The core issue is that email provides communication that is unsolicited, nearly free, universal, and amenable to automation. Nobody actually wants that, because nobody wants to communicate with everybody, and nobody has enough time to communicate with everybody. There are 8 billion people in the world, and most of us communicate in multiple roles as individuals, and as part of organizations where we work or volunteer. So we have maybe 50 billion legitimate identities for communication. It will take only a tiny subset of that 50 billion to become a denial of service attack on my mailbox or on my organization's mail infrastructure. The success of Facebook is that it provides a way to control who is allowed to communicate with you. Phone service became a lot less useful when long distance calling and international calling became free or inexpensive.
So, the purpose of email filtering is to find the tiny subset of acceptable senders from the nearly infinite universe of possible senders. The fact that some of those senders are malicious is relevant, but only part of the problem. Until I blocked them, 10% of our incoming mail was from a single ESP advertising for legitimate companies. To solve the spam problem, email has to move from allow-by-default to quarantine-by-default. We allow or block mail streams, and a mail stream is defined by the origination process and the transmit flow. That has a lot of complexity, so we try to simplify by focusing on the things we know about the delivering server and the message state at delivery. This is an imperfect strategy. Bottom line is that I like ARC because it provides a little bit of useful information about how a message arrived, and that will be useful in some contexts. Doug On Sat, May 9, 2026 at 9:53 PM Murray S. Kucherawy <[email protected]> wrote: > On Mon, Apr 27, 2026 at 12:13 PM Tero Kivinen <[email protected]> wrote: > >> > I would assume what's meant here is that ARC is only truly valuable if >> you >> > develop some kind of reputation for "good" or "bad" actors in the ARC >> chain. >> > Without that, ARC's value is arguably limited. (The same argument >> applies to >> > DKIM.) >> >> Which wrong. Forwarding does not happen without the recipient knowing >> it, thus recipient will KNOW which email serverices are forwarding >> emails to him (he has set those forwarding himself). If he has set up >> the forwardings himself, he clearly already trustes the service as he >> trusts that service to process his emails, thus he can trust the ARC >> headers generated by that service. >> >> There is no need for any kind of global trust, only per recipient >> trust and that is easy for him to configure to his recipient system. > > > This presumes a trust relationship between the operator and the user. It > amounts to saying that if I tell GMail I'm forwarding mail to this mailbox > from elsewhere, GMail must trust the source I name. But why should they > trust that I'm acting in good faith and possibly expose themselves to abuse? > > Also, any trust framework requiring the user to set something up correctly > and remember to tear it down later adds technical complexity we may not > want, frustrating users who will balk at the resulting user experience. > > > If I configure this email address to forward elsewhere, does Gmail make >> any >> > assertions about any identity in the message (apart from, say, its own >> DKIM >> > signature)? I don't think that's ever true, much less generally true. >> >> But when you set up the forwarding that means you do trust google. > > > I, personally, might. But there's no reason for the two operators to > trust each other just because I said so. > > -MSK > _______________________________________________ > dmarc mailing list -- [email protected] > To unsubscribe send an email to [email protected] >
_______________________________________________ dmarc mailing list -- [email protected] To unsubscribe send an email to [email protected]
