what does fraud proof mean?

[Coding Enthusiast]

i was browsing the forum the other day that i came across this:

#SNIP#
Nobody should use EWallets. Casual users might use lightweight nodes, preferably with fraud proofs (which don't exist yet, but hopefully will in a couple of years). As many people as possible should use their own independent full nodes -- certainly anyone who accepts transactions automatically.
#SNIP#

does it have anything to do with lightweight (SPV) wallets or is it about running core as a node but in pruning mode?

i wanted to ask what it means there but it would have been so off the topic in that thread
so i started a new topic and also because google wasn't helpful either.

[1715100722]

1715100722

[1715100722]

1715100722

[1715100722]

1715100722

[1715100722]

1715100722

[illyiller]

Satoshi described an alert system that could notify SPV nodes when an invalid block had been detected by a full node. But this was always impractical to implement up to this point. Segwit will change that. Eric Lombrozo said in an interview, "If there was a way to have fraud proofs, it would improve the security [of SPV nodes] because it would only require one whistleblower on the entire network to notice that a block is invalid and all SPV nodes could ignore that block."

So instead of SPV nodes incurring the cost of fully validating (downloading the whole blockchain to ensure validity), we could create a short proof that demonstrates a block is invalid. If so, it would only take one node to propagate this proof to the network, so that all SPV nodes would reject the invalid block.

[Coding Enthusiast]

Satoshi described an alert system that could notify SPV nodes when an invalid block had been detected by a full node. But this was always impractical to implement up to this point. Segwit will change that. Eric Lombrozo said in an interview, "If there was a way to have fraud proofs, it would improve the security [of SPV nodes] because it would only require one whistleblower on the entire network to notice that a block is invalid and all SPV nodes could ignore that block."

So instead of SPV nodes incurring the cost of fully validating (downloading the whole blockchain to ensure validity), we could create a short proof that demonstrates a block is invalid. If so, it would only take one node to propagate this proof to the network, so that all SPV nodes would reject the invalid block.

i still don't quite understand what fraud proof is.
isn't SPV wallets like Electrum connecting to a Full node (the server) to validate the transaction and also receive the headers from 10 different servers? (ref: https://en.bitcoin.it/wiki/Thin_Client_Security#Electrum)

[vdotcash]

i was browsing the forum the other day that i came across this:

#SNIP#
Nobody should use EWallets. Casual users might use lightweight nodes, preferably with fraud proofs (which don't exist yet, but hopefully will in a couple of years). As many people as possible should use their own independent full nodes -- certainly anyone who accepts transactions automatically.
#SNIP#

does it have anything to do with lightweight (SPV) wallets or is it about running core as a node but in pruning mode?

i wanted to ask what it means there but it would have been so off the topic in that thread
so i started a new topic and also because google wasn't helpful either.

https://gist.github.com/justusranvier/451616fa4697b5f25f60

[ranochigo]

No. That is not related to pruned nodes.

Satoshi described an alert system that could notify SPV nodes when an invalid block had been detected by a full node. But this was always impractical to implement up to this point. Segwit will change that. Eric Lombrozo said in an interview, "If there was a way to have fraud proofs, it would improve the security [of SPV nodes] because it would only require one whistleblower on the entire network to notice that a block is invalid and all SPV nodes could ignore that block."

So instead of SPV nodes incurring the cost of fully validating (downloading the whole blockchain to ensure validity), we could create a short proof that demonstrates a block is invalid. If so, it would only take one node to propagate this proof to the network, so that all SPV nodes would reject the invalid block.

i still don't quite understand what fraud proof is.
isn't SPV wallets like Electrum connecting to a Full node (the server) to validate the transaction and also receive the headers from 10 different servers? (ref: https://en.bitcoin.it/wiki/Thin_Client_Security#Electrum)

IIRC, Electrum now has their own server and does not connect to normal Bitcoin nodes.
Yes. However, like all SPV clients, it does not verify the blocks and just follow the longest chain that is from its peers. They will just verify the transaction against the merkle root and hope that the longest valid chain is always valid. There was a fork last year and Bitcoin Core running the lower versions were rejecting blocks that were valid to those in the higher versions. Unless, Electrum is connecting to a server with a higher version, it will also get the invalid blocks.

[Coding Enthusiast]

#SNIP#
https://gist.github.com/justusranvier/451616fa4697b5f25f60

thanks, that was so helpful.