2

I have read sometimes in the web, that in the blockchain are checkpoints, so that an 51% could not the blockchain before that checkpoint even with over 51%. Is this true?

In this answer this is missing completely: Can an attacker with 51% of hash power change old blocks?

For this question i would assume that the 51% attacker would change the client he uses any way that would help to do this attack

Cheers!

  • Which question do you mean by 'this question'? If you would like to suggest an edit on the other answer, you may do so. – morsecoder Sep 15 '15 at 19:05
  • Is it really that hard to spot? There is only one question in the post besides the hyperlink... – user54512 Sep 16 '15 at 1:18
  • user54512, I just wasn't sure if you were trying to add context for what you are looking for in the answer to 'this' question (i.e. 'Do checkpoints in the block chain...'). So, by 'this' I wasn't sure if you meant the question right here, or the question you gave a link to. Sounds like the latter. – morsecoder Sep 16 '15 at 1:53
3

Checkpoints are features of certain clients, not inherent to the blockchain itself. So, if an attacker did re-write the chain all the way back to a popular checkpoint, he could maybe get the network to disagree. Some would go with the 'chain with most work wins' approach, and some would go with the 'chain that meets my checkpoints with the most work' rule.

Checkpoints do not inherently help prevent a 51% attack though. In fact, their usefulness is predicated on the assumption that a 51% attack all the way back to a checkpoint will never happen. For example, some clients skip ECDSA verification for anything older than a certain checkpoint, so it's useful to have these checkpoints because initial-block-download syncing will be faster.

  • Sorry! for beeing a bit harsh in my comment - thank you for that good answer! – user54512 Sep 16 '15 at 1:20
-1

Checkpointing is designed to protect against 51% attacks. By storing a block in the history of the blockchain at intervals and refusing to accept divergent blockchains without these blocks, they prevent an attacker from rewriting too much of the ledger’s history. However, divergent blockchains can exist legitimately, so there is a fine line between protecting against attack and legitimate operation of the blockchain. Checkpointing also runs the risk of centralizing (if a single node generates and distributes checkpoints) or splitting (if multiple nodes generate checkpoints potentially on divergent chains) the network.

  • 1
    I downvoted, because I strongly disagree. If checkpoints are necessary to prevent a 51% attack, the assumptions underlying proof-of-work are clearly wrong. – Pieter Wuille Sep 23 at 15:41
  • He didn't say "necessary", he said what it was designed to do, which is indeed to protect against 51% attacks. I agree that they are not really much of a solution, but that's no reason for a downvote... – hedgedandlevered Sep 24 at 1:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.