[Mimedefang] GMail (was Re: stripping Received headers based on authentication)

David F. Skoll dfs at roaringpenguin.com
Wed Feb 17 13:05:29 EST 2010

Les Mikesell wrote:

>> Because running pine over SSH is not a "gateway" as defined in RFC
>> 5321, whereas running a Webmail server that accepts email using some
>> kind of transport (HTTP or HTTPS) and then delivers it using SMTP *is*
>> a gateway as defined in RFC 5321.

> Sorry, but a web interface isn't an email gateway.

Transmitting an email via HTTP from a client computer qualifies
as gatewaying by my reading of the RFC.

> The application running email isn't where the web interface displays
> it any more than it would be if you display thunderbird in a remote
> X session.

No, that's not correct at all.  Using HTTP as a transport protocol
for email is quite different from using the X protocol to draw pixels
on a remote display.

> But the RFC says nothing about remote displays and remote application
> acccess, whether provided by ssh, X, or http(s).

The RFC defines a gateway as:

   A "gateway" SMTP system (usually referred to just as a "gateway")
   receives mail from a client system in one transport environment and
   transmits it to a server system in another transport environment.

This describes webmail exactly.  You type your email into a form in the
Web browser which transmits it over HTTP to the server.  That's the
server "receiving mail from a client system".


> I wouldn't object to showing the remote display's IP address in a header
> so I'm not arguing against that, but it is somewhat imaginative to think
> a web browser is the other half of an email gateway so the RFC
> requirement would apply.

I suppose we'll have to agree to disagree.  I'd note that all other
webmail providers and open-source Webmail programs disagree with
Google.  (Hotmail has to be different, of course; it uses
X-Originating-IP: rather than a Received: header.)



More information about the MIMEDefang mailing list