So today Trump was taking about how he wouldn't mind if someone shot the media at his rallies. He also said he'd probably go along with Kennedy Jr's proposal to ban vaccines and stop water fluoridation. So uh... get your covid boosters while you still can, folks – before they go back in the Kennedy Vault.
Secrecy is not primary importance though: most important is consistency (i.e. accurate recording of ballots) is primary, followed by durability (ballot data remains committed once received). ACID compliant would be great, for all you programmers out there.
But also Pennsylvania's process does not ensure security. Providing the extra envelope, it provides some additional security through obscuring, and prevents some degree of accidental disclosure. But a voter who doesn't use the security layer is disenfranchised while still having their security compromised. This is in fact worse, because an adversarial pollworker can read the invalid ballot, then tell people who can pressure the voter to vote differently, or not to vote again provisionally. If it was simply accepted, none of extra pressure exists.
Security is always in conflict with usability. Any security method which relies on the end user to do everything correctly is going to fail if your end users are humans. This is just bad design. They are optimizing for the wrong things. The more steps, the higher the chance of failure, which only serves to disenfranchis people who are trying to vote correctly, while preventing no voter fraud.
(11-03-2024, 06:46 PM)nemonowan Wrote: Well, if you were writing the election laws from the point of view that secrecy is of primary importance, it makes a lot of sense to require a privacy envelope.
The process would then be, some workers recieve the ballots, do whatever verification is required, open them and put the ballot in its second privacy envelope in a box with all the rest of the ballots processed that shift. Then different workers will process the box with its ballots, open them and count them.
That way no election worker would know for certain who voted for which candidate. If the ballot came in a single envelope they could not avoid knowing it.
It becomes messy of course when the laws are not clear or not fair on what to do if the privacy envelope is missing but the ballot inside is otherwise valid. More so if said envelope is missing because of mistakes made by the state.
Secrecy is not primary importance though: most important is consistency (i.e. accurate recording of ballots) is primary, followed by durability (ballot data remains committed once received). ACID compliant would be great, for all you programmers out there.
But also Pennsylvania's process does not ensure security. Providing the extra envelope, it provides some additional security through obscuring, and prevents some degree of accidental disclosure. But a voter who doesn't use the security layer is disenfranchised while still having their security compromised. This is in fact worse, because an adversarial pollworker can read the invalid ballot, then tell people who can pressure the voter to vote differently, or not to vote again provisionally. If it was simply accepted, none of extra pressure exists.
Security is always in conflict with usability. Any security method which relies on the end user to do everything correctly is going to fail if your end users are humans. This is just bad design. They are optimizing for the wrong things. The more steps, the higher the chance of failure, which only serves to disenfranchis people who are trying to vote correctly, while preventing no voter fraud.
"Kitto daijoubu da yo." - Sakura Kinomoto