On October 29, 2005 at 22:44, Frank Ellermann wrote:
With DKIM policy controled by the domain owner, and not the
mailbox users, a mailbox user may be held "hostage" by the
domain owner on how the mailbox user can use their account.
If I can't use "my" user(_at_)isp(_dot_)example (or similar vanity host
alias(_at_)user(_dot_)isp(_dot_)example constructs) whereever it pleases me
in a mail header, then it's not more good enough, and I'd
find a better ISP. The mail header is user territory. The
limit for this "hostage" issue is the MAIL FROM from my POV.
(And for the "errors-to" fraction that's already insane)
I figured someone would raise this counter-argument :)
The problem is that the end-user may have little "real" choice.
For example, the end-user's main address may be from one of many email
service proviers. He may have used this address for years and all
his friends know him by this address. He also has used this address
as registration to various business that he purchases services from
along with using it subscribe to several mailing lists.
So if the ISP enables a strict DKIM policy, many may have little
choice but to go along with it since the cost of changing providers
is high and cumbersome.
But that's only my personal view, ISPs are free to enforce
a more radical approach, Others might accept it and try to
work with Reply-To.
Reply-to is limited. For example, some users do not want particular
email addresses to ever be exposed (due to address harvesting robots).
Therefore, a person may want key originating addresses to appear from
a given domain.
--ewh
_______________________________________________
ietf-dkim mailing list
http://dkim.org