[mightybyte]

One problem that I think blockchains could be well suited for is the storage of police evidence. Not physical evidence of course, but anything that could be turned into data. There was a scene in the show Billions where U.S. prosecutors tampered with evidence in the from of notes taken while interviewing persons of interest. Since blockchains are at their core a tamper-proof append-only data store, if those notes had been stored in a blockchain that evidence tampering would have been impossible. Any time the stakes are high enough a centralized technology can become vulnerable to people being bribed, etc.

Note that my use of the term "blockchain" here isn't limited to public blockchains. You most likely wouldn't want to store police evidence publicly on a blockchain like Ethereum. You'd want that to be private, but there do exist private blockchain consensus algorithms like BFT that do have the same tamper-proof properties without requiring a public proof of work network.

[peteretep]

Why does this need a blockchain rather than being simply cryptographic signing?

[monson]

Securely signing a cryptographic message with a provable timestamp still requires a trusted third party, as far as I know. A trusted third party can be corrupted.[1]

Writing data on a decentralized blockchain inherently provides a secure timestamp that cannot be modified without being noticed. Not to mention the extreme costs involved with trying to rewrite a blockchain's history. [2]

I disagree with the parent comment on one thing-- I would much rather trust a public blockchain with a respectable hashrate over a private blockchain. You could simply store a salted hash of the data on the public blockchain, and still keep the actual evidence private.

[1] https://en.wikipedia.org/wiki/Trusted_third_party [2] https://en.wikipedia.org/wiki/Trusted_timestamping

[peteretep]

> still requires a trusted third party

That would be the judiciary. If they -- the people with the guns and the resources of the state -- become corrupted, you'd be unwise to think that your cryptographic signatures are going to help much.

In the example given, having lodged the evidence with the court would have prevented the abuse.

[monson]

> you'd be unwise to think that your cryptographic signatures are going to help much

It was you who suggested simple cryptographic signatures in your initial reply to the parent comment. I was only pointing out that securely signing a message with a timestamp requires a cryptographic entity known as a "Trusted Third Party". Please see the first link in my original comment.

> That would be the judiciary. If they -- the people with the guns and the resources of the state -- become corrupted, you'd be unwise to think that your cryptographic signatures are going to help much.

If there was verifiable proof on a globally distributed blockchain that evidence had been tampered with by a judiciary member of a democratic country, I find it very hard to believe they would get away with it in the long term.

[peteretep]

> I was only pointing out that

No, you were pointing it out and then saying that the flaw of this system was that the judiciary could be corrupted.

> If there was verifiable proof on a globally distributed blockchain that evidence had been tampered with by a judiciary member of a democratic country

Putting aside the many many political and practical ways in which this fantasy will stay firmly a fantasy, why does this need a blockchain instead of simply a published list of documents, if this is globally distributed?

[notyourday]

How does a distributed public ledger help with it?

[mightybyte]

It provides more protection against deletion. But also note that my comment was explicitly not limited to distributed public ledgers.

[notyourday]

The only difference between a block chain and the database is that block chain is public so a rando can validate that nothing has been deleted.