Software of such importance should be scrutinized by the people, and ultimately owned by the people as an apparatus of the Government. It should be fully transparent, from the software it runs, to the hardware that it runs on.
A secure ballot collection and counting device should not become less secure because of its source code being viewable.
Checksums and hashes can be used to quickly validate that devices are running untainted images of the software.
It can be written by the greatest minds on either side of the aisle, and be reviewed and collaborated on by the other side of the aisle, through cyclic bipartisan committees that over see the development.
It can be versioned, tracked, and become an important part of the US Historical Libraries going forward.
It needs to go all the way to the operating system and hardware with no network connectivity whatsoever - not just a piece of software that can run on anything.
I guess it could be an open source thing still, but the whole implementation would need to be air tight to prevent any sort of confusions that can happen from external factors.
I don't disagree at all. All I'm saying is just making some software with common hardware that can run a myriad of other things will keep it so there will always be a back door. The only purpose of the machine should be to tally and report the votes with auditing capabilities. It should ONLY be able to do that and run only what is supposed to be running to do that. It has to be tamper proof...
It can be a piece of software that can run on any open source operating system, on any chipset that supports the instructions.
That's not hard to come up with then. Just make some sort of "open vote" project and do it. If you make everything abstract enough there can be a wide variety of implementations to accommodate for a multitude of situations.
Running on common hardware will just make it so it is susceptible to manipulation so I guess you can come up with another way to combat that (this would be the difficult part)... Many will say how a block chain implementation might help combat that but there could be situations where the whole chain can go to shit since it got full of invalid votes.
It should be "easy" enough to choose hardware because I believe ballot counting devices should be dumb. No internet access. Just input via screen/ballot, and output via screen/printout
I don't have the all the deets, because it's all very much theoretical.
You won't know if the same source code is loaded on the machine, or if somebody screwed with the chips themselves though
hashes, checksums, and hardware diagnostics are relatively trivial if everything is open source and works on a standard chipset.
Relatively trivial, is obviously something to be taken with a grain of salt, but software leaves a signature. Hardware people would have a better answer than me on how to diagnostic that portion, but it is entirely feasible.
That's why I was mentioning having to make everything down to the hardware a very specific thing that's not common and having it's sole purpose be for voting.
I'm going to guess there will be things mentioned about hashing and verification of sorts that could work; but that then could be defeated and made to work again and so on - now we have the cat and mouse game going on and then in the meantime we have our little friend fraud coming in wrecking it anyway since whatever would come out has the possibility of being tainted.
It needs to be on blockchain