Author: Christian Decker 2016-09-22 11:10:49
Published on: 2016-09-22T11:10:49+00:00
In an email exchange between Tom and Gregory Maxwell, Tom asked for a BIP number for his FT spec. However, Maxwell pointed out that the document was not concrete enough to be reviewed or implemented from it. The document did not specify the serialization of 'integer' nor did it specify how the presence of optional fields were signaled or the cardinality of inputs or outputs. Tom referred to an external specification which he believed provided the necessary details. However, Maxwell expressed concern about referencing external formatting documents and requiring readers to reverse engineer a reference implementation. He believes that BIPs should be self-contained, or rely on previous BIPs, whenever possible. Tom responded by stating that the presence of optional fields is signaled by encountering the tag, which contains both token type and name-reference. The encoder and decoder operations could be described better. He also suggested that Maxwell may have missed the third table in the document, which specifies the cardinality of inputs and outputs. However, Maxwell pointed out that the table was not well-formed and that Tom's proposal only addressed third-party malleability. Additionally, allowing miners to reject transactions with unknown fields makes the OP_NOPs unusable since they would result in forks: non-upgraded nodes would reject blocks from upgraded nodes. In conclusion, Maxwell believes that publishing a BIP with CMF and completing the spec with missing details would be better.
Updated on: 2023-06-11T20:03:22.941978+00:00