#1083: USEPRO 5.3: Rules for generating message-ID

[Russ Allbery <rra@xxxxxxxxxxxx>]

I propose closing this ticket with no further changes.

There seems to be a rough consensus in the working group that the $alz
convention of issuing spam cancels with message IDs formed by adding
"cancel." to the beginning of the LHS of the original message ID to allow
for intentional collisions between multiple cancels for the same message
is the sort of weird stunt that can be left as an intentional protocol
violation.

The rest of the ticket is talking about namespace issues in deciding what
message IDs to use for messages.  USEFOR now defers to RFC 2822 for this,
and I think the treatment in RFC 2822 is sufficient.  This is a tricky
area, so if we can avoid having to say more about it and use the consensus
of the RFC 2822 working group, that saves us a lot of work and tricky
wording discussion.

-- 
Russ Allbery (rra@xxxxxxxxxxxx)             <http://www.eyrie.org/~eagle/>