Alias privacy
 

Hi -
How do aliases work in terms of privacy or identification of the base email account? I'm not looking to do anything out of line, I simply want to understand what to expect.

If I have an alias of joebob@fastmail.com, is the actual account identifable in the headers or some other way?

Let's say I wanted to send an email reporting a dangerous delivery driver to the delivery business but I don't want to use my normal account. Is using aliases the way to go?

Thanks!

If you are using the web interface, your recipients have no way of seeing the account name in the headers for a send. (FastMail has a special header that allows them to trace back to the originating account if necessary.) On a reply or forward, I believe na x-delivered-to header from the original message receipt may give you away. When using an external mail client, there can be additional headers can could give away the account name.

I assume this means for legal matters, subpoenas, etc?

Thank you.

Headers and server logs
 

As BritTim correctly states, when using the Fastmail web system the alias will be shown as the From address but your main (login) Fastmail account address will not be shown in the transmitted email.

Email services use log files to keep track of each email which is sent or received. This is required to discover problems and improve performance of the system. The service provider can also examine the logs if they suspect breach of their terms of service or attacks on the system. Use for legal purposes is probably very rare.

Fastmail uses the Postfix MTA (Mail Transfer Agent) software package. When you send a message using the Fastmail web interface, one of the headers added to the message is similar to the following:
The Message-Id header is unique. No other message sent on the internet should ever have that identical Message-Id content. So if there is a question about a message, Fastmail can go back and find that Message-Id in their server logs and see what happened - which account sent the message, when it was sent, which server it was delivered to, etc.

But all the message recipient knows is the From address you use (the alias created by an identity entry). If they look at the full headers they can also see that the message was sent by the Fastmail webmail interface:
So both the X-Mailer and Message-Id headers show the recipient that the message was sent via webmail by the MessagingEngine.com system (which it's easy to discover is run by Fastmail). But your account name (login address) is never revealed.

Bill

Unlike forwarding via FastMail web UI redirecting an email via Sieve will retain all of FastMail's X-headers in the email header. "X-Resolved-to" would give the base email account away.

If I send an email from one of my aliases (using FM's webmail system) nothing in that email tells the recipient about my FM account email address.


If someone else sends an email to my alias, then in the email I receive there are headers which show it was sent to the alias, but placed in the differently-named mailbox for my account. For example such an incoming email would contain

X-Resolved-to: myaccountuserid@oneofthefmdomains
X-Delivered-to: thatalias@whicheverfmdomainitis

If you're trying to keep the underlying account userid a secret (which I certainly do, NEVER using it for email) then it would be stupid to forward such an incoming email anywhere else.

I just tried forwarding a mail and could not find the X-Resolved-to or X-Delivered-to in the forwarded message. Maybe they are removed by the FastMail system?

When forwarding via FastMail web UI Fastmail's X-headers are stripped from the email. When redirecting an email via Sieve (redirect or redirect :copy) all of FastMail's X-headers are retained.

I already have an open support ticket for the latter issue but maybe it might help if others complain of redirect's behavior as well.

PS: I forgot to mention forwarding via email client using Fastmail's SMTP server does also strip FM's X-headers (tested with MailMate).

Okay, good. Sorry, I read over your previous comment.