Re: [Evolution] X-evolution-account setting

[W Van Snyder <van snyder sbcglobal net>]

I looked at another message that I had sent.  I had copied it from my
"Sent" folder to one of my storage folders.  In addition to
X-evolution-account, it has

Message-id: <1333140299 29286 3 camel math jpl nasa gov>

Where does my machine name (math.jpl.nasa.gov) come from?  My smtp
server is a different (institutional) server, running Outlook (I think),
named jpl.nasa.gov.  The IT guys tell me it's my computer's name
appearing in headers (I hadn't noticed it in Message-id earlier) that is
causing trouble with DMARC when I try to send to some accounts.  The
DMARC complaint is that my message isn't authenticated.  But my
Evolution settings require a "login" transaction with smtp.jpl.nasa.gov.
The IT guys did something to the smtp server that makes it possible to
send to gmail.com, but now I can't send to berkeley.edu.

Are they still blowing hot air at me, or is it possible that my
computer's name appearing in Message-id is causing the DMARC problem?
Can I convince Evolution to use the smtp server's name, not my
computer's name?  That seems dangerous because a message ID generated by
my computer could, in principle, clash with an ID for a message ID
generated by the smtp server for an entirely different message.  Or is
the smtp server supposed to replace the message ID from my computer with
one of its own?

On Wed, 2019-01-09 at 12:46 +0100, Milan Crha via evolution-list wrote:
> On Wed, 2019-01-09 at 11:22 +0000, Pete Biggs wrote:
> > The X-Evolution-* headers are not exposed to the outside world -
> > certainly not on recent versions.  Some (perhaps all?) of the X-
> > Evolution-* headers don't even really exist, they are inserted by
> > Evolution when the mail is displayed and are not present in the
> > stored emails.
>
>       Hi,
> that's true. There is a set of "helper" headers used by Evolution
> itself when sending mails. Those can be visible in the Outbox folder
> and they are removed before the message is passed to the transport
> service. It's not that obvious when the Outbox folder is skipped, which
> is the default behavior for some time now (the message is sent
> immediately when the Send is clicked in the composer window).
>
> Such headers are X-Evolution-Identity (the From account UID),
> X-Evolution-Fcc (the sent folder), X-Evolution-Transport (the transport
> service UID, usually associated with the From account). Those UIDs used
> to look like an email address, containing the "localhost". Newly
> created UIDs look like SHA256 hash, thus they do not expose anything
> from the local machine (some users consider it a private information).
> The X-Evolution-Account header used to be used in the past and it
> contains the same value as the X-Evolution-Identity.
>
> I suppose the headers are left in the message only because:
>
> On Tue, 2019-01-08 at 16:04 -0800, Van Snyder via evolution-list wrote:
> > My outgoing mail isn't sent directly from my own computer.  It's sent
> > from an smtp server.
>
> which I understand as the messages are stored into some specific Outbox
> (not the Evolution's) and some other server/sendmail/script picks
> messages from there and pushes them into the world. I cannot imagine
> other reason why that particular header would be left in the outgoing
> message, especially if evolution itself would connect to the SMTP
> server (+/- bugs in the code). I can be wrong, though.
>       Bye,
>       Milan
>
>
> _______________________________________________
> evolution-list mailing list
> evolution-list gnome org
> To change your list options or unsubscribe, visit ...
> https://mail.gnome.org/mailman/listinfo/evolution-list