Currently, when generating an "id" because of missing/duplicate message-id header, the location of the mail (path to the mbox, and position) is used. This is dependent on how the path is specified (absolute, relative etc.), also messages with +- identical content are not noticed as being the same thing. (See also message identification.)
Instead, hash selected message headers and entities, so as to stay independent from source of message, and avoid duplicates from getting into the output if they miss message-id, also avoid giving a warning if the message does have a message-id and is indeed virtually identical.
Cost: $120.
⇒ Cost range: $120