Re: The need for two headers
Hadmut Danisch <[email protected]> Wed, 11 Feb 2004 23:06:36 +0100
| Newsgroups | gmane.mail.ng |
|---|---|
| Message-ID | <[email protected]> |
On Sun, Feb 08, 2004 at 09:42:02AM -0600, Steven F Siirila wrote: > > Be careful here. If I send a message to multiple recipients at a given site, > all as "Bcc:" or equivalent, I may not want all of the recipients to see the > complete delivery path information if it's going to contain all of the > recipient's addresses. That would defeat the purpose of "Bcc:". I don't think so. Bcc should be a function of the MUA as today which says: Write this address on an envelope (outer metadata), but not to the inner header, which is sent to every recipient. This has some consequences for the metadata structure. When we send a message to several recipients, one message body (and inner header) could be attached to a bundle of separate outer headers (routing metadata) for every single recipient. Whenever one MTA decides to choose different routes for the given recipients, then the routing metadata are to be split and the body copied. This means that we would need two different Bcc mechanisms. Both of them mean that the recipients do not see (in the outer header/ routing metadata or in the message header) the Bcc addresses. The weak Bcc could mean to leave the recipients bundled until different routes need to be taken. (Advantage: Less traffic. Disadvantage: MTAs see the Bcc address) The strong Bcc would mean that the sending MUA launches distinct messages from the beginning of the path. regards Hadmut