Of course cable *is* (technologically) inferior to fiber. There's no doubt about it. 100Mbps would be trivial on fiber, heck 1Gbps would be trivial on fiber.
Correction: 1Gbps is trivial on fiber.
In France the number one ISP, Orange, is deploying fiber using the G-PON technology for residential service. This means 2Gbps downstream and 1Gbps upstream. Of course they don't give you access the the full bandwidth, mostly for commercial reasons. However the point is that the 'optical modem' they send you already communicates at gigabit speeds while being cheap enough to be deployed on a large scale.
#1. DSL is shared just like cable, just at a slightly different point. Several DSL customers connect to the same node and from there, they are given times slices.
It's unclear what you call a 'node'. If you mean a local exchange then yes obviously a lot of people are going to connect to it. However each ISP usually has their own DSLAMs there and the customer's line plugs directly into the ISPs DSLAM. Then upstream of the DSLAM of course every thing travels on a single 'cable'. However that is normally a fiber optic line with far more bandwidth than the DSL lines that connect to it so it's no issue. There are still cases where the DSL plugs into a competitor's DSLAM and where the traffic is then transmitted to your ISP, at cost to the ISP, via fiber optic lines. However, at least in France, that's less than something like 10% of the lines and is being eliminated as fast as possible (remember the 'at cost' part).
example. My brother has aDSL, he has ~20 people on his local node and he had an 80ms ping to his first hop. Yes, his ISP that is just down the road is an 80ms jump. If he had this dedicated connection you talked about, it would be impossible to have anything much more than 1ms to his ISP.
The extra latency you see is mostly created by the DSL error correction algorithms. Some ISPs (e.g. Free in France) let you choose the level of error correction on your DSL line which lets you adjust the tradeoffs between bandwidth, latency and packet loss.
The *only* difference between cable and other techs is that the connection between you and your node is shared, but your node to ISP is still fiber and has the same limitations of all the other techs.
That's a significant difference. With cable the choke point is where it is hardest to fix: the last mile. With fiber the choke point is where it is trivial to fix: between the local exchange and the rest of the Internet.
Case in point: some poster mentioned 'still' getting 5-6Mbps during peak hours out of 50Mbps theoretical. I have ADSL and I still get 12Mbps during peak hours which is the maximum my line supports given its length. So either his ISP really skimped on upstream bandwidth, or more likely his choke point is still in the last mile cable connection.
here in Minnesota, with the Al Franken debacle, even having paper ballots didn't help so much. With every successive challenge and recount, more ballots were magically found with Franken's name on them that had been previously undiscovered. Oh, but these were left in someone's car! Oh, these were found in a closet!
That just shows it is madness to move the ballots around before counting them. And it's even more crazy to accept new ballots that were 'found' hours if not days after the election! That's why in France the ballots must be counted right on the spot when the polling place closes (by volunteers, overseen by party representatives and anyone else who wants to). This is also why once the results are announced for that polling place the ballots are no longer authoritative.
Paper doesn't ensure integrity.
Sure paper is not sufficient to ensure integrity. No technology can provide integrity without a good set of procedures to follow. However paper is the only known technology that provides transparency, the one that has the best verifiability characteristics out of all those deployed, and with the right procedures in place it also provides pretty good integrity, with again better characteristics than any electronic system (i.e. with electronic systems flipping one thousand votes is just as easy as flipping a single vote, with paper it is a thousand times harder).
The Eyetracker also has applications in computer games where players could look around themselves without requiring a joystick to change their viewing direction
Hum, if I look to my left all I see is the wall. So will this system also move the computer screen / tv around so I see what it is trying to show me in game?
Essentially, the system relies on most people being honest, but it still seems to work reasonably well.
More importantly the system relies on an attacker having to bride a lot of people to have a meaningful impact on the election result, thus making it pretty hard to not get caught. Electronic voting does not have this type of security: bribe the right guy and you change the election result. That's why it's dangerous.
ensures anonymity by giving each voter a randomised token [...] numbers and votes are published after the vote
This also makes vote selling possible and thus is no better than having the votes be fully public (it's just more insidious).
If you consider jailbreaking the iPhone a favor to the user.
The users who are doing it would, that's why they are doing it!
It's not because it's doing what it claims to do that it's not a Trojan. It could be doing other things too. I'm not saying it does but it would certainly be the perfect vector. Hopefully someone has investigated this already...
When everyone parks their car at home at 6pm, it doesn't cause a massive power surge larger than our entire towns take at the moment.
Or it could be used to erase the demand peaks just by using your car's battery to provide the electricity needed to cook your diner and watch TV. Your car would then recharge later in the night when demand is the lowest. In essence you would be using your car battery as a house-wide UPS, meaning that this is well-known tech and does not require any paperwork or agreement from the electricity company. Also note that the typical car battery pack has more than enough juice to power your house for a good 4 hours (4800Wh in a small 100 battery pack so enough for 1kW for 4 hours).
Do you think Apple, Dell, HP, or the damn OLPC project actually develop anything?
I'll agree with you as far as Apple, Dell and HP are concerned. But the OLPC was truly innovative:
The display works in both backlit (color) and reflective (black and white) modes. As mentioned in another comment a startup is now try to industrialize this technology (Pixel Xi) and we should finally have it within a couple of years.
The display can stay on while the rest of the computer is turned off, thus allowing big power savings when reading stuff.
If it had not been for this minor boom problem, Concorde would have been a much bigger success.
The other major problem of the Concorde was its range. It would have had a much bigger success if it had had a range that allowed it to do non-stop routes linking the US to Asia, e.g. LA-Tokyo.
I'm fairly sure they can put the software on a CD and let a committee of computer scientists verify the content of the disc prior to installing it into the system so it's also a matter of proper procedure and trust in the elections committee.
But I don't have to make any such assumption with paper ballots. I can let anyone provide the transparent election box on election day, even a crook or my worst enemy. All I need is confidence that there will be opposing party representatives verifying that the transparent box is indeed empty at election start and who will oversee the whole process up until the ballots are counted. And if I don't trust the party representatives I can come and oversee everything myself if I want, it's just one day after all.
But with voting computers it does not matter that there are opposing party representatives on election day or that I can watch them start the voting computers and that the computers claim their 'virtual ballot box' is empty. Nobody in that room, even the best computer scientist, would not be able to detect fraud. So now I need to worry about who could have tampered with the computers when bringing them to the voting place, who had access to them while in storage, who programmed them, who checked the programs, etc. And obviously for most of those neither I nor party representatives were present: you cannot have 5 opposing party representatives watching the computers 24/7 all year round while they are in storage.
That's the fundamental difference: with paper ballots all that matters is what happens in the voting place on election day. With voting computers everything that happened since their conception is relevant.
Only problem is any time you can check your vote someone can force you to prove you voted 'right', so it's a bit of a conflict there...
Exactly. It removes the anonymity of the vote so it's a big change with huge implications, and is the problem what makes electronic voting so hard (and different from everything else we do in computer science).
60% of the voters don't notice if their vote is modified on the confirmation screen. I'm not sure they'll do any better with the printed ballot.
The above means hacking the voting computers at the polling station is pretty much sufficient to hack the whole election (since the machines cannot be wrong the 5% of voters who complain their printed ballot is wrong must be dumbasses who don't know what they type, this gives an 8% margin to hackers).
Barring that, your only hope of detecting fraud is that hackers (mainly town and manufacturer employees) did not have access to both systems and thus that the two computers give different results.
The result given at the polling station is given by a computer and and the counting of the ballots is done by another computer. There is a good chance they are from the same manufacturer (if you make the razor you also sell the blades, single contract for the town, no blaming it on the other manufacturer when the scanners cannot read the printed ballots). This all boils down to no real redundancy and simplifies hacking both systems.
Nobody counted the ballots by hand and as long as the results from the two machines agree there's no way you will get anyone to spend time to recount even a sample by hand.
The wait time is not related to whether you have machines or not but to the number of voters per voting booth.
Perhaps you missed the detail about computerized tabulation being much faster?
So? I too can give election results within minutes of the election close with my eight ball. But for democracy's sake I think it's more important that the people be able to participate in and oversee the counting process and convince themselves that there has been no cheating. It's even one of the defining elements of democratic elections: it's called transparency.
Because people fuck up? Their hands can shake, they can have trouble seeing the little circles to fill out, their hands can slip, all manner of fuck ups can happen when a human has to fill out a little piece of paper with time pressure added to the top.
They could get another ballot and try again. Or you could have people only vote on one issue at a time so all they have to do is fold a pre-printed piece of paper.
With this design the person makes their choice on a large screen, with easy to read categories, and gets an "are you sure this is what you want?" (which you don't have with paper)
Unfortunately 60% of voters don't notice if you change their vote on this confirmation screen! So it seems there's not much point.
The sheet has BOTH human readable and machine code so it can be scanned quickly and if there is any doubt can be recounted by hand.
Who will want to manually recount the ballots when the result is known already (except for the sore loser and conspiracy nuts who claim otherwise).
After the whole election 2000 fiasco, it should be obvious to anyone that trusting a human to make a perfect circle in a tiny box, something they only have a chance to do once every 4 years and no practice in between, is a bad idea.
Maybe they should train by voting every year instead of skipping three elections out of four.
Thinking the software never needs an update (that would be somewhat true with a trivial voting machine such as the Indian one, but not with the one you propose) so that with your scheme it's necessary to regularly remove the tamper-proof stickers to upgrade the software, then put new ones and yet maintain the illusion that the machine has not been tampered with!
Thinking the stickers protect you against tampering by the government organizing the elections (the ones with the most to lose).
Thinking the stickers protect the machines for the 11 months they spend in storage waiting for the next election, or thinking any hack done during that period would be detected before the election.
Not figuring out that the tamper proof stickers can be used for denial of service: break them and you can cast suspicion on all the votes of that computer (only to be done in precincts where your opponent is expected to make a good score).
Thinking open-source helps: this is no guarantee that the published source is the one running on the day of the election.
Thinking it's possible to verify that the voting computers are working and have not been tampered with: that's akin to verifying your computer has no virus just by browsing a couple of websites.
Thinking voting computers are inherently more usable than paper or that people actually verify their vote.
Thinking you can remove the memory cards containing the votes and ship them around without creating a thousand opportunities for tampering.
But I get the feeling that for a techie-site there is a strangely high amount of no-can-do people here...
Maybe because Slashdot has a higher proportion of software developers and that is people who understand why it's not possible.
You just assume it's not possible no matter what you do.
I claim it is possible, and any issue can be addressed.
In 2008 or so I have seen scientists in the field claim no solution exists yet (even at a theoretical stage) and that there is no proof that a solution is possible. But if you say it's possible then I guess I'll have to take your word for it. As for Internet voting: Vulnerability analysis of three remote voting methods
And come on, even fucking soda brands can print unique codes inside bottlecaps which can be used anonymously online to win or buy shit.
This has nothing to do with voting. The soda brand knows in advance that code 1234 gives you access to a special tune for your phone. There are ample confidential records to prove it somewhere in a vault. If you come to their website or in person and claim it gives you access to the Mercedes they will laugh in your face and have the means to prove you are lying.
In contrast your voting key 1234 is not associated to any candidate when it is generated. You are the one who does that when you vote. But because your vote is secret you cannot prove to anyone, not even the government, how you voted. This means that when you complain that the online records show key 1234 as voting for candidate A instead of candidate B they will laugh in your face because you have, by definition, no proof.
We currently have a flawed system which relies on perfect cooperation of thousands of individuals each one of which can influence the voting process without us knowing it.
This has lead to voting fraud not only in third world countries with 'broken election laws'
but in all western nations (at least on some local level).
In the US you are right, votes can be tampered during transport, probably by a single individual (blame the 'broken election laws'). In France (where we have transparent ballot boxes and where ballots are counted on the spot when the election closes) no single individual can tamper with the voting process. But yes, tell me how to commit fraud in France and not get caught.
There is a single point of failure, and if that fails you should to be able to detect it.
How will you know that there has been fraud if candidate A wins with 52% of the votes instead of losing with 49%? Do you pretend to know the results in advance?
Massive fraud can never occur without detection.
With voting computers it can. You only need to corrupt a handful of people to get your hack into the voting machine's software or hardware. Then you have control of how every single vote is counted.
And face it, a guy who would be coerced by his wife at home would also vote 'her' party in the polling booth, don't act like shit does not happen.
The lack of ballot secrecy has real impacts. See the impact on Chile's election results when they switched in 1958.
You say 99,9% of the people won't know if they can trust the system, but you need to trust some people's intention to honestly host a democratic election.
With any computer voting system you specifically need to trust the people who organize the elections, that is the people with the most at stake if they lose. With paper ballots you only need to be confident that any fraud that will impact the result will have to involve thousand of people and will thus be detected. Even better, you can make sure that no fraud takes place at your voting place.
You now have a technical solution for the problem, but you still need to trust the people that organize the election, this is a basic premise for any election.
With paper ballots I don't need to trust the people manning the polling place or counting the ballots because I can verify myself that they are not cheating. With voting computers I would not be able to detect cheating even if I was in the same room. With Internet voting it's even worse.
P.S. you wrote: 'can the online server write to the disk or not? If it can, then it can mess with all its content, that's all'... This is overly simplistic. If you create a standalone machine
If *I* create such a machine then *I* may be confident it will behave the way I decided. Would you trust me to have designed the machine according to specifications? Would you trust the government to use the machine as I designed it and not with some modifications that allow cheating? Do you also trust the lowly employee who installed it? The one who transported it? You have to trust *all* of them because *any one* of them could have tampered with it (or substituted it).
Of the WORM devices (CD-R, DVD-R, PROM and punch cards) only punch cards could provide the kind of physical guarantee you want, and only because it would be easy to verify the lack of a punching mechanism in the read-only columns (and here I assume there's a way to prevent feeding a card in reverse).
The problem in Indian elections is not manipulation of voting machines, but the corruption of the voter by purchasing their votes through paying money, especially to the poor and illiterate masses.
The solution to voter coercion is of course not voting computers but secret ballots. Apparently ballots are already supposed to be secret but maybe the polling places need bigger curtains.
Not true. India did have paper voting. The fraud happened before the vote count. Criminals would simply "capture" polling booths and stuff the ballot boxes.
How do the new voting machines help in this regard?
In the past they captured the booth, stuffed it, returned it and nobody noticed or corrupt officials used the stuffed ballots anyway.
Now instead they have to capture the voting machine, stuff it by pressing the right two buttons as shown on the video, and return it so that nobody notice it was missing or so that corrupt officials use the hacked results anyway.
Sorry, I fail to see the difference.
Here's a better solution:
Use transparent ballot boxes. That way it's obvious if they are stuffed before the start of the election.
Have volunteers count the votes right at the polling place. Do not ever move the ballot boxes around before counting them.
What stops people from selling their vote and going to the polling booth to vote?
What stops them is the lack of a buyer: the potential buyer knows he will have no way to verify you really voted the way he wanted so he won't waste 'money' on trying to outright buy votes (he will instead resort to propaganda).
And how do you think it's a good idea to just have a few points where people go to vote, it makes it *very* easy for people to disrupt, influence or plainly destroy votes there.
And centralizing all the votes on one server in one place solves this problem? Sure if you have massive safety and order problems then you will get voter suppression. Note though that such in countries people would typically be lacking the means to vote from home anyway.
To my knowledge polling stations are responsible for all irregularities with voting up until now
Irregularities which will be obvious to any observer (such as party representatives or regular citizens overseeing the elections).
and most of these irregularities can be prevented with proper implemented transparent e-voting.
The problem being that know one knows what this mythical 'properly implemented transparent e-voting' scheme is.
There are plenty of techniques to create a one-time code that isn't linked to you personally and can't be traced back
Except all the proposals I have seen call for the unique key being generated by the government (and generally snail mailed to you). So you have no proof that such techniques have been used by the government.
But you can't honestly tell me you're so paranoid about this that you now vote with gloves on because they might trace the fingerprints on the ballot?
I don't wear gloves because I help count the votes so my fingerprints are on all the ballots!
All jokes aside, they don't know which ballot is yours. So they would have to scan the fingerprints on a substantial percentage of the ballots to find out and they would have a hard time doing that in secret. In contrast installing a small 'security' patch that records either the votes or matches the unique keys with your identity would be pretty easy. Much easier than bugging the phone of Greece's prime minister along with those of a hundred other high ranking officials for months without getting caught for instance.
Only problem there is the unique key needs to be disposed for you to remain anonymous... but I guess you could instruct people to do so after casting their vote (if they wish to remain anonymous).
Forcing the voters to take action for their vote to remain anonymous is equivalent to making their votes public. If they erased the proof that they voted right, then they will get get their knees broken all the same. Note that this is not just a theoretical issue, it has real world effects on votes as proven by Chile's switch to secret ballots in 1958.
Here is my solution to make the process as open as possible:
[...]
- To vote at home you can use the supplied voting live-CD or use your own (it's recommended instead of your default OS), or use the kiosks supplied at voting locations.
Make it possible to vote at home and a lot of people will be coerced to vote a certain way by their spouse / parent (or you're out of this house) / children (elderly people). Make it possible to vote from any computer and companies will nicely provide computers for you, will even help you. You would be free vote the way you wanted and they would not even put you on top of the list for the next round our layoffs if you voted wrong. Vote at the kiosk against the wishes of the above parties and be assured they will be very understanding of your reservations and will surely not take any action against you.
- The voting consists of going to the voting website, verifying the origin of the site and after that select a candidate and enter the key to store the vote.
So you send your vote and the unique id the government gave you back to the government. But your vote is still anonymous because the government would never stoop so low as to match your voting key with your identity, right? And anyway if they say it cannot be done it must be true, right?
- These votes are stored on the same 'offline' drive that is currently online only with a serial cable connected to the webserver.
Who cares whether it's a serial cable or a SCSI / IDE / SATA / USB one. All that matters is: can the online server write to the disk or not? If it can, then it can mess with all its content, that's all. And if it can't... well, how do you, the average joe, know it cannot in the first place? Did you check that drive / cable in person or did you just trust some government official?
- The drive containing the votes as well as the server(s) that serves the website are on public display and the code is all opened to public scrutiny.
And the code which is on public display is the same one that's running on the server, right? You know because you compiled and installed it yourself (and so did the other 100 million plus voters).
- The server should be behind a firewall that specifically looks for any and all attacks (it should be fairly easy if you tightly define only the packets that may get trough), if there is any reason to doubt the results because of a possible breach we will know.
It's almost as simple as making sure a login procedure is secure. And login procedures have never had any security issue... well, not very often anyway.
- The results as well as the timeline of the votes is made public from the start, when the voting closes the results are known *immediately*.
So the server shows you whatever it wants you to believe the votes are in real time. So what? Besides that, do you propose to show partial results during election day? Are you sure that's a good idea? You do know that's a radical departure from current practice, right?
Before talking about how insecure the web is please note that this problem is known and well understood, so we have know what to harden the system against attacks...
What you missed totally is that the server is set up by the government and thus cannot be trusted. If you really trusted the government you would not hold elections. You would just write into law that at the end of his mandate the head of state must designate his heir^H^H^H^Hsuccessor based on the people's will.
The current voting solutions are much worse in my opinion since there are attack vectors too,
Your proposal did not eliminate any attack vector. You just added at least half a dozen even more serious vectors!
But we do know for a fact that paper elections have been rigged (desp
Nice system. So once my party governs I can simply block any further election to ever finish, just by touching a single machine.
Even better, tamper with a machine, accuse your main competitor of it thus causing a huge public outcry against them, rerun the election and win by a landslide.
The moral of the story is: you cannot rerun an election without modifying its result.
It's nice they've developed a way to transfer data at ridiculous speeds, but it does the average user no good as long as we're using mechanical hard drives.
You do realize that LightPeak is also meant to replace things like HDMI cables, right? HDMI 1.3 has a peak video bandwidth of 8.16Gb/s so that already leaves you with only 1.84Gb/s for your file transfers, something which SSDs will easily fill. Or you could just max it out by connecting two lower resolutions screens.
But LightPeak has advantages other than bandwidth. With it you can connect your computer to the TV on the other side of your living-room, or even on the other side of your house. Try doing that with HDMI or DVI!
Slashdot Mirror
No, there is another one. For time being, POE works way better on copper than on fibre.
How many cable ISPs provide POE? How many DSL ISPs do so for that matter? Right. None. So that point is irrelevant in this discussion.
Of course cable *is* (technologically) inferior to fiber. There's no doubt about it. 100Mbps would be trivial on fiber, heck 1Gbps would be trivial on fiber.
Correction: 1Gbps is trivial on fiber.
In France the number one ISP, Orange, is deploying fiber using the G-PON technology for residential service. This means 2Gbps downstream and 1Gbps upstream. Of course they don't give you access the the full bandwidth, mostly for commercial reasons. However the point is that the 'optical modem' they send you already communicates at gigabit speeds while being cheap enough to be deployed on a large scale.
This is VERY wrong.
#1. DSL is shared just like cable, just at a slightly different point. Several DSL customers connect to the same node and from there, they are given times slices.
It's unclear what you call a 'node'. If you mean a local exchange then yes obviously a lot of people are going to connect to it. However each ISP usually has their own DSLAMs there and the customer's line plugs directly into the ISPs DSLAM. Then upstream of the DSLAM of course every thing travels on a single 'cable'. However that is normally a fiber optic line with far more bandwidth than the DSL lines that connect to it so it's no issue. There are still cases where the DSL plugs into a competitor's DSLAM and where the traffic is then transmitted to your ISP, at cost to the ISP, via fiber optic lines. However, at least in France, that's less than something like 10% of the lines and is being eliminated as fast as possible (remember the 'at cost' part).
example. My brother has aDSL, he has ~20 people on his local node and he had an 80ms ping to his first hop. Yes, his ISP that is just down the road is an 80ms jump. If he had this dedicated connection you talked about, it would be impossible to have anything much more than 1ms to his ISP.
The extra latency you see is mostly created by the DSL error correction algorithms. Some ISPs (e.g. Free in France) let you choose the level of error correction on your DSL line which lets you adjust the tradeoffs between bandwidth, latency and packet loss.
The *only* difference between cable and other techs is that the connection between you and your node is shared, but your node to ISP is still fiber and has the same limitations of all the other techs.
That's a significant difference. With cable the choke point is where it is hardest to fix: the last mile. With fiber the choke point is where it is trivial to fix: between the local exchange and the rest of the Internet.
Case in point: some poster mentioned 'still' getting 5-6Mbps during peak hours out of 50Mbps theoretical. I have ADSL and I still get 12Mbps during peak hours which is the maximum my line supports given its length. So either his ISP really skimped on upstream bandwidth, or more likely his choke point is still in the last mile cable connection.
here in Minnesota, with the Al Franken debacle, even having paper ballots didn't help so much. With every successive challenge and recount, more ballots were magically found with Franken's name on them that had been previously undiscovered. Oh, but these were left in someone's car! Oh, these were found in a closet!
That just shows it is madness to move the ballots around before counting them. And it's even more crazy to accept new ballots that were 'found' hours if not days after the election! That's why in France the ballots must be counted right on the spot when the polling place closes (by volunteers, overseen by party representatives and anyone else who wants to). This is also why once the results are announced for that polling place the ballots are no longer authoritative.
Paper doesn't ensure integrity.
Sure paper is not sufficient to ensure integrity. No technology can provide integrity without a good set of procedures to follow. However paper is the only known technology that provides transparency, the one that has the best verifiability characteristics out of all those deployed, and with the right procedures in place it also provides pretty good integrity, with again better characteristics than any electronic system (i.e. with electronic systems flipping one thousand votes is just as easy as flipping a single vote, with paper it is a thousand times harder).
The Eyetracker also has applications in computer games where players could look around themselves without requiring a joystick to change their viewing direction
Hum, if I look to my left all I see is the wall. So will this system also move the computer screen / tv around so I see what it is trying to show me in game?
Essentially, the system relies on most people being honest, but it still seems to work reasonably well.
More importantly the system relies on an attacker having to bride a lot of people to have a meaningful impact on the election result, thus making it pretty hard to not get caught. Electronic voting does not have this type of security: bribe the right guy and you change the election result. That's why it's dangerous.
ensures anonymity by giving each voter a randomised token [...] numbers and votes are published after the vote
This also makes vote selling possible and thus is no better than having the votes be fully public (it's just more insidious).
If you consider jailbreaking the iPhone a favor to the user.
The users who are doing it would, that's why they are doing it!
It's not because it's doing what it claims to do that it's not a Trojan. It could be doing other things too. I'm not saying it does but it would certainly be the perfect vector. Hopefully someone has investigated this already...
When everyone parks their car at home at 6pm, it doesn't cause a massive power surge larger than our entire towns take at the moment.
Or it could be used to erase the demand peaks just by using your car's battery to provide the electricity needed to cook your diner and watch TV. Your car would then recharge later in the night when demand is the lowest. In essence you would be using your car battery as a house-wide UPS, meaning that this is well-known tech and does not require any paperwork or agreement from the electricity company. Also note that the typical car battery pack has more than enough juice to power your house for a good 4 hours (4800Wh in a small 100 battery pack so enough for 1kW for 4 hours).
Actually, there are millions of people who really do need a larger vehicle.
Right. However what they need is not a 4x4 or a pickup truck. What they need is something like this (or a smaller version thereof).
Do you think Apple, Dell, HP, or the damn OLPC project actually develop anything?
I'll agree with you as far as Apple, Dell and HP are concerned. But the OLPC was truly innovative:
If it had not been for this minor boom problem, Concorde would have been a much bigger success.
The other major problem of the Concorde was its range. It would have had a much bigger success if it had had a range that allowed it to do non-stop routes linking the US to Asia, e.g. LA-Tokyo.
I'm fairly sure they can put the software on a CD and let a committee of computer scientists verify the content of the disc prior to installing it into the system so it's also a matter of proper procedure and trust in the elections committee.
But I don't have to make any such assumption with paper ballots. I can let anyone provide the transparent election box on election day, even a crook or my worst enemy. All I need is confidence that there will be opposing party representatives verifying that the transparent box is indeed empty at election start and who will oversee the whole process up until the ballots are counted. And if I don't trust the party representatives I can come and oversee everything myself if I want, it's just one day after all.
But with voting computers it does not matter that there are opposing party representatives on election day or that I can watch them start the voting computers and that the computers claim their 'virtual ballot box' is empty. Nobody in that room, even the best computer scientist, would not be able to detect fraud. So now I need to worry about who could have tampered with the computers when bringing them to the voting place, who had access to them while in storage, who programmed them, who checked the programs, etc. And obviously for most of those neither I nor party representatives were present: you cannot have 5 opposing party representatives watching the computers 24/7 all year round while they are in storage.
That's the fundamental difference: with paper ballots all that matters is what happens in the voting place on election day. With voting computers everything that happened since their conception is relevant.
Only problem is any time you can check your vote someone can force you to prove you voted 'right', so it's a bit of a conflict there...
Exactly. It removes the anonymity of the vote so it's a big change with huge implications, and is the problem what makes electronic voting so hard (and different from everything else we do in computer science).
The problems:
Perhaps you missed the detail about computerized tabulation being much faster?
So? I too can give election results within minutes of the election close with my eight ball. But for democracy's sake I think it's more important that the people be able to participate in and oversee the counting process and convince themselves that there has been no cheating. It's even one of the defining elements of democratic elections: it's called transparency.
Because people fuck up? Their hands can shake, they can have trouble seeing the little circles to fill out, their hands can slip, all manner of fuck ups can happen when a human has to fill out a little piece of paper with time pressure added to the top.
They could get another ballot and try again. Or you could have people only vote on one issue at a time so all they have to do is fold a pre-printed piece of paper.
With this design the person makes their choice on a large screen, with easy to read categories, and gets an "are you sure this is what you want?" (which you don't have with paper)
Unfortunately 60% of voters don't notice if you change their vote on this confirmation screen! So it seems there's not much point.
The sheet has BOTH human readable and machine code so it can be scanned quickly and if there is any doubt can be recounted by hand.
Who will want to manually recount the ballots when the result is known already (except for the sore loser and conspiracy nuts who claim otherwise).
After the whole election 2000 fiasco, it should be obvious to anyone that trusting a human to make a perfect circle in a tiny box, something they only have a chance to do once every 4 years and no practice in between, is a bad idea.
Maybe they should train by voting every year instead of skipping three elections out of four.
But I get the feeling that for a techie-site there is a strangely high amount of no-can-do people here...
Maybe because Slashdot has a higher proportion of software developers and that is people who understand why it's not possible.
You just assume it's not possible no matter what you do.
I claim it is possible, and any issue can be addressed.
In 2008 or so I have seen scientists in the field claim no solution exists yet (even at a theoretical stage) and that there is no proof that a solution is possible. But if you say it's possible then I guess I'll have to take your word for it. As for Internet voting: Vulnerability analysis of three remote voting methods
And come on, even fucking soda brands can print unique codes inside bottlecaps which can be used anonymously online to win or buy shit.
This has nothing to do with voting. The soda brand knows in advance that code 1234 gives you access to a special tune for your phone. There are ample confidential records to prove it somewhere in a vault. If you come to their website or in person and claim it gives you access to the Mercedes they will laugh in your face and have the means to prove you are lying.
In contrast your voting key 1234 is not associated to any candidate when it is generated. You are the one who does that when you vote. But because your vote is secret you cannot prove to anyone, not even the government, how you voted. This means that when you complain that the online records show key 1234 as voting for candidate A instead of candidate B they will laugh in your face because you have, by definition, no proof.
We currently have a flawed system which relies on perfect cooperation of thousands of individuals each one of which can influence the voting process without us knowing it.
This has lead to voting fraud not only in third world countries with 'broken election laws'
but in all western nations (at least on some local level).
In the US you are right, votes can be tampered during transport, probably by a single individual (blame the 'broken election laws'). In France (where we have transparent ballot boxes and where ballots are counted on the spot when the election closes) no single individual can tamper with the voting process. But yes, tell me how to commit fraud in France and not get caught.
There is a single point of failure, and if that fails you should to be able to detect it.
How will you know that there has been fraud if candidate A wins with 52% of the votes instead of losing with 49%? Do you pretend to know the results in advance?
Massive fraud can never occur without detection.
With voting computers it can. You only need to corrupt a handful of people to get your hack into the voting machine's software or hardware. Then you have control of how every single vote is counted.
And face it, a guy who would be coerced by his wife at home would also vote 'her' party in the polling booth, don't act like shit does not happen.
The lack of ballot secrecy has real impacts. See the impact on Chile's election results when they switched in 1958.
You say 99,9% of the people won't know if they can trust the system, but you need to trust some people's intention to honestly host a democratic election.
With any computer voting system you specifically need to trust the people who organize the elections, that is the people with the most at stake if they lose. With paper ballots you only need to be confident that any fraud that will impact the result will have to involve thousand of people and will thus be detected. Even better, you can make sure that no fraud takes place at your voting place.
You now have a technical solution for the problem, but you still need to trust the people that organize the election, this is a basic premise for any election.
With paper ballots I don't need to trust the people manning the polling place or counting the ballots because I can verify myself that they are not cheating. With voting computers I would not be able to detect cheating even if I was in the same room. With Internet voting it's even worse.
P.S. you wrote: 'can the online server write to the disk or not? If it can, then it can mess with all its content, that's all'... This is overly simplistic. If you create a standalone machine
If *I* create such a machine then *I* may be confident it will behave the way I decided. Would you trust me to have designed the machine according to specifications? Would you trust the government to use the machine as I designed it and not with some modifications that allow cheating? Do you also trust the lowly employee who installed it? The one who transported it? You have to trust *all* of them because *any one* of them could have tampered with it (or substituted it).
This can even be enforced physically with WORM: http://en.wikipedia.org/wiki/Write_Once_Read_Many
Of the WORM devices (CD-R, DVD-R, PROM and punch cards) only punch cards could provide the kind of physical guarantee you want, and only because it would be easy to verify the lack of a punching mechanism in the read-only columns (and here I assume there's a way to prevent feeding a card in reverse).
The problem in Indian elections is not manipulation of voting machines, but the corruption of the voter by purchasing their votes through paying money, especially to the poor and illiterate masses.
The solution to voter coercion is of course not voting computers but secret ballots. Apparently ballots are already supposed to be secret but maybe the polling places need bigger curtains.
Not true. India did have paper voting. The fraud happened before the vote count. Criminals would simply "capture" polling booths and stuff the ballot boxes.
How do the new voting machines help in this regard?
Sorry, I fail to see the difference.
Here's a better solution:
What stops people from selling their vote and going to the polling booth to vote?
What stops them is the lack of a buyer: the potential buyer knows he will have no way to verify you really voted the way he wanted so he won't waste 'money' on trying to outright buy votes (he will instead resort to propaganda).
And how do you think it's a good idea to just have a few points where people go to vote, it makes it *very* easy for people to disrupt, influence or plainly destroy votes there.
And centralizing all the votes on one server in one place solves this problem? Sure if you have massive safety and order problems then you will get voter suppression. Note though that such in countries people would typically be lacking the means to vote from home anyway.
To my knowledge polling stations are responsible for all irregularities with voting up until now
Irregularities which will be obvious to any observer (such as party representatives or regular citizens overseeing the elections).
and most of these irregularities can be prevented with proper implemented transparent e-voting.
The problem being that know one knows what this mythical 'properly implemented transparent e-voting' scheme is.
There are plenty of techniques to create a one-time code that isn't linked to you personally and can't be traced back
Except all the proposals I have seen call for the unique key being generated by the government (and generally snail mailed to you). So you have no proof that such techniques have been used by the government.
But you can't honestly tell me you're so paranoid about this that you now vote with gloves on because they might trace the fingerprints on the ballot?
I don't wear gloves because I help count the votes so my fingerprints are on all the ballots!
All jokes aside, they don't know which ballot is yours. So they would have to scan the fingerprints on a substantial percentage of the ballots to find out and they would have a hard time doing that in secret. In contrast installing a small 'security' patch that records either the votes or matches the unique keys with your identity would be pretty easy. Much easier than bugging the phone of Greece's prime minister along with those of a hundred other high ranking officials for months without getting caught for instance.
Only problem there is the unique key needs to be disposed for you to remain anonymous... but I guess you could instruct people to do so after casting their vote (if they wish to remain anonymous).
Forcing the voters to take action for their vote to remain anonymous is equivalent to making their votes public. If they erased the proof that they voted right, then they will get get their knees broken all the same. Note that this is not just a theoretical issue, it has real world effects on votes as proven by Chile's switch to secret ballots in 1958.
Here is my solution to make the process as open as possible:
[...]
- To vote at home you can use the supplied voting live-CD or use your own (it's recommended instead of your default OS), or use the kiosks supplied at voting locations.
Make it possible to vote at home and a lot of people will be coerced to vote a certain way by their spouse / parent (or you're out of this house) / children (elderly people). Make it possible to vote from any computer and companies will nicely provide computers for you, will even help you. You would be free vote the way you wanted and they would not even put you on top of the list for the next round our layoffs if you voted wrong. Vote at the kiosk against the wishes of the above parties and be assured they will be very understanding of your reservations and will surely not take any action against you.
- The voting consists of going to the voting website, verifying the origin of the site and after that select a candidate and enter the key to store the vote.
So you send your vote and the unique id the government gave you back to the government. But your vote is still anonymous because the government would never stoop so low as to match your voting key with your identity, right? And anyway if they say it cannot be done it must be true, right?
- These votes are stored on the same 'offline' drive that is currently online only with a serial cable connected to the webserver.
Who cares whether it's a serial cable or a SCSI / IDE / SATA / USB one. All that matters is: can the online server write to the disk or not? If it can, then it can mess with all its content, that's all. And if it can't... well, how do you, the average joe, know it cannot in the first place? Did you check that drive / cable in person or did you just trust some government official?
- The drive containing the votes as well as the server(s) that serves the website are on public display and the code is all opened to public scrutiny.
And the code which is on public display is the same one that's running on the server, right? You know because you compiled and installed it yourself (and so did the other 100 million plus voters).
- The server should be behind a firewall that specifically looks for any and all attacks (it should be fairly easy if you tightly define only the packets that may get trough), if there is any reason to doubt the results because of a possible breach we will know.
It's almost as simple as making sure a login procedure is secure. And login procedures have never had any security issue... well, not very often anyway.
- The results as well as the timeline of the votes is made public from the start, when the voting closes the results are known *immediately*.
So the server shows you whatever it wants you to believe the votes are in real time. So what? Besides that, do you propose to show partial results during election day? Are you sure that's a good idea? You do know that's a radical departure from current practice, right?
Before talking about how insecure the web is please note that this problem is known and well understood, so we have know what to harden the system against attacks...
What you missed totally is that the server is set up by the government and thus cannot be trusted. If you really trusted the government you would not hold elections. You would just write into law that at the end of his mandate the head of state must designate his heir^H^H^H^Hsuccessor based on the people's will.
The current voting solutions are much worse in my opinion since there are attack vectors too,
Your proposal did not eliminate any attack vector. You just added at least half a dozen even more serious vectors!
But we do know for a fact that paper elections have been rigged (desp
Nice system. So once my party governs I can simply block any further election to ever finish, just by touching a single machine.
Even better, tamper with a machine, accuse your main competitor of it thus causing a huge public outcry against them, rerun the election and win by a landslide.
The moral of the story is: you cannot rerun an election without modifying its result.
It's nice they've developed a way to transfer data at ridiculous speeds, but it does the average user no good as long as we're using mechanical hard drives.
You do realize that LightPeak is also meant to replace things like HDMI cables, right? HDMI 1.3 has a peak video bandwidth of 8.16Gb/s so that already leaves you with only 1.84Gb/s for your file transfers, something which SSDs will easily fill. Or you could just max it out by connecting two lower resolutions screens.
But LightPeak has advantages other than bandwidth. With it you can connect your computer to the TV on the other side of your living-room, or even on the other side of your house. Try doing that with HDMI or DVI!