Author: Dan Gould 2023-08-11 17:03:02+00:00
Published on: 2023-08-11T17:03:02+00:00
The email is a response to someone named "waxwing" who provided a detailed response. The sender appreciates the feedback and mentions that they have fixed the flaws mentioned in the response. They discuss the security of the protocol and mention that even though they handle certain information as secret, it should not be assumed secure. The sender explains their decision to use a symmetric key over DH (Diffie-Hellman) for receiver authentication due to the inconvenience of another round of communication. They propose a solution to mitigate an attack by having the receiver share a public key of a per-request keypair. They also mention the use of BIP 47 codes and ephemeral keys to enroll multiple buffers at a relay simultaneously. The sender acknowledges the concern of relays having metadata that could be used for timing attacks and suggests a random delay based on a Poisson distribution to mitigate this problem. They refer to a research study by S. Ghesmati in 2020 which states that a significant number of transactions conform to the type of heuristic that payjoins conform to. The sender expresses reluctance to require Tor for deployment and suggests considering Oblivious HTTP instead. They mention the biggest intersection attack being timing correlation of two linked potential payjoin transactions related to one IP address and suggest a specified delay to mitigate this concern. The sender agrees that padding should be a requirement and discusses the buffer size, noting that PSBTs have significant overhead compared to consensus transactions. They thank the recipient, Dan, for the feedback.
Updated on: 2023-08-12T01:48:18.310649+00:00