>>45083
Well, that's one particular device/implementation, and tbh rather more like an API endpoint providing consensus-by-convention than an oracle, because its decisions are an algorithmic result of whatever data is fed into its inputs
In a general sense, "oracle" is often used in cryptography and infosec to refer to a process or program that can give you (useful) information in an a-priori opaque manner. In the case of cryptography it could be for example that you may via some cryptographic operation determine whether or not certain ciphertext encodes a given piece of data, while anyone else may not, or may get a different anwser (so the oracle can give you a custom answer as long as you present the appropriate offering, such as a key). Or in security parlance, oracle can be used to describe a kind of vulnerability that lets an attacker confirm and/or refute theories about secret information (thus reducing entropy of the secret) while itself not outright disclosing the secret or how the answer is determined

>>45085
>Although there is no complete solution for proving that you didn't mess with certificate (and blaming you for this while silently replacing the chain).
Hm. I guess they could just accuse you of having changed your private key after the fact, while they re-sign your ballot/certificate with another key. So it's basically the problem of proving that you are/aren't the legitimate owner of a given key. And I suppose adding a prior pre-committment phase is just pushing the problem down another layer... Here's is where distributed consensus becomes useful, except that something as important as the state relying on unproven technology is too risky
>adding another independent blockchain into voting system may solve issues
The problem with this is that saying "independent" is easy while in practice hard to define and even harder to implement and ensure