🚀 go-pugleaf

kyle@uunet.uu.net (Kyle Jones) writes:

    > Exactly.  The operative term is "the message."  No need to send the
    > Received lines back to the sender, since that person doesn't care
    > about them, even though s/he hasn't seen them.  Likewise for a third
    > party recipient.  If you think it's important for a third party to see
    > these mail transport debugging lines and other cruft then it's just as
    > important for the original sender to see them.

   Augh, no.  That's different.  The trace information is good for
   routing replies, if you're presented with a mangled From address.
   If a person sent a message, they hardly need to be sent back the
   Received headers.  They won't be sending mail to themselves.

The problem is that there are several different forwarding scenarios:

1) You are forwarding a message to a third party because you want them to
know about it, but you don't want them to reply.  In this case, VM should
be able to strip unwanted headers, as specified by a regexp variable.

2) You are forwarding a message to a third party because you want them to
handle it and they not on the same system as the sender of the original
message.  Perhaps the message should have been sent to them in the first
place.  In this case, it is important for the third party to be able to
reply.

3) You are forwarding a message to a third party because you want them to
handle it.  message.  Perhaps the message should have been sent to them in
the first place.  However, all the parties are on machines that reliably
communicate with each other, perhaps they are all in the same building.
Clearly in this case it is best to strip the unwanted headers and
OFFENSIVE and OBNOXIOUS to the third party not to.  With VM, one ends up
doing this by hand and thinking about switching to MH.

It appears to me that Kyle is arguing as though he thinks #2 is the only
reason anyone ever forwards a message.

However, I think if we all pour massive amounts of e-mail in Kyle's
mailbox he'll break down and give us the feature.  If necessary we can
mail him core dumps and train schedules.  :-)  The header-stripping code
already exists, so we just need a variable like
vm-forwarded-unwanted-headers-regexp.  (hint)

Simple poll to VM users: what percentage of the time do you wish to have
some headers stripped automatically and what percentage of the time would
you rather forward all the headers in the message?  (For me the answer is
about 95% vs. 5%.)

-- 
Enjoy,

Joe Wells <jbw@bu.edu>