Re: [imap5] Feature set? - was Re: Designing a new replacement protocol for IMAP

[Bron Gondwana <[e-mail]>]

On Sun, Feb 19, 2012 at 08:35:57PM +0100, Sebastian Hagedorn wrote:
> >I've seen the argument of "what if you want to forward a copy on to
> >someone else for inspection, whatever" - without forming a new copy
> >of the message.  That's actually really hard to do in the current
> >universe, because the headers will make it look extra spammy.
> 
> Not necessarily. I just resent your message to myself. Mulberry, IMO
> still the best email client, adds the following headers:
                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^

> Resent-Date: Sun, 19 Feb 2012 20:30:30 +0100
> Resent-From: Sebastian Hagedorn <[e-mail]>
> Resent-To: Sebastian Hagedorn <[e-mail]>
> Resent-Message-ID: <[e-mail]>
> X-Resent-Mailer: Mulberry/4.1.0a1 (Mac OS X)
> 
> Doesn't look spammy at all, because it's completely clear what has
> happened. I use that feature all the time.

Doesn't sound like it's the same IMAP UID to me.

So it looks like using "Resent-To" rather than "To" if Resent-To
exists handily covers your usecase.

> > You're
> >much better off forming a new message.
> 
> No, because the idea is precisely to preserve the original headers.

Sounds like you're forming a new RFC822 message stored on the server
rather than redirecting the actual original bytes...

> >Again, I really think it's the rare case.  I'd rather not complicate
> >message sending (and knowing which recipients had the message sent
> >to them) for this case.  There will still be SMTP.  Most clients
> >don't provide it for good reason - it's weird, advanced,
> >non-intuative stuff.  Do Exchange or Facebook provide such interfaces?
> 
> Facebook not, but Outlook/Exchange does ... it's called "Resend This
> Message...".

Do they attach the original?  Is the destination address stored in
a header somewhere in the saved copy so that you can see it?  Assuming
that information is stored somewhere, requiring it to be so and
parsing from that rather than extending the command is much cleaner.
It provides an audit trail of sorts.

Bron.