[arc-discuss] [dmarc-discuss] A bit quiet?

Roland Turner roland at rolandturner.com
Thu Nov 26 17:40:21 PST 2015


On 11/27/2015 03:01 AM, Dave Warren via arc-discuss wrote:

> On 2015-10-29 10:49, Shal Farley via arc-discuss wrote:
>> Hence the need for deliverability monitoring, as Roland points out.
>>
>> That is, if you (the list operator) have the manpower to salt the 
>> world's mailbox providers with test accounts to monitor. Even if 
>> limited to the systems identified by the list service's outbound mail 
>> logs, that could get daunting.
>>
>> If that's not already a service for hire, maybe it should be. 
>> Probably more useful than just listing who does and who does not 
>> examine ARC headers.
>
> I'm not sure how useful it will be anyway, with all the large 
> providers making filtering decisions based on individual recipient's 
> behaviour.
>
> You can certainly get a rough idea of what is going on, but all of the 
> past mail sent to the aforementioned test account and the way the 
> "user" acts toward that (or doesn't) will impact future deliverability 
> to that one mailbox in a way that may not reflect on other mailboxes.

I suspect that you're looking for a little too much precision. Bear in 
mind that there can be additional downstream forwarding, some of it also 
DKIM-breaking. At best, using this sort of monitoring to make inferences 
about DMARC behaviour in order to make decisions about list processing 
is about the 95% case[1], you'll never get a perfect picture. That said, 
anyone operating seedboxes in this way would presumably treat them in 
such a way as to get as accurate a read about the receiver's default 
behaviour as possible (e.g. neither putting senders in their address 
book, nor complaining about them).

- Roland

1: Perhaps higher, but not 100%.


More information about the arc-discuss mailing list