensor has an intensity of 1/337.5 units/mm^2. Note that it is more intense, as we expect from our f-stops. Again, divide the smaller sensor into 10 megapixels, each with an area of 337.5 / 10^6. Again, each pixel will get 1 / 10^6 units of light.
That is, the larger and smaller sensors, with properly designed lenses and neglecting wasted space on the sensor, get equal amounts of light per pixel. So what changes? The focal length and the f-stop do. But the physical dimensions of the lens (except for the thickness) do not change. So for a given light gathering area and a properly designed lens, neglecting wasted space on the sensor and non-dominant sources of noise that do not scale with sensor size, noise performance isn't affected by the size of the sensor.
But wait! The apertures aren't the same! Well, aperture in digital photography doesn't mean what it did in film photography. Given a piece of film, whether 35 mm size or large format, for the same film design and chemistry a 1/100 s @ f8 shot will give the same exposure no matter the format because the film grains (the "pixels") are all the same size. The situation is NOT the same in digital photography. Since the pixels change size in digital sensors depending on the size of the sensor, f8 for one sensor may not be the same as f8 for another,
I'm very familiar with processing RAW.
It does take a certain amount of time. However, you can e-mail your photos to the office and someone can do it on a computer there, or you can do it on your own laptop on your way back from the shoot, or in the hotel room that night. Anyone reasonably familiar with the process can take a RAW and spit out something at least as good as a first-pass developing/printing operation in ten or fifteen seconds.
Compare that to film processing. It requires chemicals (that have a limited shelf life, particularly for colour), a special room, other special equipment, parts of it have to be done in complete darkness by feel (or automatically by a very expensive machine), the reactions take several minutes and once you're done you have to wait for everything to dry. Then you have to scan the negatives and you're right back with your RAW image file.
All that is if you even have a darkroom (never mind a medium format capable darkroom). If not you have to send everything away and wait for the mail.
Note in my post I said "in terms of noise."
The smaller the sensor, given a particular light gathering capability, the greater the signal. You measure signal to noise ratio, not noise.
In practice some space is wasted so bigger sensors really do perform better, but the relationship is not as you imagine.
I very, very rarely enter the menu system on my 30D. Certainly not to adjust shutter speed or aperture. There's one dial for shutter speed, another for aperture. Pretty simple. Are you sure you've used a digital SLR?
If you really want to use your old lenses, go ahead. Depending on the age and make of your lens you can either a) just stick it right on a modern SLR from the same manufacturer b) buy a simple adapter ring and go to (a) or c) file off a particular bit and go to (a).
Not to mention you don't have to wait for processing with digital. For a professional the time saved might add up to $14K on the first job.
"Secondly as I understand it (i've done a little bit on optics but i'm not an expert) depth of field is related to the ratio between aperture and sensor size."
More or less, but the relationship is the opposite of what you've said. A bigger aperture and/or a bigger sensor means less depth of field. Small sensor cameras can be desirable for things like macro photography where you're normally starved for depth of field, or landscape where you want everything in focus. They're undesirable for things like portraits, where you want to make sure the background is well blurred by achieving a narrow depth of field.
One of the reasons compact and cell phone camera shots look amateurish is because the depth of field is very wide, so everything is in focus. More professional looking photos generally use narrower depths of field.
A bigger sensor doesn't do you the least bit of good without a bigger lens, generating a bigger image, to go with it.
Neglecting certain inefficiencies, a small sensor with a big lens will give you just as good an image (in terms of noise) as a big sensor with a big lens. The aperture determines how much light is gathered and focused on the sensor.
Who lost billions? Do you suppose the airlines gave back all the money they were paid for tickets? Even if they did, do you suppose all the people who were stranded decided to stay where they were and not fly home when flights started up again?
It's even possible the airlines made money. This time of year many flights might well go half full, but you can bet everything is stuffed to capacity catching up on the backlog.
Doing a highly simplified analysis, the one data point we have is four engines, all failing, one permanently. So in that case there was a 100% chance of (dangerous) temporary failure, and a 25% chance of permanent failure.
Suppose only the permanent failure of all your engines kills you. That 4 engine 747 then had a 0.25*0.25*0.25*0.25 = 0.4% chance of all it's engines failing completely. A two engine aircraft would have had a 6% chance. The two engine aircraft would have been 16 times more at risk.
Perhaps you could link to some evidence for this? Richard Branson is a businessman and adventurer who has partnered with some excellent engineers including Burt Rutan, who does design and build spaceships. I didn't find any evidence that he's an engineer himself, nor a pilot, although it seems fairly likely he has a private pilot's license. That's a far cry from being an experienced airline pilot with the lives of a few hundred passengers directly in your hands. Wikipedia says Branson was a fairly poor student who suffered from dyslexia and got ahead by connecting with people, not by being technically educated.
Branson is a businessman. He seems to be one of the better ones, but he is probably not qualified to assess the danger to an aircraft himself, and he certainly has a strong conflict of interest in this case. Not to mention he's piping up after the fact - it's easy to criticize someone else's decisions after the danger is past.
It's highly unlikely Agnilux is building processors from scratch. It's much more probable that whatever they are doing starts with... an ARM core.
So no, it doesn't make sense. Google may very well want to start building low power ARM-based servers but it's highly unlikely they bought Agnilux so they can start putting non-ARM processors in their phone.
Also, North America is pretty big. It's unlikely the ash will have much direct effect on Mexico or Florida, for example.
The ash could well be limited to Newfoundland.
"which anyway would take about 3 months IIRC to get from Europe to USA hoho"
If you're sailing in a square rigger, maybe. Cruise ships today generally make the trip in about a week.