[ietf-dkim] Are subdomains like parent domains?
johnl at iecc.com
Mon Apr 28 17:25:26 PDT 2008
>> But I have to say, without any sort of domain blanket/coverage
>> option, it seems like something is really missing here.
I'm seeing an implicit assumption that if someone has an opinion about
mail from foo.com, they will have a similar opinion of mail from
subdomains a.foo.com or a.b.foo.com, or a.b.c.foo.com. I've been
thinking about the mail I actually see, and I am having great
difficulty finding even a small set of real life scenarios where that
For corporate or ISP domains, they either use one domain for all the
mail, e.g. aol.com, or a handful of regional or divisional subdomains,
e.g. twcny.rr.com or watson.ibm.com. The other 99% of the names under
the main domain are random hosts whose names are unlikely to appear as
mail addresses* in real mail. Even companies that outsource some of
their mail to ESPs tend to use a small number of subdomains, e.g.,
Orbitz' weekly newsletters come from my.orbitz.com. Maybe I'm
suffering from a lack of imagination, but I just can't think of
realistic scenarios where you would want to accept mail from a whole
lot of otherwise unknown subdomains of known domains.
So here's a question for people who run MTAs: when you filter, score,
sort, or otherwise manage your inbound mail stream, do you treat
subdomains the same as parent domains? Why or why not?
* - as opposed to, e.g., host names in received headers, which neither
DKIM nor ADSP care about.
More information about the ietf-dkim