[EthanHeilman]

> Or if a nation state or the central banks see it as an existential threat, they could consider it the cost of doing business? Maybe $30B to take out Algo or Solana and destroy trust in all PoS networks? That's a rounding error for them.

While you are correct that burning $30 billion dollars to destroy trust in PoS blockchains isn't that much money, I disagree that such an action would actually destroy trust in PoS blockchains. We have seen serious attacks on a number of blockchains, Ethereum for instance had enormous amounts of money stolen or destroyed via weaknesses in the blockchain. Yet Ethereum is still going strong. Bitcoin suffered 51% attacks that were used to perform double spends and Bitcoin is more valuable than ever.

It might be cheap to burn $30B to destroy a blockchain, but what if you burn $30B and the blockchain recovers 12 hours later.

[PretzelPirate]

> Ethereum for instance had enormous amounts of money stolen or destroyed via weaknesses in the blockchain.

These weaknesses weren’t due to consensus failures or protocol failures, but bugs in applications running on Ethereum. If Ethereum’s protocol allowed arbitrary funds to be stolen, that could certainly cause a loss of trust.

[wizzwizz4]

You overestimate the amount that people “investing” in crypto actually care about what it is. If my friends are anything to go by, anyway.

[PretzelPirate]

Most investors don’t care at all and blow off things like “the blockchain you’re using requires this fully centralized component”, but many players in the ecosystem that enable the speculation we see now, do care about protocol safety. If the Ethereum protocol was shown to be unsafe, they’d publicly promote safer alternatives and push their users to move.

[EthanHeilman]

"Blockchains get knocked down but they get up again." - Chumbawamba, ...probably

So two of the Bitcoin examples I gave was a consensus failure which already establishes the point, but lets do a very recent example from Ethereum:

A few months ago in August 2021 when Ethereum had a serious consensus failure and about three quarters of the clients in the network and some miners [0] forked off from the miners. How many people even noticed? [1]

> "Ethereum has weathered a bug that split the world’s most-used blockchain and opened up the risk of counterfeit Ether tokens." [2]

The issue at play is that the ability to cripple the consensus of a blockchain for the most part only impacts its availability not its security or the trust placed in that blockchain. Social consensus can just reset the bad transactions. If the theft or doublespend is big enough. We've seen that happen time and time again. They are somewhat robust but highly resilient.

Now it is possible that perhaps someone could perform an action that can not be so easily reset. For instance a huge doublespend where both parties receiving the funds are honest and have traded an object of extreme value for the doublespent funds. That is very hard to pull off. For instance how do you non-reversibly send something of that much value before the fork/doublespend/consensus bug is discovered? If you are moving something worth say 1 billion dollars in a single transaction you should probably be using an escrow service. Perhaps someone will invent a better technique for turning consensus failures into blockchain killers but so far I'm not aware of such a technique.

[0]: https://twitter.com/TimBeiko/status/1431278258222338056

[1]: https://www.theblockcrypto.com/post/115822/bug-impacting-ove...

[2]: https://www.bloomberg.com/news/articles/2021-08-30/ethereum-...

[PretzelPirate]

You’re now moving the goal post.

You said there were “enormous amounts of money stolen or destroyed” as a result of “ weaknesses in the [Ethereum] blockchain.”

The consensus issue where one client forked off isn’t evidence of that at all. Even the article you link to says it seems that the network was stable and the impact was minimal. Even in this particular attack, doing a double spend would be rather difficult.

[inter_netuser]

Why destroy it, when you can co-opt and control?

Spend $X billion, then just bleed everyone without power. Sort of like what we do now.

[davidmanheim]

Because nothing-at-stake attacks only allow double spending, not arbitrary forged transactions.