Published on: 2018-08-16T17:32:25+00:00
In a recent conversation on the bitcoin-dev mailing list, concerns were raised about whether miners can identify transactions coming from specific software and censor them. The use of IPFS, a distributed file system, was discussed as a potential solution. It was noted that miners cannot identify transactions by their hashes until after they have already been confirmed and added to the blockchain.The discussion also touched on whether choosing not to mine transactions is considered censorship. It was argued that miners have the right to choose transactions based on their own policies, and that the system is designed to have many miners with differing policies. It was pointed out that even if transaction fees were higher, no one is obliged to accept payment for a service they don't want to provide.The topic of pruning op_return outputs was also brought up in the conversation. It was suggested that these outputs, which are not spendable, could be removed. However, it was concluded that using hash in witness script data would not be helpful and might make things worse. Recommendations were made to use pay 2 contract instead of op_return for timestamping purposes. OpenTimestamps was also recommended for actual timestamping needs. It was emphasized that op_return is used not only for timestamping but also for storing URLs and other elements that affect the hash. Batching OTS uses and IPFS directories were found to be incompatible.The possibility of miners identifying transactions from specific software and censoring them was discussed. It was acknowledged that miners could potentially censor transactions if they were able to identify them. However, it was noted that trying to inspect every OP_RETURN of every transaction would be a high-cost operation. It was also argued that choosing not to mine transactions may not be censorship economically, but politically it could be seen as such. The discussion concluded with the assertion that choosing not to mine transactions is not censorship.In a separate discussion, the use of op_return prefixes was debated. It was advised against using them due to the potential for transaction censorship. One suggestion was to remove IPFS multihash prefix information in order to minimize a miner's ability to identify transactions. Responsible ways to publish the hash, such as including it in the witness script data, were considered. However, no solid proposal for best practices has been made yet.The conversation also touched on the issue of claiming an op_return prefix. It was recommended not to use a prefix at all, as it is unnecessary from a censorship resistance and anonymity perspective. There is no formal or standard process to claim a prefix, and no existing BIP has claimed an op_return prefix.In response to a message from Lautaro Dragan, Tech Lead of Po.et, seeking guidance on using colored coins or storing data on the Bitcoin blockchain with a prefix, Luke Dashjr suggested collaborating with the authors of BIP 160 or writing a new BIP if BIP 160 is insufficient. It was emphasized that BIPs need to specify an actual protocol, not just claim a prefix.Overall, the email conversations covered various topics related to transaction censorship, op_return outputs, timestamping, and claiming op_return prefixes. Different perspectives were shared, and recommendations were made for responsible practices.
Updated on: 2023-08-01T23:42:03.316186+00:00