Of course Canada is a state; they just think they're a country. It's kinda cute, actually - like when little kids play grown-up. Just don't try to tell them they're not really a country, or they'll get all upset about it.
No one really wants a free market; everyone wants to "level the playing field" - in a direction that completely coincidentally benefits them at the expense of others.
That's not to say that a free market is bad, just that fox is no more and no less hypocritical than most other companies.
Yes, getting 100+ miles high is the easy part - getting that high with 17,000 MPH of sideways velocity is where it gets tricky. Without that much sideways velocity, you just fall back down.
So kids, how does it feel to be your parents talking about how this e-mail thing is a waste of time and when you want to talk to someone you should pick up the phone or write a letter?
Phones? You had phones? Why, back in my day, we had to yell really loud!
Yes, that was my first thought - extending my mouse's lifespan would be great, because I hate trying to find another one when the buttons stop working or the wheel stutters.
But NOOOOO, it's not anything useful like that - just some crap about making people live forever.
Excel has certain settings (e.g. selected page, column widths, etc) that are saved with the document, but which do not cause the dirty bit to be set when they are changed. This way you can open a spreadsheet and look around, but it doesn't prompt you to save unless you change something "important". It seems like "last printed date" should be treated similarly.
True, modern radio buttons don't have the same tactile feel as old-fashioned ones do - but computer "radio buttons" have never had that tactile feel, so in that sense they're actually more like modern radio buttons than they were like old-fashioned ones. The analogy is getting more accurate, not less.
And yes, I remember the partial-press phenomenon - but I always thought of that as a harmless "bug" due to the mechanical nature of the mechanism, not a "feature" to be emulated. Certainly computer "radio buttons" have never exhibited that sort of behavior.
I open one as a template, modify it and then save it with a new name. "Autosave & undo" would immediately overwrite the file I use as a template.
Instinctively saving (which I tend to do about every other sentence) without changing the name first has the same effect - I finally learned (after screwing up WAY too many times) to ALWAYS change the name FIRST ("save as..."), and THEN make the changes and save again. That usage pattern is now burned into my brain, to the point that I get nervous when I see someone else modifying a template without doing a "save as" first!
What I don't understand is why word thinks that printing a document constitutes some sort of change, and it asks me to save (even if the doc is in my "temporary internet files" folder!)
OK, I get that printing might (might!) adjust the pagination or whatever, but surely any moron can see that any such changes are changes it made on my behalf, not changes I actually made! Yet year after year this bug remains...
who actually still has a radio with preset stations, where once one button is operated, all other buttons mechanically jump back into the "off" position?
But it doesn't have to be mechanical buttons - I've got a modern electronic radio, but when I press one of the presets, it switches to that station and simultaneously switches away from whatever station I was previously listening to.
For contrast, imagine if they were "checkboxes" instead of "radio buttons" - pressing one would result in listening to two stations at once, and then you'd have to go back and de-select the previous station. Only a geek could rationalize a UI like that: "But it's better that way: what if you want to listen to two stations at once? Radio buttons inhibit your freedom!")
I think radio buttons are a useful analogy, though perhaps there are better names out there - maybe something like "channels", but that focuses on one particular application of the control, rather than the functionality of the control itself.
The warning doesn't bother me - I rent from netflix, use DvdShrink to rip the disc (movie only) to an ISO file, then send the disk back. I literally never see the screens they've already got telling me not to do what I'm doing; adding more (that I also won't see) doesn't change the situation.
So a natural birth with identifiable mother and father is the same as a child born via a research institute? In that case, you wouldn't mind if the US gov phased out the volunteer army and gets darpa to fill the ranks via clones
Who's being a dick now? Of course you can't own a person - we settled that about 150 years ago (at least in the US; YMMV). If the darpa clones WANT to join the army, they're entitled to do so same as anyone else; if not, they can't be forced into it any more than anyone else. If you start with the common-sense premise that the rights and responsibilities of a person arise because of their personhood, and are independent of how they came to be a person, it all pretty much sorts itself out.
Perhaps there are as many as three separate issues:
- are babies whose genes come from one parent legally different from babies whose genes come from two parents? I can't think of any reason they would be.
- are babies who develop outside of wombs legally different from babies who develop inside wombs? Again, I can't think of any reason they would be.
- are babies born of surrogate mothers who were paid by a corporation legally different from babies born of surrogate mothers paid by an individual? I still can't think of a good reason why they would be different.
Plus of course there are many possible combinations, but even if you mix all 3 (a corporation pays for a single person's DNA to be cloned into a human and somehow has the fetus develop without the aid of a womb) I still can't see how the legal system could or should allow the corporation to "own" the person (or kill them, or whatever). Once the person reaches adulthood they're no different than anyone else; the only thing even slightly tricky is the first 18 years.
Yes, if the company somehow ditches responsibility for the person's upbringing the expense would fall on society at large - but that already happens all the time (via orphanages, adoptions, etc) and given the likely expense of the cloning process I can't see there being a significant additional burden to society.
As for being physically normal, random birth defects do and probably always will occur, but if it can be shown that the company's actions (or inactions) were negligent, I'd expect the system to treat the company much like it does a pregnant woman who smokes, drinks, does other drugs, etc (I'll admit I don't actually know how high the bar is, or what happens to the parents who exceed the threshold). Companies could go out of business, but then again mothers who smoke and drink and whatnot die and somehow we muddle through.
What rights does a clone have? Trying to build sane regulation for cloning would be a nightmare that no politician could navigate successfully.
Yes, we face that problem all the time with identical twins - deciding what rights they have is a political nightmare! The fact that clones are not just identical twins, but time-shifted identical twins makes the problem more complex than the human mind could unravel.
But no mention was made of using the money to pay artists or record labels or whatever - the summary at least says they want to avoid becoming the next greece, so it sounds like just another tax to help fund various govt operations. In that case I can't see that it changes the copyright/liability situation at all.
However, at 103 euro for a 2TB drive, that's pretty close to a 100% tax, and Moore's law says the tax rate will double every 18 months. Yikes! (yes, I know Moore's law applies to transistors, not spinning rust, but the effect is similar)
Instead, you use a data access layer, that always binds parameters.
Kinda like I said above. Only you claim that you will miss sanitizing something. So what if you forget to use bound parameters? Oh that's right, things work perfectly in your view of the world but everyone else is wrong. Use a data access layer, access everything the same way.
I don't so much care how "thick" your data access layer is - a thousand layers of code or just a rule - the important thing is that at the bottom you MUST use bound parameters instead of doubling all your quotes and wrapping it in quotes.
In-band signaling... I'll leave that for others if they want to rip it apart. I assume you mean escape sequences, replacing control characters with escapes specifically. There are common ways of replacing, and common ways to defeat common ways of replacing. It has nothing to do with in or out of band signaling.
Poor choice of words, perhaps - what it really boils down to is, don't let your users write your source code. Seems pretty obvious when you say it that way, but so many things like SQL injection attacks, XSS browser problems, etc, all come down to taking a string of user input and putting it into an environment where it gets evaluated as executable code. People see that it's happening (usually the first time Mr. O'Brien registers), and they try to patch it, but they usually fail one way or another.
For example, go back a few weeks and find the slashdot article about the voting machine being hacked (legally, during a public eval period) by some researchers. It turned out to be the wrong kind of quotes used in a shell script, which meant that a carefully crafted input ended up being executed as code. Watch over the next few weeks/months as various as-yet-unknown exploits are discussed in academic or real-world settings, and 99% of the time it ends up that user input is being executed in some way. And more often than not, there was some sort of attempt at "sanitizing" the input, which failed to account for something.
IF you have a bug in the binding, such as the case here, it doesn't matter if it's in or out of band. There is a bug, and it will likely be discovered sooner or later.
Yes, there could always be a bug in an underlying library. If the bug is in a subroutine that supposedly sanitizes your data, you're screwed (and note that there's a decent chance you won't know about the bug until someone else uses it on you). If the bug is in the SQL binding code, and the 8 bytes that's supposed to represent am IEEE floating point number happens to end up containing 'or1=1-- , then it probably doesn't matter, because no part of an SQL driver is likely to be expecting to execute the binary data of a bound parameter. And if there is someone a problem where the data packets DO try to get evaluated, you're far more likely to find it before the system hits production, because the vast majority of such attempted evaluations will fail miserably due to syntactical errors or whatever.
I only know of ONE environment where you really have no choice but to "escape" a bunch of strings, glue them together, and hope for the best: HTML. There's no equivalent of a bound parameter. And this fundamental flaw is why web pages designed by careless people (realistically, that's most of them) will always be easily exploitable, and web pages designed by careful people will also be exploitable, just not as easily and somewhat less often.
Mark my words: ten years from now, if people are still using HTML, there will still be major new types of attacks being discovered and utilized every other month or so. It's inherent in the architecture, and every new feature (javascript, CSS, etc) just introduces new escaping rules for people to fuck up..
Why oh why do people still make and use systems/apps/tools/interfaces/etc that use in-band signaling and thus require that their inputs be "sanitized"? Can't everyone see that sanitizing inputs is a fool's errand? You'll ALWAYS miss something, or the next version will have a feature you forgot to screen for, or something. In-band signaling is BAD BAD BAD and any system that uses it is doomed to an endless series of X-injection attacks.
For example (and yes, I realize this has nothing to do with SQL, it's just an example) don't even try to sanitize your SQL inputs; use bound parameters instead - not only is it guaranteed 100% safe, it's easier and faster too! As much as I love XKCD, little Bobby Tables really screwed the pooch on that one.
Remember, folks: when it comes to any sort of in-band signaling: JUST SAY NO. If you think you need to sanitize your inputs, you're doing something completely wrong. Stop and figure out what it is, and figure out how to do it right; don't just throw in some half-assed regex or character translation/stripping or whatever and hope that no one is cleverer than you are.
Did you READ TFA? He's not talking about lifting the ban on talking on the phone - he's talking about lifting the ban on having gizmos powered on during the takeoff and landing. If you can tolerate someone next to you reading a kindle or playing angry birds for 10 hours, you can tolerate it for another 30 minutes.
His real point is that he's too weak to turn his gizmos off when he wants some down time, so he wants to make sure no one else can use theirs either.
Actually, I wonder - if that's the only time he can get away from his gizmos, does he book pointless flights back and forth across the country, with as many stops as possible, just to get some quiet time?
This doesn't help much for those of us with crappy internet - I've only got about 300K (bits) upload speed, and at that speed backing up 1TB would take around a year.
FWIW, my strategy is to keep truly important stuff on a raid enclosure (and backup to other disks periodically), and to just live with the fact that there's really nothing irreplaceable about the rest.
electrons are on the same level as protons and neutrons, but quarks of of a lower order yet that phrase seems to lump them all in the same set.
It depends on which direction you're counting from: from the top down, electrons are on the same level as protons and neutrons ("constituents of atoms") - but from the bottom up, electrons and quarks are on the same level (level 0 = fundamental [as far as we know] particles ), but protons and neutrons are a level above electrons (level 1 = stuff built directly from fundamental particles)
Ah, so then the argument is that these nefarious henchmen are going to station themselves at the home of every voter that they wish to bribe while they watch them click on the pre-selected candidate? Since one or two wouldn't turn the tide of an election, we're talking 10s of thousands of conspirators who must operate in absolute secrecy and keep quiet for the rest of their lives.
An argument against internet voting due to interference by aliens from another planet would make more sense.
Yes, thousands of nefarious henchmen are implausible... but more mundane threats exist, too. Imagine an abusive husband telling his wife (and/or kids) who to vote for... Right now they can agree with him (to save their skins) and then go into the voting booth and do whatever the hell they want to do. But with him looking over their shoulders, that option disappears. Multiply this by the number of overbearing spouses out there, and it could become significant.
The problem is that the cost of securing such a system (which has to be accessible to the general populace) is very very high compared to the cost of compromising such a system.
Actually you've got a problem no matter how much you spend or how secure it actually is: even if no shenanigans have taken place - even if none COULD take place, due to some amazing design - the losers will claim that the system was rigged against them. They've always done that, of course, but have generally gotten limited traction because undetectable large-scale fraud with paper ballots is implausible. The difference is that with a fully-computerized system, fraud will seem plausible to the average person. No amount of gibberish spewed by so-called "expert cryptographers" will convince the public that fraud didn't take place. Without any brakes on the conspiracy theories, huge sections of the population will convince themselves that the gov't is a fraud (remember, it doesn't matter whether fraud actually took place or not) and that could be very bad for society.
I'll stick with paper ballots.
P.S. To prove my point about how people want to deny the results of elections, I guarantee I'll get lots of responses detailing how past elections have already been stolen. I'm not saying they have or haven't - and it's completely irrelevant to my point - but I'll get lots of responses anyway. Now imagine if there were no effective way for the average person to estimate the likelihood of these claims being true - just 49% of the country believing it was all a scam, and 51% telling the 49% to shut the hell up and quit being sore losers.
Slashdot Mirror
I'd like to volunteer to spend a month in space - purely for scientific purposes, you understand.
The secret to a long life is actually very simple: Keep Breathing. That's all there is to it!
Of course Canada is a state; they just think they're a country. It's kinda cute, actually - like when little kids play grown-up. Just don't try to tell them they're not really a country, or they'll get all upset about it.
No one really wants a free market; everyone wants to "level the playing field" - in a direction that completely coincidentally benefits them at the expense of others.
That's not to say that a free market is bad, just that fox is no more and no less hypocritical than most other companies.
Yes, getting 100+ miles high is the easy part - getting that high with 17,000 MPH of sideways velocity is where it gets tricky. Without that much sideways velocity, you just fall back down.
Venture capitalists' ability to make billions of dollars for no effort is being threatened!!11
Yes, those greedy bastards - providing money to startups who otherwise might never get off the ground. I hate people like that!
P.S. The hip term is "vulture capitalists" - if you're going to be snarky, at least do it right.
So kids, how does it feel to be your parents talking about how this e-mail thing is a waste of time and when you want to talk to someone you should pick up the phone or write a letter?
Phones? You had phones? Why, back in my day, we had to yell really loud!
(now get off my lawn!)
Yes, that was my first thought - extending my mouse's lifespan would be great, because I hate trying to find another one when the buttons stop working or the wheel stutters.
But NOOOOO, it's not anything useful like that - just some crap about making people live forever.
Wow... yes, that's pointless.
Excel has certain settings (e.g. selected page, column widths, etc) that are saved with the document, but which do not cause the dirty bit to be set when they are changed. This way you can open a spreadsheet and look around, but it doesn't prompt you to save unless you change something "important". It seems like "last printed date" should be treated similarly.
True, modern radio buttons don't have the same tactile feel as old-fashioned ones do - but computer "radio buttons" have never had that tactile feel, so in that sense they're actually more like modern radio buttons than they were like old-fashioned ones. The analogy is getting more accurate, not less.
And yes, I remember the partial-press phenomenon - but I always thought of that as a harmless "bug" due to the mechanical nature of the mechanism, not a "feature" to be emulated. Certainly computer "radio buttons" have never exhibited that sort of behavior.
I open one as a template, modify it and then save it with a new name. "Autosave & undo" would immediately overwrite the file I use as a template.
Instinctively saving (which I tend to do about every other sentence) without changing the name first has the same effect - I finally learned (after screwing up WAY too many times) to ALWAYS change the name FIRST ("save as..."), and THEN make the changes and save again. That usage pattern is now burned into my brain, to the point that I get nervous when I see someone else modifying a template without doing a "save as" first!
What I don't understand is why word thinks that printing a document constitutes some sort of change, and it asks me to save (even if the doc is in my "temporary internet files" folder!)
OK, I get that printing might (might!) adjust the pagination or whatever, but surely any moron can see that any such changes are changes it made on my behalf, not changes I actually made! Yet year after year this bug remains...
who actually still has a radio with preset stations, where once one button is operated, all other buttons mechanically jump back into the "off" position?
But it doesn't have to be mechanical buttons - I've got a modern electronic radio, but when I press one of the presets, it switches to that station and simultaneously switches away from whatever station I was previously listening to.
For contrast, imagine if they were "checkboxes" instead of "radio buttons" - pressing one would result in listening to two stations at once, and then you'd have to go back and de-select the previous station. Only a geek could rationalize a UI like that: "But it's better that way: what if you want to listen to two stations at once? Radio buttons inhibit your freedom!")
I think radio buttons are a useful analogy, though perhaps there are better names out there - maybe something like "channels", but that focuses on one particular application of the control, rather than the functionality of the control itself.
The warning doesn't bother me - I rent from netflix, use DvdShrink to rip the disc (movie only) to an ISO file, then send the disk back. I literally never see the screens they've already got telling me not to do what I'm doing; adding more (that I also won't see) doesn't change the situation.
So a natural birth with identifiable mother and father is the same as a child born via a research institute? In that case, you wouldn't mind if the US gov phased out the volunteer army and gets darpa to fill the ranks via clones
Who's being a dick now? Of course you can't own a person - we settled that about 150 years ago (at least in the US; YMMV). If the darpa clones WANT to join the army, they're entitled to do so same as anyone else; if not, they can't be forced into it any more than anyone else. If you start with the common-sense premise that the rights and responsibilities of a person arise because of their personhood, and are independent of how they came to be a person, it all pretty much sorts itself out.
Perhaps there are as many as three separate issues:
- are babies whose genes come from one parent legally different from babies whose genes come from two parents? I can't think of any reason they would be.
- are babies who develop outside of wombs legally different from babies who develop inside wombs? Again, I can't think of any reason they would be.
- are babies born of surrogate mothers who were paid by a corporation legally different from babies born of surrogate mothers paid by an individual? I still can't think of a good reason why they would be different.
Plus of course there are many possible combinations, but even if you mix all 3 (a corporation pays for a single person's DNA to be cloned into a human and somehow has the fetus develop without the aid of a womb) I still can't see how the legal system could or should allow the corporation to "own" the person (or kill them, or whatever). Once the person reaches adulthood they're no different than anyone else; the only thing even slightly tricky is the first 18 years.
Yes, if the company somehow ditches responsibility for the person's upbringing the expense would fall on society at large - but that already happens all the time (via orphanages, adoptions, etc) and given the likely expense of the cloning process I can't see there being a significant additional burden to society.
As for being physically normal, random birth defects do and probably always will occur, but if it can be shown that the company's actions (or inactions) were negligent, I'd expect the system to treat the company much like it does a pregnant woman who smokes, drinks, does other drugs, etc (I'll admit I don't actually know how high the bar is, or what happens to the parents who exceed the threshold). Companies could go out of business, but then again mothers who smoke and drink and whatnot die and somehow we muddle through.
What rights does a clone have? Trying to build sane regulation for cloning would be a nightmare that no politician could navigate successfully.
Yes, we face that problem all the time with identical twins - deciding what rights they have is a political nightmare! The fact that clones are not just identical twins, but time-shifted identical twins makes the problem more complex than the human mind could unravel.
This is going to be really popular with the fundamentalists. Using stem cells to oppose God's Will?
Yeah, that was my first thought: They're going to kill babies to help gays? This should be fun to watch!
But no mention was made of using the money to pay artists or record labels or whatever - the summary at least says they want to avoid becoming the next greece, so it sounds like just another tax to help fund various govt operations. In that case I can't see that it changes the copyright/liability situation at all.
However, at 103 euro for a 2TB drive, that's pretty close to a 100% tax, and Moore's law says the tax rate will double every 18 months. Yikes! (yes, I know Moore's law applies to transistors, not spinning rust, but the effect is similar)
Instead, you use a data access layer, that always binds parameters.
Kinda like I said above. Only you claim that you will miss sanitizing something. So what if you forget to use bound parameters? Oh that's right, things work perfectly in your view of the world but everyone else is wrong. Use a data access layer, access everything the same way.
I don't so much care how "thick" your data access layer is - a thousand layers of code or just a rule - the important thing is that at the bottom you MUST use bound parameters instead of doubling all your quotes and wrapping it in quotes.
In-band signaling... I'll leave that for others if they want to rip it apart. I assume you mean escape sequences, replacing control characters with escapes specifically. There are common ways of replacing, and common ways to defeat common ways of replacing. It has nothing to do with in or out of band signaling.
Poor choice of words, perhaps - what it really boils down to is, don't let your users write your source code. Seems pretty obvious when you say it that way, but so many things like SQL injection attacks, XSS browser problems, etc, all come down to taking a string of user input and putting it into an environment where it gets evaluated as executable code. People see that it's happening (usually the first time Mr. O'Brien registers), and they try to patch it, but they usually fail one way or another.
For example, go back a few weeks and find the slashdot article about the voting machine being hacked (legally, during a public eval period) by some researchers. It turned out to be the wrong kind of quotes used in a shell script, which meant that a carefully crafted input ended up being executed as code. Watch over the next few weeks/months as various as-yet-unknown exploits are discussed in academic or real-world settings, and 99% of the time it ends up that user input is being executed in some way. And more often than not, there was some sort of attempt at "sanitizing" the input, which failed to account for something.
IF you have a bug in the binding, such as the case here, it doesn't matter if it's in or out of band. There is a bug, and it will likely be discovered sooner or later.
Yes, there could always be a bug in an underlying library. If the bug is in a subroutine that supposedly sanitizes your data, you're screwed (and note that there's a decent chance you won't know about the bug until someone else uses it on you). If the bug is in the SQL binding code, and the 8 bytes that's supposed to represent am IEEE floating point number happens to end up containing 'or1=1-- , then it probably doesn't matter, because no part of an SQL driver is likely to be expecting to execute the binary data of a bound parameter. And if there is someone a problem where the data packets DO try to get evaluated, you're far more likely to find it before the system hits production, because the vast majority of such attempted evaluations will fail miserably due to syntactical errors or whatever.
I only know of ONE environment where you really have no choice but to "escape" a bunch of strings, glue them together, and hope for the best: HTML. There's no equivalent of a bound parameter. And this fundamental flaw is why web pages designed by careless people (realistically, that's most of them) will always be easily exploitable, and web pages designed by careful people will also be exploitable, just not as easily and somewhat less often.
Mark my words: ten years from now, if people are still using HTML, there will still be major new types of attacks being discovered and utilized every other month or so. It's inherent in the architecture, and every new feature (javascript, CSS, etc) just introduces new escaping rules for people to fuck up..
Why oh why do people still make and use systems/apps/tools/interfaces/etc that use in-band signaling and thus require that their inputs be "sanitized"? Can't everyone see that sanitizing inputs is a fool's errand? You'll ALWAYS miss something, or the next version will have a feature you forgot to screen for, or something. In-band signaling is BAD BAD BAD and any system that uses it is doomed to an endless series of X-injection attacks.
For example (and yes, I realize this has nothing to do with SQL, it's just an example) don't even try to sanitize your SQL inputs; use bound parameters instead - not only is it guaranteed 100% safe, it's easier and faster too! As much as I love XKCD, little Bobby Tables really screwed the pooch on that one.
Remember, folks: when it comes to any sort of in-band signaling: JUST SAY NO. If you think you need to sanitize your inputs, you're doing something completely wrong. Stop and figure out what it is, and figure out how to do it right; don't just throw in some half-assed regex or character translation/stripping or whatever and hope that no one is cleverer than you are.
Did you READ TFA? He's not talking about lifting the ban on talking on the phone - he's talking about lifting the ban on having gizmos powered on during the takeoff and landing. If you can tolerate someone next to you reading a kindle or playing angry birds for 10 hours, you can tolerate it for another 30 minutes.
His real point is that he's too weak to turn his gizmos off when he wants some down time, so he wants to make sure no one else can use theirs either.
Actually, I wonder - if that's the only time he can get away from his gizmos, does he book pointless flights back and forth across the country, with as many stops as possible, just to get some quiet time?
This doesn't help much for those of us with crappy internet - I've only got about 300K (bits) upload speed, and at that speed backing up 1TB would take around a year.
FWIW, my strategy is to keep truly important stuff on a raid enclosure (and backup to other disks periodically), and to just live with the fact that there's really nothing irreplaceable about the rest.
electrons are on the same level as protons and neutrons, but quarks of of a lower order yet that phrase seems to lump them all in the same set.
It depends on which direction you're counting from: from the top down, electrons are on the same level as protons and neutrons ("constituents of atoms") - but from the bottom up, electrons and quarks are on the same level (level 0 = fundamental [as far as we know] particles ), but protons and neutrons are a level above electrons (level 1 = stuff built directly from fundamental particles)
Ah, so then the argument is that these nefarious henchmen are going to station themselves at the home of every voter that they wish to bribe while they watch them click on the pre-selected candidate? Since one or two wouldn't turn the tide of an election, we're talking 10s of thousands of conspirators who must operate in absolute secrecy and keep quiet for the rest of their lives.
An argument against internet voting due to interference by aliens from another planet would make more sense.
Yes, thousands of nefarious henchmen are implausible... but more mundane threats exist, too. Imagine an abusive husband telling his wife (and/or kids) who to vote for... Right now they can agree with him (to save their skins) and then go into the voting booth and do whatever the hell they want to do. But with him looking over their shoulders, that option disappears. Multiply this by the number of overbearing spouses out there, and it could become significant.
The problem is that the cost of securing such a system (which has to be accessible to the general populace) is very very high compared to the cost of compromising such a system.
Actually you've got a problem no matter how much you spend or how secure it actually is: even if no shenanigans have taken place - even if none COULD take place, due to some amazing design - the losers will claim that the system was rigged against them. They've always done that, of course, but have generally gotten limited traction because undetectable large-scale fraud with paper ballots is implausible. The difference is that with a fully-computerized system, fraud will seem plausible to the average person. No amount of gibberish spewed by so-called "expert cryptographers" will convince the public that fraud didn't take place. Without any brakes on the conspiracy theories, huge sections of the population will convince themselves that the gov't is a fraud (remember, it doesn't matter whether fraud actually took place or not) and that could be very bad for society.
I'll stick with paper ballots.
P.S. To prove my point about how people want to deny the results of elections, I guarantee I'll get lots of responses detailing how past elections have already been stolen. I'm not saying they have or haven't - and it's completely irrelevant to my point - but I'll get lots of responses anyway. Now imagine if there were no effective way for the average person to estimate the likelihood of these claims being true - just 49% of the country believing it was all a scam, and 51% telling the 49% to shut the hell up and quit being sore losers.