Not really. Most atmospheric oxygen has marine origins.
What's more you can replant forests, although you lose the ecological benefits if the forest you just cut down was virgin as opposed to managed timberland.
One of the remarkable facts of Egyptology is how nearly impossible it was to prevent tombs from being robbed over the course of thousands of years. There's never been a tomb found that hasn't been robbed at some point, even Tutankhamen's tomb. Most are picked clean of anything that might be of interest to anyone other than an archaeologist.
But it always seemed to me, given the scale of the pyramids, that there was an obvious option for deterring robbers: make the scale of the engineering project necessary to find and reach the burial chamber more costly than the value of the goods in the chamber. It's not unreasonable; the cost of even a small pyramid must have outweighed the cost of the funerary goods in it by thousands of times. I'm not talking about sealing the burial chamber with a ten ton slab of rock; I'm thinking in terms of hundreds of thousands of tons.
It has to have occurred to anyone who's pondered the pyramids that there might be things still left hidden inside all that volume. The thing is there is no way to investigate such speculation without some means of being able to see through solid stone. For that reason the matter of undiscovered chambers in the pyramids has become to Egyptology a bit like questions about perpetual motions machines are to physicists. I even saw one Egyptologist say in response to this news that there was "zero chance" of anything remaining undiscovered in the Great Pyramid.
But maybe speculation isn't so pointless, now that we in the 21st century actually *can* in a fashion see through solid stone.
If they'd said "fewer than 1 in 10 experience problems" I'd be somewhat skeptical. If they'd claimed "fewer than 1 in 100" I'd be *very* skeptical.
But I have no doubts whatsoever about a claim of fewer than 1 in 100,0000 devices failing. It's bogus.
The only question about a figure like that is whether it was produced by outright fabrication or by some process deliberately designed to discard unflattering data. Nothing that complex achieves anything like that level of perfection.
I remember when PERL was a new thing, being distributed on USENET (a pre-Internet modem-to-modem connection) on comp.sources.misc. It filled a very useful niche for writing filter-type programs that heretofore had been written in a mix of shell script, awk, and sed. What it was good for was written right there on the tin: Practical Extraction and Report Language.
I also remember when the cgi-bin interface was proposed; before that there was no way to handle user input and programmatically generate dynamic output with HTML and the web. PERL fit in very nicely then as a substitute for writing standalone C programs to handle cgi-bin requests, because that was exactly the kind of stuff people had already been using it for.
But very soon thereafter we began to think of web programming less in terms of discrete input-handling programs, and more in terms of systems that had complex, persistent state and complex multi-step behaviors. And people continued to write in PERL into this era because they knew it, and it was still very handy for many of the kinds of "generate HTML from a stream of data" tasks you still had to do in a "web application". Yet it seems to me that it was around this time the wheels began to come off the bus.
Armchair language designers probably have their theories , and I suppose many of those theories have some merit, but I've begun to suspect the problem with languages is often rooted in a kind of culture that grows up around each language. Many of the problems with Java, for example, I believe to come not from the features of the language, although there is certainly grounds for criticizing it, particularly earlier versions. But really Java's problem is that the people who used it developed a culture of over-engineering things; a preference for the verbose and clunky.
PERL, I suspect, has the opposite kind of cultural problem. Of course it's been over ten years since I've had to look at any PERL code so I can't speak to what people have been up to recently; but back in the day when PERL was hot for web applications I had to dig into the code for a number of PERL applications. While there might have been "more than one way to do it", for a lot of working PERL programmers that was an excuse for declaring the job done when a correct output had been generated. And that may be a legacy of the early days of PERL, when most of our jobs were generating discrete, stand-alone information products and your productivity was measured in how many you churned out. Back then people didn't spend much time thinking about non-functional requirements.
I've never hated PERL, but I sure have hated working on PERL programs.
Some people find they need the prospect of human extinction to motivate them to do something about pollution, but in truth that scenario is extremely unlikely. We are the most adaptable multi-cellular organism that four billion years of evolution has produced.
Climate change isn't going to kill us (at least collectively); it's going to cost us a lot of money. In fact it's going to cost some of us more than others, while causing it will benefit some of us more than others. If the costs and benefits of climate change were fairly distributed, then we'd automatically adopt a reasonable compromise. But people like you and me are going to pay a greater share of the costs than we received of the benefits, because things aren't run with our good in mind.
And it's going to cost us in ways that nobody's bothering to measure, but should be. I've been fishing for forty years now, and when I started people still remembered taking native, wild brook trout from streams in my state where they haven't been seen since the 1960s. Water pollution killed them off, but even though we've cleaned up those streams, the waters are too warm now to reestablish them. Streams where the average summertime temperature was 65 twenty years ago are pegging average temps in the 75 range now, and brook trout die at 68-77F depending on maturity. Even the brain-dead hatchery rainbow trout near-clones that are put out to replace the native brookies aren't surviving past June, and they were chosen because they're more heat-tolerant.
It's not just fishing; there have been declines in game species in the lower 48 due to temperature driven habitat stress and parasites. One study found a 75% mortality rate in moose calves due to parasite infection, and a 45% drop in adult population in the past fifteen years. During that time centuries old eastern hemlock groves where I used to hike went from having no sky visible overhead to being largely denuded because of the spread of parasites formerly kept in check. In a decade or two groves with trees predate the signing of the Declaration of Independence will be gone, replaced by alien Noway maples.
Can you put a price on that stuff? I suppose you can in terms of lost economic activity, but more to the point we're losing something you literally can't put a price on: tradition. Heritage. Our natural legacy. Maybe some people will be able to afford to take a month to go on safari, but for the average person these things are disappearing.
The notion that Lucas turned to a historical uniform expert for his costume and ended up with his bad guys in Stallhelms and Kriegsmarine tropical caps doesn't sound so ridiculous to me.
Of course you have to go back to 1977; Nazis weren't a serious contemporary political issue back then. They'd been irrelevant long enough to be a joke; just the thing if you wanted a campy villain.
Hydrogen sulfide is a reactive gas that normally has a half-life in the atmosphere measured in days -- too short to be a long-term greenhouse gas. The reaction is as follows
H2S + OH HS + H20.
Sulfanyl (HS) is highly reactive and converts mainly to sulfur dioxide (SO2), which carries the sulfur out of the atmosphere as acid rain. So again the total half life of sulfur in the atmosphere is a matter of days or weeks at most. This precludes it H2S being a greenhouse gas concern.
In comparison the half-life of CO2 is centuries; methane has a half-life in decades (and decomposes to CO2). Chlorofluorocarbons have half lives that run from under a year to thousands of years depending on the compound. The most common CFC, freon (dichlorodifluoromethane) has a half life of about 100 years.
Because they are angry and fearful. It's an old, reliable formula: scare people, or take advantage of their existing insecurities, and then put a face on it: the Jew. The Auslander. The immigrant. The Mexican.
The formula works because it feels simple. There's no complicated policy or economics involved, you know its right because it feels right. But feelings *always* feel right. There's no such thing as critical feeling, only critical thinking.
Basically if you need a part of a plane that plane is grounded for half a year. If that's only *twice* the "program's objective" I have to believe they mean "the program's objective for this point in time."
Well, you're supposing people are always rational in their hiring decisions. There's a lot of emotion and "gut" thinking involved in most hires, so much so that I think of rationality as a kind of competitive advantage.
And by the way irrationality is a two way street when it comes to employment. Take ageism. If you believed that employees and employers were entirely rational it wouldn't be a problem. An older employee would take a lower paid engineering job rather than an even lower paying non-engineering job. And faced with the reality of ageism a smart employer would go bargain hunting for experienced older employees.
So you do have engineers leaving the field while there is a shortage of people with their skills, usually ones over 40. If hiring decisions (and job acceptance decisions) were made by expert systems which found the cheapest (best paying available) fit of skills to requirements, ageism wouldn't exist.
Token ring is time division multiplexed and Ethernet is stastically multiplexed. Neither is fundamentally better than the other, it depends on your requirements.
It's not just marketing in that case; Ethernet really does perform better when overall network segment use is less than 30%. But it's moot now because everyone uses hubs.
How a number is constructed is more important than how good they sound.
So 78% of F35s were able to *fly*. Doesn't mean they're ready to actually *do* anything other than fly, like use their weapons systems, or appear as small on enemy radars as they're supposed to. So that 78% may include aircraft that *could* fly, but which would be pointless to fly.
We should take any statitics on a too-big-to-fail project with a large grain of salt, because they can easily be affected by tweaking your success criteria.
Consider: for FY16, the F35 had a 56% availability rate, which for a plane so early in its deployment is quite impressively good. But because of the program's unique concurrency strategy, in which deliveries of aircraft started years before the design was finalized, 187 of the aircraft will never be capable of combat operations -- not unless they're sent back to the factory and re-manufactured.
As of March of this year, 231 F35s have been delivered, so if "available" meant "capable of flying a combat mission", the highest the FY '16 availability figure could possibly have been is 19% 81% of the fleet were semi-functioning prototypes.
So clearly "available" means "capable of flying the mission you planned for them"; and you can adjust the rate of availability by planning your missions accordingly. Gun not working? Plan a mission with no shooting and the plane is still available.
Without repair capabilities, availability for actual combat is probably zero; but it's probably zero anyway until the software improves.
It depends on whether you goal is to have a certain number plains, or whether you want to be able to use them.
If having the plane in inventory is the only thing you care about, then you're right: the quickest and cheapest route is to concentrate on production rather than maintenance. If your goal is to have planes ready to use, your production rate cannot outstrip your repair capabilities, because something as complex as a modern fighter aircraft is constantly breaking down.
The requirements for something like this is different from connecting your laptop to peripherals.
Firewire is a peer to peer bus that supports data transfer via DMA, minimizing interrupts. Since response time consistency isn't a big deal for your laptop, and it has CPU bandwidth sitting unused most of the time, something like USB 3 is just as good if not better.
The US Marines are screwed without the F35 too. They have 9 amphibious assault ships, each larger than a WW2 fleet carrier. Each of these ships are supposed to be able to debark highly mobile , self-contained "expeditionary units" of 2200 troops, each of which has a squadron of ground attack aircraft which have to operate from improvised air strips.
The thing is, the air component of that doesn't work against modern, mobile air defenses, like those possessed by Iraq unless you have a stealth aircraft that can take off and land vertically or nearly so. This will leave the Marine units tied to air support from carriers.
Well, estimating the wealth of someone who deals in contraband is by its very nature impossible. It's not like the file 10-Qs for their business. Their money is laundered and stashed in assets that can't be traced to them.
Still, buying out Jeff Bezos at 90+ billion net worth isn't chump change for anyone, even Vladmir Putin is who by some estimated to have a net worth of over 200 billion.
Also, it's probably worth questioning whether people who have to manage their assets this way are richer in any meaningful way, because in a sense those assets are not really available to them to use. Jeff Bezos could sell of most of his stock and retire to an orbiting space station if he wanted to. If Putin wanted to do that it would reveal where he's got his money stashed.
For criminal billionaires a lot of their assets are in effect a kind of insurance. If the cops discover their Swiss bank accounts there's still that London real estate they own through various proxies. If they get their London assets there's still that box of gold ingots you stashed in your grandma's grave.
This is similar in a way to the bug behind the famous Therac-25 incident. The Therac-25 as a medical radiation machine which had software which was supposed to prevent patients receiving dangerous doses of radiation. However it turned out the operators entered configuration command far faster than testers did, creating a race condition that could result in the machine delivering over 100x the safe dosage.
The bug never showed up in testing because the testers never got as fast at input as the operators, and in any case the specific keystroke combination that caused it was rare.
This isn't a math problem. It's basic programming; you don't rely on tests of equality between the results of floating point calculations and integer quantities.
What's going on here is that the sqrt function yields a floating point number that's close to the integer 2, and that's rounded for display down to 2. The round for display routine clearly chooses not to round if the difference is large relative to the amount. For example take the number 1000, and add.00000000001. It will display 1000.00000000001. Now do the same with 1,000,000; it will display 1,000,000 even though internally the number is 1,000,000.00000000001.
It's not an entirely unreasonably way of doing it, although most people would tend to round to a fixed number of digits.
Manipulating voter registration records can also affect the outcome of an election, particularly in a situation where you are performing mass voter roll purges -- which I believe Georgia did. Tweaking the purge in the partisan way is one of the things that could be hidden by this.
Many voters who are denied a ballot may not choose to cast a provisional ballot, or come out to vote next time. Although this may seem like they get what they deserve if they're not sufficiently dedicated, voting can be more of a sacrifice for some people, particularly those with limited time to vote and where the polling stations are located inconveniently. If you're risking late for work you might choose just to leave. There have been documented attempts to manipulate turnout by inconveniently locating polling places.
As far as paper ballots, the obvious choice is optically scanned paper ballots. That said, I have never seen any evidence that human recounters don't perform acceptably. They are not perfect, but neither are machines -- for that matter how could you possibly know if a machine is perfect? In any case human limitations can be dealt with using statistics, to any desirable level of confidence.
However the editors should be informed that due to the phenomenon of metaphor inflation the phrase "This machine kills captchas" no longer carries any discernible meaning to the average reader. The closest you can get would be "This machine eviscierates captchas," which would be taken to mean "says uncomplimentary things about".
Strangely, people who push this kind of conspiracy theory don't seem to think there would be funding for scientists who could show that humans weren't contributing to climate change.
They also don't know any working scientists. It's not about money, it's about reputation, and the way you make your reputation is you prove other scientists wrong.
Well, that's kind of the poster's point. People are afraid of investing in the common good because they can't make ends meet even on practically unlimited money.
Slashdot Mirror
Not really. Most atmospheric oxygen has marine origins.
What's more you can replant forests, although you lose the ecological benefits if the forest you just cut down was virgin as opposed to managed timberland.
One of the remarkable facts of Egyptology is how nearly impossible it was to prevent tombs from being robbed over the course of thousands of years. There's never been a tomb found that hasn't been robbed at some point, even Tutankhamen's tomb. Most are picked clean of anything that might be of interest to anyone other than an archaeologist.
But it always seemed to me, given the scale of the pyramids, that there was an obvious option for deterring robbers: make the scale of the engineering project necessary to find and reach the burial chamber more costly than the value of the goods in the chamber. It's not unreasonable; the cost of even a small pyramid must have outweighed the cost of the funerary goods in it by thousands of times. I'm not talking about sealing the burial chamber with a ten ton slab of rock; I'm thinking in terms of hundreds of thousands of tons.
It has to have occurred to anyone who's pondered the pyramids that there might be things still left hidden inside all that volume. The thing is there is no way to investigate such speculation without some means of being able to see through solid stone. For that reason the matter of undiscovered chambers in the pyramids has become to Egyptology a bit like questions about perpetual motions machines are to physicists. I even saw one Egyptologist say in response to this news that there was "zero chance" of anything remaining undiscovered in the Great Pyramid.
But maybe speculation isn't so pointless, now that we in the 21st century actually *can* in a fashion see through solid stone.
If they'd said "fewer than 1 in 10 experience problems" I'd be somewhat skeptical. If they'd claimed "fewer than 1 in 100" I'd be *very* skeptical.
But I have no doubts whatsoever about a claim of fewer than 1 in 100,0000 devices failing. It's bogus.
The only question about a figure like that is whether it was produced by outright fabrication or by some process deliberately designed to discard unflattering data. Nothing that complex achieves anything like that level of perfection.
I remember when PERL was a new thing, being distributed on USENET (a pre-Internet modem-to-modem connection) on comp.sources.misc. It filled a very useful niche for writing filter-type programs that heretofore had been written in a mix of shell script, awk, and sed. What it was good for was written right there on the tin: Practical Extraction and Report Language.
I also remember when the cgi-bin interface was proposed; before that there was no way to handle user input and programmatically generate dynamic output with HTML and the web. PERL fit in very nicely then as a substitute for writing standalone C programs to handle cgi-bin requests, because that was exactly the kind of stuff people had already been using it for.
But very soon thereafter we began to think of web programming less in terms of discrete input-handling programs, and more in terms of systems that had complex, persistent state and complex multi-step behaviors. And people continued to write in PERL into this era because they knew it, and it was still very handy for many of the kinds of "generate HTML from a stream of data" tasks you still had to do in a "web application". Yet it seems to me that it was around this time the wheels began to come off the bus.
Armchair language designers probably have their theories , and I suppose many of those theories have some merit, but I've begun to suspect the problem with languages is often rooted in a kind of culture that grows up around each language. Many of the problems with Java, for example, I believe to come not from the features of the language, although there is certainly grounds for criticizing it, particularly earlier versions. But really Java's problem is that the people who used it developed a culture of over-engineering things; a preference for the verbose and clunky.
PERL, I suspect, has the opposite kind of cultural problem. Of course it's been over ten years since I've had to look at any PERL code so I can't speak to what people have been up to recently; but back in the day when PERL was hot for web applications I had to dig into the code for a number of PERL applications. While there might have been "more than one way to do it", for a lot of working PERL programmers that was an excuse for declaring the job done when a correct output had been generated. And that may be a legacy of the early days of PERL, when most of our jobs were generating discrete, stand-alone information products and your productivity was measured in how many you churned out. Back then people didn't spend much time thinking about non-functional requirements.
I've never hated PERL, but I sure have hated working on PERL programs.
Economic burden is actually an important point.
Some people find they need the prospect of human extinction to motivate them to do something about pollution, but in truth that scenario is extremely unlikely. We are the most adaptable multi-cellular organism that four billion years of evolution has produced.
Climate change isn't going to kill us (at least collectively); it's going to cost us a lot of money. In fact it's going to cost some of us more than others, while causing it will benefit some of us more than others. If the costs and benefits of climate change were fairly distributed, then we'd automatically adopt a reasonable compromise. But people like you and me are going to pay a greater share of the costs than we received of the benefits, because things aren't run with our good in mind.
And it's going to cost us in ways that nobody's bothering to measure, but should be. I've been fishing for forty years now, and when I started people still remembered taking native, wild brook trout from streams in my state where they haven't been seen since the 1960s. Water pollution killed them off, but even though we've cleaned up those streams, the waters are too warm now to reestablish them. Streams where the average summertime temperature was 65 twenty years ago are pegging average temps in the 75 range now, and brook trout die at 68-77F depending on maturity. Even the brain-dead hatchery rainbow trout near-clones that are put out to replace the native brookies aren't surviving past June, and they were chosen because they're more heat-tolerant.
It's not just fishing; there have been declines in game species in the lower 48 due to temperature driven habitat stress and parasites. One study found a 75% mortality rate in moose calves due to parasite infection, and a 45% drop in adult population in the past fifteen years. During that time centuries old eastern hemlock groves where I used to hike went from having no sky visible overhead to being largely denuded because of the spread of parasites formerly kept in check. In a decade or two groves with trees predate the signing of the Declaration of Independence will be gone, replaced by alien Noway maples.
Can you put a price on that stuff? I suppose you can in terms of lost economic activity, but more to the point we're losing something you literally can't put a price on: tradition. Heritage. Our natural legacy. Maybe some people will be able to afford to take a month to go on safari, but for the average person these things are disappearing.
The notion that Lucas turned to a historical uniform expert for his costume and ended up with his bad guys in Stallhelms and Kriegsmarine tropical caps doesn't sound so ridiculous to me.
Of course you have to go back to 1977; Nazis weren't a serious contemporary political issue back then. They'd been irrelevant long enough to be a joke; just the thing if you wanted a campy villain.
Hydrogen sulfide is a reactive gas that normally has a half-life in the atmosphere measured in days -- too short to be a long-term greenhouse gas. The reaction is as follows
H2S + OH HS + H20.
Sulfanyl (HS) is highly reactive and converts mainly to sulfur dioxide (SO2), which carries the sulfur out of the atmosphere as acid rain. So again the total half life of sulfur in the atmosphere is a matter of days or weeks at most. This precludes it H2S being a greenhouse gas concern.
In comparison the half-life of CO2 is centuries; methane has a half-life in decades (and decomposes to CO2). Chlorofluorocarbons have half lives that run from under a year to thousands of years depending on the compound. The most common CFC, freon (dichlorodifluoromethane) has a half life of about 100 years.
That's a thing too, but not for getting people to vote against their own interests.
Because they are angry and fearful. It's an old, reliable formula: scare people, or take advantage of their existing insecurities, and then put a face on it: the Jew. The Auslander. The immigrant. The Mexican.
The formula works because it feels simple. There's no complicated policy or economics involved, you know its right because it feels right. But feelings *always* feel right. There's no such thing as critical feeling, only critical thinking.
Basically if you need a part of a plane that plane is grounded for half a year. If that's only *twice* the "program's objective" I have to believe they mean "the program's objective for this point in time."
Well, you're supposing people are always rational in their hiring decisions. There's a lot of emotion and "gut" thinking involved in most hires, so much so that I think of rationality as a kind of competitive advantage.
And by the way irrationality is a two way street when it comes to employment. Take ageism. If you believed that employees and employers were entirely rational it wouldn't be a problem. An older employee would take a lower paid engineering job rather than an even lower paying non-engineering job. And faced with the reality of ageism a smart employer would go bargain hunting for experienced older employees.
So you do have engineers leaving the field while there is a shortage of people with their skills, usually ones over 40. If hiring decisions (and job acceptance decisions) were made by expert systems which found the cheapest (best paying available) fit of skills to requirements, ageism wouldn't exist.
Token ring is time division multiplexed and Ethernet is stastically multiplexed. Neither is fundamentally better than the other, it depends on your requirements.
It's not just marketing in that case; Ethernet really does perform better when overall network segment use is less than 30%. But it's moot now because everyone uses hubs.
How a number is constructed is more important than how good they sound.
So 78% of F35s were able to *fly*. Doesn't mean they're ready to actually *do* anything other than fly, like use their weapons systems, or appear as small on enemy radars as they're supposed to. So that 78% may include aircraft that *could* fly, but which would be pointless to fly.
We should take any statitics on a too-big-to-fail project with a large grain of salt, because they can easily be affected by tweaking your success criteria.
Consider: for FY16, the F35 had a 56% availability rate, which for a plane so early in its deployment is quite impressively good. But because of the program's unique concurrency strategy, in which deliveries of aircraft started years before the design was finalized, 187 of the aircraft will never be capable of combat operations -- not unless they're sent back to the factory and re-manufactured.
As of March of this year, 231 F35s have been delivered, so if "available" meant "capable of flying a combat mission", the highest the FY '16 availability figure could possibly have been is 19% 81% of the fleet were semi-functioning prototypes.
So clearly "available" means "capable of flying the mission you planned for them"; and you can adjust the rate of availability by planning your missions accordingly. Gun not working? Plan a mission with no shooting and the plane is still available.
Without repair capabilities, availability for actual combat is probably zero; but it's probably zero anyway until the software improves.
It depends on whether you goal is to have a certain number plains, or whether you want to be able to use them.
If having the plane in inventory is the only thing you care about, then you're right: the quickest and cheapest route is to concentrate on production rather than maintenance. If your goal is to have planes ready to use, your production rate cannot outstrip your repair capabilities, because something as complex as a modern fighter aircraft is constantly breaking down.
The requirements for something like this is different from connecting your laptop to peripherals.
Firewire is a peer to peer bus that supports data transfer via DMA, minimizing interrupts. Since response time consistency isn't a big deal for your laptop, and it has CPU bandwidth sitting unused most of the time, something like USB 3 is just as good if not better.
The US Marines are screwed without the F35 too. They have 9 amphibious assault ships, each larger than a WW2 fleet carrier. Each of these ships are supposed to be able to debark highly mobile , self-contained "expeditionary units" of 2200 troops, each of which has a squadron of ground attack aircraft which have to operate from improvised air strips.
The thing is, the air component of that doesn't work against modern, mobile air defenses, like those possessed by Iraq unless you have a stealth aircraft that can take off and land vertically or nearly so. This will leave the Marine units tied to air support from carriers.
Well, estimating the wealth of someone who deals in contraband is by its very nature impossible. It's not like the file 10-Qs for their business. Their money is laundered and stashed in assets that can't be traced to them.
Still, buying out Jeff Bezos at 90+ billion net worth isn't chump change for anyone, even Vladmir Putin is who by some estimated to have a net worth of over 200 billion.
Also, it's probably worth questioning whether people who have to manage their assets this way are richer in any meaningful way, because in a sense those assets are not really available to them to use. Jeff Bezos could sell of most of his stock and retire to an orbiting space station if he wanted to. If Putin wanted to do that it would reveal where he's got his money stashed.
For criminal billionaires a lot of their assets are in effect a kind of insurance. If the cops discover their Swiss bank accounts there's still that London real estate they own through various proxies. If they get their London assets there's still that box of gold ingots you stashed in your grandma's grave.
This is similar in a way to the bug behind the famous Therac-25 incident. The Therac-25 as a medical radiation machine which had software which was supposed to prevent patients receiving dangerous doses of radiation. However it turned out the operators entered configuration command far faster than testers did, creating a race condition that could result in the machine delivering over 100x the safe dosage.
The bug never showed up in testing because the testers never got as fast at input as the operators, and in any case the specific keystroke combination that caused it was rare.
This isn't a math problem. It's basic programming; you don't rely on tests of equality between the results of floating point calculations and integer quantities.
What's going on here is that the sqrt function yields a floating point number that's close to the integer 2, and that's rounded for display down to 2. The round for display routine clearly chooses not to round if the difference is large relative to the amount. For example take the number 1000, and add .00000000001. It will display 1000.00000000001. Now do the same with 1,000,000; it will display 1,000,000 even though internally the number is 1,000,000.00000000001.
It's not an entirely unreasonably way of doing it, although most people would tend to round to a fixed number of digits.
The Invisible Hand provides, and corporations are the minister of Its grace.
Manipulating voter registration records can also affect the outcome of an election, particularly in a situation where you are performing mass voter roll purges -- which I believe Georgia did. Tweaking the purge in the partisan way is one of the things that could be hidden by this.
Many voters who are denied a ballot may not choose to cast a provisional ballot, or come out to vote next time. Although this may seem like they get what they deserve if they're not sufficiently dedicated, voting can be more of a sacrifice for some people, particularly those with limited time to vote and where the polling stations are located inconveniently. If you're risking late for work you might choose just to leave. There have been documented attempts to manipulate turnout by inconveniently locating polling places.
As far as paper ballots, the obvious choice is optically scanned paper ballots. That said, I have never seen any evidence that human recounters don't perform acceptably. They are not perfect, but neither are machines -- for that matter how could you possibly know if a machine is perfect? In any case human limitations can be dealt with using statistics, to any desirable level of confidence.
Scientists are only in it for the money.
However the editors should be informed that due to the phenomenon of metaphor inflation the phrase "This machine kills captchas" no longer carries any discernible meaning to the average reader. The closest you can get would be "This machine eviscierates captchas," which would be taken to mean "says uncomplimentary things about".
Strangely, people who push this kind of conspiracy theory don't seem to think there would be funding for scientists who could show that humans weren't contributing to climate change.
They also don't know any working scientists. It's not about money, it's about reputation, and the way you make your reputation is you prove other scientists wrong.
Well, that's kind of the poster's point. People are afraid of investing in the common good because they can't make ends meet even on practically unlimited money.