I too like the idea of giving SPV clients the tools to make effective fee estimates in a truly decentralized manner, but this seems to me like a very difficult problem given that the miners generating the data are also the people who would benefit from manipulating it.
6/ This attack, if carried out, can also be easily spotted on the blockchain.
-
-
7/ It also can be easily mitigated, as Charlie has pointed out, as long as SPV clients don’t use the same rule for estimating fee. Min feerate is only a fee signal, it’s up to each SPV client how to interpret it.https://twitter.com/SatoshiLite/status/949802676056834048 …
-
8/ In conclusion, I think this attack is unrealistic, easily spotted, and easily mitigated against.
-
I think it's unlikely, but not unrealistic. My main point was that I think the risk of this attack, or more clever variants, means that SPV clients would be better off using existing semi-centralized fee estimation services rather than the miner-curated minimum block fee info.
-
I would argue that it’s easier to manipulate these centralized fee estimation services than to manipulate min feerate consistently over X blocks- where X is a large number :-)
-
Btw, there is no reason SPV fee estimation can’t get smarter. “Median over X blocks” is just the naive algorithm. Charlie already mentioned another one like having a fallback.
-
If you want to get fancy, you can do anomaly detection / cluster analysis over a large distribution of min feerates. Given enough blocks, this kind of analysis should produce a fee estimate not too far off from a full node’s estimate.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
