There appears to be a use case for a layout for undecoded data that, like UTF8, is variable length. An idiom of using a null layout handles top-level cases (the value is the buffer itself), but explicit expression would support variable length unions.
There appears to be a use case for a layout for undecoded data that, like UTF8, is variable length. An idiom of using a null layout handles top-level cases (the value is the buffer itself), but explicit expression would support variable length unions.