On 28/04/2026 19:01, John R Levine wrote:
On Tue, 28 Apr 2026, Alessandro Vesely wrote:
Sometimes. My wife and I have forwarding addresses at our
universities, and I have one at the IEEE ([email protected].) They
all used to handle the forwarding
themselves, but over the years they outsourced it. Yale outsourced to
Google, Cornell to Microsoft. The IEEE has a strange system where the
mail goes to
Google, is relayed to an IEEE server at AWS, and is then reinjected
into Google to send back out.
When a forwarder moves, frequently without telling the forward
recipients, whose job is it to update your users' configuration?
If the user’s configuration is tied to a stable identifier such as a
List-ID,
Um, I think it should be self-evident that those forwarders are not
mailing lists.
Agreed. It is also self-evident that those forwarders did not obtain
any form of explicit authorization from receivers when they set up their
forwarding behavior.
That seems to be the underlying issue: we have authentication
mechanisms, but no corresponding authorization model for intermediaries
like forwarders.
IMHO, an authorization protocol should provide a stable identifier for
this kind of forwarding role, analogous (in spirit) to List-ID for
mailing lists. Without something along those lines, receivers have no
principled way to distinguish legitimate forwarding from unauthorized
mediation.
Perhaps this is getting into design details, but it relates directly to
the analysis of ARC failures: focusing on authentication alone overlooks
that authentication without authorization is of limited value —arguably
only sufficient for direct (non-mediated) delivery.
Best
Ale
--
_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]