On Tue, 1 May 2007, Frank Ellermann wrote:
Pete Resnick wrote:
...
I use MIME when I forward mails or news, not Resent-*
Resent-* are *never* used when you "forward". That is also said in
quite clearly in 3.6.6 (in the Note).
Sure, but the desired effect is often the same, I got a mail and want
you to read it. I can MIME-forward or resend it to you. With MIME I
can add a reason why I think that it might interest you, with MIME I
can also forward more than mail. But if I don't pick any additional
option offered by MIME the effect is almost the same, you get the
original mail incl. original header and original body.
I'm not sure I would even call the effects of resending vs encapsulating
_similar_, much less "almost the same". The very fact that they behave
differently is a Good Thing, as it means they can express differences in
intent. When I want to add someone to a conversational thread, I resend
to them the latest message (occasionally more than one) so that their
replies naturally go back to the current set of recipients. When I want
to inform someone of something or start a side conversation, I forward it
to convey that replies probably should go back to me. They can reply
however they want, of course, but it conveys my intent and helps avoid
misaddressing.
Heck, the difference between resending and forwarding is *much* greater
than the difference between To and Cc recipients.
I don't see why there could be multiple "authors"
So you've backed off of your original argument about the need for
Resent-Sender (because it is needed when Resent-From is not identical
to the sender) and you're just complaining about Ned's example?
Not really. But it would be easy to convince me that the Resent-From
is actually unnecessary, keeping the Resent-Sender.
That would completely break backwards compatibility, rendering all
software that generates Resent-* non-compliant. If your intent is simply
to break a feature you don't like, please say so.
I think it's more important for interoperability of implementations
to keep the syntax of Resent-From and From identical (and the way
it's always been) than to go verify that nobody is using this
particular feature.
Have you seen my proposed fix for RFC 4409 8.1 ? It's clumsy if there
are both kinds, Resent-From and Resent-Sender.
...
Here's an older version of the 4409 proposal:
<http://permalink.gmane.org/gmane.ietf.rfc822/11805>
A clumsy fix is better than a wrong fix. Off hand, I think discussion of
Resent-* in a 4409bis should not be restricted to Resent-Sender, as it may
be appropriate for the MSA to add Resent-Date and/or Resent-Message-Id as
well. Also, the last two paragraphs of your suggested change at the above
URL seem undesirable or out of scope to me.
And I, and a bunch of other folks that I know, will scream bloody
murder if you take that feature out of our mail clients. Forwarding
(MIME or otherwise) does not have the same functionality as resending
MIME has more functionality, you're not forced to use it if you don't
like it. I often send mails with Content-Type: message/rfc822, with
an indicator in the subject why I did that if it's for a human reader.
1) that's not "more functionality"; that's *different* functionality.
Encapsulation is not a superset of resending.
2) "not forced to use it" Saying that after you suggest deprecating
or breaking Resent-* seems disingenuous.
There's no "Resent-subject", what are you doing that can't be done
with MIME, threading based on the original references ?
Yeah, you're right, that's *another* feature of resending that
encapsulation can't reproduce.
Philip Guenther