I'll share your viewpoint to the letter when MS and the rest of the TCI cadre announce their intent to certify trust providers and publicize clear, universal, and low-cost methods for certification.
Until then, this is just another attempt to monopolize a market. (If you view inter-system communication as a market, the economics become quite clear.)
Sorry, the DMCA prevents. As long as the trusted computing crap is used to protect against copyright violations, you're prevented from circumventing it. And I don't plan on ever seeing a trusted computing scheme that doesn't include DRM, so that's that.
Unless, of course, you write an entire BIOS from the ground up... then maybe you could argue that you're simply replacing the stock software rather than disabling a copyright protection mechanism. However, I don't see how this could realistically be accomplished on a wide scale.
As long as there are multiple competing trust providers, and administrators can choose which ones to certify for interoperability with their systems, I don't see much of a problem.
Of course, the problem is that right now there is essentially only one trust provider, and its previous behavior doesn't incline me toward trusting it.
The benefit of using multiple trust certifications is that OSS could get in on the game... if someone wanted to set up a way to submit source and receive signed compiled binaries for a small fee. A bit of a hassle and in effective in the event a licensee wants to modify the code, but then again the licensee could pay the original OSS coders or submit the modified source for signing themselves.
:::So if I were to ask you for a copy of a book report you wrote for school when you were 12 years old, you would gladly and quickly provide me one?
Unless you DESTROYED PART OF HISTORY by throwing that crap away once you no longer saw value in it.
Worst. Argument. Ever.
Seriously... his gradeschool book report wasn't intended for public distribution, had no intended impact upon social evolution, and was not developed for profit.
There's a huge difference between works that were knowingly released to the public and later were "rescinded" by the creator and works that either never were intended for at-large consumption or never were deemed fit for it.
The original iPods had their share of problems too, although newer firmware revisions have pretty much taken care of it AFAIK.
I'm not familiar with the reset procedure or diagnostics built into the minis because I haven't worked on one yet, but I bet they'll end up working pretty well in about another 6 months or so. This time, Apple finally took the opportunity to follow someone else's lead in product implementation.:)
There were actually four Kenshin anime releases. The original series (95 eps IIRC), a 4-episode OVA that takes place in Japan during the Revolution, a 2-episode OVA that takes place after the TV series, and a movie that takes place after the series.
There was also a boxed set of CDs containing the soundtracks and themes.
Anyway, Rurouni Kenshin is the name of the series except for the mangled version originally released in the US. Maybe not intentionally mangled in the same way that Sailor Moon or Cardcaptora Sakura were supposed to be, just bad translations and voice acting.
Simply print out or punch out a paper ballot that the voter drops into a ballot box like they do right now. Include a machine ID+time+location+date stamp on each ballot in both plaintext and as a barcode (let the barcode use AES so that the barcode can be used to verify the time/date/ID as well) so that it is not possible to stuff the ballot box with fakes.
Since the machines are already equipped with printers and are apparently required to output individual ballots within the next few years, this should be possible.
1. 1/1000 = 0.1% of the vote. Even the Bush/Gore election in Florida wasn't this close IIRC. I suggest external audits of the code (and compilation of the binaries by the auditors) along with pre-election and post-election verification of the HD contents to virtually negate the possibility of this happening. An extremely large and complicated conspiracy could work around this, but the same could be said of the current method.
2. I was going to go point-by-point, but I guess #1 takes care of this as well.
I do agree, however, that open source would be a better option, but there seems to be some general resistance to it (misunderstanding of it, I would say). Open source simply makes everyone a potential code auditor, which would be ideal as long as you don't some some uber-perfectionists making noise or crying "wolf" every single year about perfectly serviceable code.
Actually, IIRC, were't Diebold ATMs the ones they found with blaster on them last year? Yet another reason to enforce verification of the drive's integrity before allowing the machine's result to be included.
External code auditing. If they don't like it and refuse to submit, another company gets the contract for the electronic voting systems. Steep penalties for "errors" (in addition to contract cancellation) for failing to maintaining a complete and accurate code repository would be good as well. Also, code auditors compile and retain the master HD (or whatever ) image before going gold.
To reduce the probability of a conspiracy to produce fraudulent results, the auditors could all be taken from different parties (i.e., a seven-person committee would have to have 2 reps, 2 dems, a libertarian, an independent, and two people either of other parties or at-large). As long as any group of two has veto power, that should maintain integrity as well as be a reasonably small number in order to keep Diebold "trade secrets" confidential, plus the expense should be minimal compared to the cost of deploying and maintaining the machines.
The article mentioned dropping off the ballots and equipment at the collection center; in context, I took that to mean there were some individual physical artifacts being moved around, especially after reading that the machines had an enclosed printer. If this is not the case, then the article should be clarified a bit on the point.
At any rate, as long as the machines do start printing paper ballots as required by law, this issue doesn't generalize to electronic voting as a whole anyway. Although they should fire the idiot who decided to have the capability disabled during this election if it was already implemented so he doesn't do anything else monumentally stupid.
RTFA... the machine prints a paper ballot after each voter completes the process. So unless the machines are systematically compromised there shouldn't be an issue.
Physically compromising the paper storage area presents the same risk as traditional ballots, so the risk presently by the undoubtedly insecure Diebold implementation is the only real concern. Of course, if the machines aren't accessible by any remote hosts, then bit-by-bit or hash verification of the drives before and after voting should provide some degree of safety against tampering that would generate false paper ballots (assuming as always that the master itself is unaltered in the case of bit-for-bit checking).
To clarify: I do not support present copyright laws. Providing incentive to create is very important. Little-to-no copyright protection does not provide this, but neither do the current laws. Why produce if you can continue making money off the things you developed 60 years ago?
I'm not arguing that the current system is unflawed; I'm only pointing out that ignoring the need for incentives will hamper creativity and innovation nearly as much as the idiotic maze of patents and legal obstacles we have at present.
Not so... in Spanish, WinXP's IRQ_NOT_LESS_THAN_OR_EQUAL bluescreen would be QIR_NI_MENOS_NI_EGUALDAD (I think).
Even for the newest, most stable version of Windows, it is very important for users to have equal access to the content of their error messsages. Really, you ought to be more culturally sensitive.
::: Government and Academic institutions would be more than happy to pay extra. It isn't their money.:::
Knee-jerk anti-tax reactions aside, some academic and government institutions have to monitor their spending very closely in order to remain within their budgets. While I doubt the CIA has issues of this sort, I'm sure there are at least a dozen places in your own community that do.
The major point about this I would mention is that it's necessary to provide some means of financial reward to creators so they can create at will instead of being compelled to take regular jobs. That way, truly talented creators will be able to do so full-time if desired, and other people who are either not as interested or not as talented could still receive something to encourage them to produce.
While the most recognized creative endeavours probably were not done with a profit motive (or at least not much of a profit motive), there are a lot of things that people have come up with simply because they needed to find a way to pay their bills (or maintain positive cashflow, in the case of those unlucky enough to be corporate drones:) ).
::: Legit companies can't do it for obvious reason. Black hats will not do it because if everyone takes virii seriously and install firewalls, delets mail attachments, etc., then they can't take advantage of security holes. That leaves white hats. But, writing virus that damages people's reputation is not something that white hats do, isn't it?:::
I think you're oversimplifying. There's no "this guy is black hat, and that guy is straight white hat". Even white hats disagree on what's acceptable sometimes... it's not like there is an Official Bible of Heavenly Hackerdom describing the different ranks of the hacker angelic host. I mean, some things are obvious no-no's like selling downloaded prototype docs on eBay, but not everything is pure black and white.
What I'm saying is I think it's entirely conceivable that someone who has a bit of a vigilante tendency could view such a trojan/worm as helping law enforcement or copyright holders. I mean, seriously, whoever wrote welchia probably fancied himself a white hat (or if not, at least not a black hat). Not saying I agree or disagree beause I'm not even 100% sure myself on that one, but there are probably as many different "fits" for "white hats" as there are people who want to wear them.
:::First, his bugs are really, really stupid. I wouldn't even call them intelligent. They are reactive in the same way a plant is reactive.:::
Some psychologists see this as being true of all organisms, including humans, with the difference being that more intelligent organisms are less observably reactively (alternative, more proactive).
:::Also, talk of crazy robots, is crazy itself. They have bugs, but being mal-adjusted is a gross anthropomorphism.:::
Maladjusted is the most appropriate word that comes to mind for "an entity that fails to achieve its purpose, particularly in such as way as to establish conditions opposing this purpose". Granted, the bugs could not manifest a purpose internally (lacking a will as they do); I meant this in a more general sense.
:::Simply put, we aren't even close to getting anything like human visual processing.:::
Complete agreement on this one. I was attempting to show that human visual processing (as it is currently modeled) isn't even a good model to use for computational visual processing. My background is more in the field of human capabilities than software development, so please bear with me if the explanation is somewhat unclear when crossing that boundary.
:::Regardless of what you thought you were doing, don't make computer vision researchers' problems for granted: unlike our cognitive capabilities that rely of reason and judgment, vision is a black box. We have very little understanding for what humans do to solve this problem.
?
Are you saying that human visual processing is poorly understood?
If so, I would argue with that statement (although there isn't really a true measure of when we know little vs a good bit vs virtually everything). However, I would agree that since the human visual system relies on highly parallel hardware which is partially asynchronous (as compared to computing machines) the data collected regarding its capabilities and operating parameters is largely useless for AI purposes. At least as long as current the processing model is in place. And even later on, from what I've read of asynchronous computing, it does not appear to be identical or even very similar to human cognitive processing (though it is somewhat closer).
To put this into electronics terms as best as I can, the human brain is simultaneous massively parallel and analog in its design. With all of the mess to go with both. While I am by no means a leading anything in the field, I'd go so far as to say I lost my faith in eye-witness testimony after my undergrad sensation and perception course. The brain will essentially change the input as the stimulus response propagates through the perception pathway based upon conscious and subconscious expectations or prompts (because some of those myriad parallel-processing connections exist between processing levels/stages/regions).
I'd agree that recognition and visual processing will be the key, but trying to do it "the human way" would be an godly mess of a project. Even neural networks (which are based upon earlier models of human cognitive function) differ in some significant ways from actual human neural functioning, and at best could serve as some kind of emulation environment for human-type visual processing... which sounds rather suboptimal.
The guy with the little robot bugs (I forget his name) has shown some rather remarkable results by creating a fairly simple neural network in them and then letting them learn (a) from go-live (b) by interacting with the environment (c) without any prior behavioral protocols. I do wonder if he's ended with any psychotic or otherwise maladjusted bugs from this though... it would definitely be a major drawback for this particular approach when the final product is supposed to be high-end military hardware.:)
:::(a) the democratic process is not being manipulated (by, say, intimidation of voters).:::
Hmm... it's not exactly the case that rich corporations and dirty politicans are like oil and water. You could make a case for (a), although I admit I don't have any means of digging around for information personally.
The way it looks now... IP laws will only become more idiotically restrictive (to developers) and cumbersome (to enforcement agencies). Especially with the US harassing other nations into adopting the same stupid stance, just look at the mess we pushed Australia into calling a law.
Less flexible and more expensive to maintain... what a brilliant trend to maintain with regard to the productive utilization of our knowledge base! O, hail the wisdon of these brave pioneers! [/soapbox]
:::It's irrelevant and pointless to spend time discussing the speed differences now between 64-bit and 32-bit.:::
Unless you need to make a decision on which machines to deploy in the near future...
It does matter, just not to most people, and probably not even to the people who are most rabidly touting 64-bit computing **(cough)**Athlon64fanboys**(cough)**.
:::I like the idea of being about to verify that my vote counted, but how will everyone being able to verify their vote stop dead people from voting?:::
The Smart cards are issued beginning on (insert go-live date) to everyone living who is entitled to vote, and thereafter they are issued at birth along with SSN. The cards are deactivated when a death notice is filed at (insert name of place where hospitals, coroners, law enforcement, etc send file this information).
Oh yeah, and all cards have a (pseudo)random number appended to the actual ID number that is recorded along with it when the card is created so that if someone's card is lost, stolen, or duplicated, you can still generate a new and distinguishable card by updating the random number associated with its ID# in the master database. Naturally, each card's record in the master database will have a turn-on date for when that person can vote and a turn-off bit for removing people who can no longer vote (certain types of felons, people who renounced their citizenship, whatever).
Of course, this doesn't account for methods of dealing with the inevitable human errors that will occur, but every system has those so there are already processes in place for that.
Slashdot Mirror
I'll share your viewpoint to the letter when MS and the rest of the TCI cadre announce their intent to certify trust providers and publicize clear, universal, and low-cost methods for certification.
Until then, this is just another attempt to monopolize a market. (If you view inter-system communication as a market, the economics become quite clear.)
Sorry, the DMCA prevents. As long as the trusted computing crap is used to protect against copyright violations, you're prevented from circumventing it. And I don't plan on ever seeing a trusted computing scheme that doesn't include DRM, so that's that.
Unless, of course, you write an entire BIOS from the ground up... then maybe you could argue that you're simply replacing the stock software rather than disabling a copyright protection mechanism. However, I don't see how this could realistically be accomplished on a wide scale.
As long as there are multiple competing trust providers, and administrators can choose which ones to certify for interoperability with their systems, I don't see much of a problem.
Of course, the problem is that right now there is essentially only one trust provider, and its previous behavior doesn't incline me toward trusting it.
The benefit of using multiple trust certifications is that OSS could get in on the game... if someone wanted to set up a way to submit source and receive signed compiled binaries for a small fee. A bit of a hassle and in effective in the event a licensee wants to modify the code, but then again the licensee could pay the original OSS coders or submit the modified source for signing themselves.
Heck, you can even make a computer desk where the desk is, itself, the computer case.
Does a motherboard duct-taped to the side of the hutch count?
:::So if I were to ask you for a copy of a book report you wrote for school when you were 12 years old, you would gladly and quickly provide me one? Unless you DESTROYED PART OF HISTORY by throwing that crap away once you no longer saw value in it.
Worst. Argument. Ever.
Seriously... his gradeschool book report wasn't intended for public distribution, had no intended impact upon social evolution, and was not developed for profit.
There's a huge difference between works that were knowingly released to the public and later were "rescinded" by the creator and works that either never were intended for at-large consumption or never were deemed fit for it.
The original iPods had their share of problems too, although newer firmware revisions have pretty much taken care of it AFAIK.
:)
I'm not familiar with the reset procedure or diagnostics built into the minis because I haven't worked on one yet, but I bet they'll end up working pretty well in about another 6 months or so. This time, Apple finally took the opportunity to follow someone else's lead in product implementation.
There were actually four Kenshin anime releases. The original series (95 eps IIRC), a 4-episode OVA that takes place in Japan during the Revolution, a 2-episode OVA that takes place after the TV series, and a movie that takes place after the series.
There was also a boxed set of CDs containing the soundtracks and themes.
Anyway, Rurouni Kenshin is the name of the series except for the mangled version originally released in the US. Maybe not intentionally mangled in the same way that Sailor Moon or Cardcaptora Sakura were supposed to be, just bad translations and voice acting.
::: For that price, one could get a low-end laptop.:::
Which also comes nicely packaged with a set of CTRL + ALT + DELETE keys.
Simply print out or punch out a paper ballot that the voter drops into a ballot box like they do right now. Include a machine ID+time+location+date stamp on each ballot in both plaintext and as a barcode (let the barcode use AES so that the barcode can be used to verify the time/date/ID as well) so that it is not possible to stuff the ballot box with fakes.
Since the machines are already equipped with printers and are apparently required to output individual ballots within the next few years, this should be possible.
Hmm... let me take a stab at this.
1. 1/1000 = 0.1% of the vote. Even the Bush/Gore election in Florida wasn't this close IIRC. I suggest external audits of the code (and compilation of the binaries by the auditors) along with pre-election and post-election verification of the HD contents to virtually negate the possibility of this happening. An extremely large and complicated conspiracy could work around this, but the same could be said of the current method.
2. I was going to go point-by-point, but I guess #1 takes care of this as well.
I do agree, however, that open source would be a better option, but there seems to be some general resistance to it (misunderstanding of it, I would say). Open source simply makes everyone a potential code auditor, which would be ideal as long as you don't some some uber-perfectionists making noise or crying "wolf" every single year about perfectly serviceable code.
Actually, IIRC, were't Diebold ATMs the ones they found with blaster on them last year? Yet another reason to enforce verification of the drive's integrity before allowing the machine's result to be included.
External code auditing. If they don't like it and refuse to submit, another company gets the contract for the electronic voting systems. Steep penalties for "errors" (in addition to contract cancellation) for failing to maintaining a complete and accurate code repository would be good as well. Also, code auditors compile and retain the master HD (or whatever ) image before going gold.
To reduce the probability of a conspiracy to produce fraudulent results, the auditors could all be taken from different parties (i.e., a seven-person committee would have to have 2 reps, 2 dems, a libertarian, an independent, and two people either of other parties or at-large). As long as any group of two has veto power, that should maintain integrity as well as be a reasonably small number in order to keep Diebold "trade secrets" confidential, plus the expense should be minimal compared to the cost of deploying and maintaining the machines.
The article mentioned dropping off the ballots and equipment at the collection center; in context, I took that to mean there were some individual physical artifacts being moved around, especially after reading that the machines had an enclosed printer. If this is not the case, then the article should be clarified a bit on the point.
At any rate, as long as the machines do start printing paper ballots as required by law, this issue doesn't generalize to electronic voting as a whole anyway. Although they should fire the idiot who decided to have the capability disabled during this election if it was already implemented so he doesn't do anything else monumentally stupid.
RTFA... the machine prints a paper ballot after each voter completes the process. So unless the machines are systematically compromised there shouldn't be an issue.
Physically compromising the paper storage area presents the same risk as traditional ballots, so the risk presently by the undoubtedly insecure Diebold implementation is the only real concern. Of course, if the machines aren't accessible by any remote hosts, then bit-by-bit or hash verification of the drives before and after voting should provide some degree of safety against tampering that would generate false paper ballots (assuming as always that the master itself is unaltered in the case of bit-for-bit checking).
To clarify: I do not support present copyright laws. Providing incentive to create is very important. Little-to-no copyright protection does not provide this, but neither do the current laws. Why produce if you can continue making money off the things you developed 60 years ago?
I'm not arguing that the current system is unflawed; I'm only pointing out that ignoring the need for incentives will hamper creativity and innovation nearly as much as the idiotic maze of patents and legal obstacles we have at present.
Not so... in Spanish, WinXP's IRQ_NOT_LESS_THAN_OR_EQUAL bluescreen would be QIR_NI_MENOS_NI_EGUALDAD (I think).
Even for the newest, most stable version of Windows, it is very important for users to have equal access to the content of their error messsages. Really, you ought to be more culturally sensitive.
::: Government and Academic institutions would be more than happy to pay extra. It isn't their money.:::
Knee-jerk anti-tax reactions aside, some academic and government institutions have to monitor their spending very closely in order to remain within their budgets. While I doubt the CIA has issues of this sort, I'm sure there are at least a dozen places in your own community that do.
The major point about this I would mention is that it's necessary to provide some means of financial reward to creators so they can create at will instead of being compelled to take regular jobs. That way, truly talented creators will be able to do so full-time if desired, and other people who are either not as interested or not as talented could still receive something to encourage them to produce.
:) ).
While the most recognized creative endeavours probably were not done with a profit motive (or at least not much of a profit motive), there are a lot of things that people have come up with simply because they needed to find a way to pay their bills (or maintain positive cashflow, in the case of those unlucky enough to be corporate drones
More competition = lower prices = more people who end up buying stuff to watch TV. Which is just what this country needs, of course.
/. during prime time, which may not constitute much of an improvement...
So says the man who is idling on
::: Legit companies can't do it for obvious reason. Black hats will not do it because if everyone takes virii seriously and install firewalls, delets mail attachments, etc., then they can't take advantage of security holes. That leaves white hats. But, writing virus that damages people's reputation is not something that white hats do, isn't it?:::
I think you're oversimplifying. There's no "this guy is black hat, and that guy is straight white hat". Even white hats disagree on what's acceptable sometimes... it's not like there is an Official Bible of Heavenly Hackerdom describing the different ranks of the hacker angelic host. I mean, some things are obvious no-no's like selling downloaded prototype docs on eBay, but not everything is pure black and white.
What I'm saying is I think it's entirely conceivable that someone who has a bit of a vigilante tendency could view such a trojan/worm as helping law enforcement or copyright holders. I mean, seriously, whoever wrote welchia probably fancied himself a white hat (or if not, at least not a black hat). Not saying I agree or disagree beause I'm not even 100% sure myself on that one, but there are probably as many different "fits" for "white hats" as there are people who want to wear them.
:::# Checks to see if it is allowed to send mail to AOL, for spamming purposes:::
Best. Feature. Ever.
:::First, his bugs are really, really stupid. I wouldn't even call them intelligent. They are reactive in the same way a plant is reactive.:::
:::Also, talk of crazy robots, is crazy itself. They have bugs, but being mal-adjusted is a gross anthropomorphism.:::
:::Simply put, we aren't even close to getting anything like human visual processing.:::
Some psychologists see this as being true of all organisms, including humans, with the difference being that more intelligent organisms are less observably reactively (alternative, more proactive).
Maladjusted is the most appropriate word that comes to mind for "an entity that fails to achieve its purpose, particularly in such as way as to establish conditions opposing this purpose". Granted, the bugs could not manifest a purpose internally (lacking a will as they do); I meant this in a more general sense.
Complete agreement on this one. I was attempting to show that human visual processing (as it is currently modeled) isn't even a good model to use for computational visual processing. My background is more in the field of human capabilities than software development, so please bear with me if the explanation is somewhat unclear when crossing that boundary.
:::Regardless of what you thought you were doing, don't make computer vision researchers' problems for granted: unlike our cognitive capabilities that rely of reason and judgment, vision is a black box. We have very little understanding for what humans do to solve this problem.
:)
?
Are you saying that human visual processing is poorly understood?
If so, I would argue with that statement (although there isn't really a true measure of when we know little vs a good bit vs virtually everything). However, I would agree that since the human visual system relies on highly parallel hardware which is partially asynchronous (as compared to computing machines) the data collected regarding its capabilities and operating parameters is largely useless for AI purposes. At least as long as current the processing model is in place. And even later on, from what I've read of asynchronous computing, it does not appear to be identical or even very similar to human cognitive processing (though it is somewhat closer).
To put this into electronics terms as best as I can, the human brain is simultaneous massively parallel and analog in its design. With all of the mess to go with both. While I am by no means a leading anything in the field, I'd go so far as to say I lost my faith in eye-witness testimony after my undergrad sensation and perception course. The brain will essentially change the input as the stimulus response propagates through the perception pathway based upon conscious and subconscious expectations or prompts (because some of those myriad parallel-processing connections exist between processing levels/stages/regions).
I'd agree that recognition and visual processing will be the key, but trying to do it "the human way" would be an godly mess of a project. Even neural networks (which are based upon earlier models of human cognitive function) differ in some significant ways from actual human neural functioning, and at best could serve as some kind of emulation environment for human-type visual processing... which sounds rather suboptimal.
The guy with the little robot bugs (I forget his name) has shown some rather remarkable results by creating a fairly simple neural network in them and then letting them learn (a) from go-live (b) by interacting with the environment (c) without any prior behavioral protocols. I do wonder if he's ended with any psychotic or otherwise maladjusted bugs from this though... it would definitely be a major drawback for this particular approach when the final product is supposed to be high-end military hardware.
:::(a) the democratic process is not being manipulated (by, say, intimidation of voters).:::
Hmm... it's not exactly the case that rich corporations and dirty politicans are like oil and water. You could make a case for (a), although I admit I don't have any means of digging around for information personally.
The way it looks now... IP laws will only become more idiotically restrictive (to developers) and cumbersome (to enforcement agencies). Especially with the US harassing other nations into adopting the same stupid stance, just look at the mess we pushed Australia into calling a law.
Less flexible and more expensive to maintain... what a brilliant trend to maintain with regard to the productive utilization of our knowledge base! O, hail the wisdon of these brave pioneers! [/soapbox]
:::It's irrelevant and pointless to spend time discussing the speed differences now between 64-bit and 32-bit.:::
Unless you need to make a decision on which machines to deploy in the near future...
It does matter, just not to most people, and probably not even to the people who are most rabidly touting 64-bit computing **(cough)**Athlon64fanboys**(cough)**.
:::I like the idea of being about to verify that my vote counted, but how will everyone being able to verify their vote stop dead people from voting?:::
The Smart cards are issued beginning on (insert go-live date) to everyone living who is entitled to vote, and thereafter they are issued at birth along with SSN. The cards are deactivated when a death notice is filed at (insert name of place where hospitals, coroners, law enforcement, etc send file this information).
Oh yeah, and all cards have a (pseudo)random number appended to the actual ID number that is recorded along with it when the card is created so that if someone's card is lost, stolen, or duplicated, you can still generate a new and distinguishable card by updating the random number associated with its ID# in the master database. Naturally, each card's record in the master database will have a turn-on date for when that person can vote and a turn-off bit for removing people who can no longer vote (certain types of felons, people who renounced their citizenship, whatever).
Of course, this doesn't account for methods of dealing with the inevitable human errors that will occur, but every system has those so there are already processes in place for that.