Alan Hodgson wrote:


Rejecting the message during receipt causes the sending server to generate a bounce. If it's at all functional.
----

        That used to happen on poorly implemented mailing
lists -- a delivery error would be bounced back to the email list as a reply that would get sent out to all the subscribers.
        Would it even be practical to bounce a deliver
failures back to the original poster of the message to the list?

On a list with 10,000 subscribers or more with maybe ~1000 messages/day, how many people would be getting back mail-delivery failures that they could do nothing about.

        Many or most hosted email services provide a
user-controllable spam filter.  The problem is, if an
email is not accepted, rather than being delivered and filtered
by the "per user-filtering", the user can not tailor such filters to their own mail.

        It preempts the oft needed ability for the user
to judge what is spam and what is not.

        I am not suggesting sending a "bounce back message",
but have the default be to do what the user configures
via their account options.

        If a given user wants emails to be dropped at the
border -- that would be fine.  *I* would not mind configuring
a filter that dropped some incoming emails, but if it is
going to make the incoming mail server too slow to handle
per-user options, it might not be doable.

        Even without per-user options, many servers
that try to do filtering between reading the message and sending a response end up having periods of unacceptable
response time.

NTL, running filters over incoming email when the user has explicitly ask for unfiltered email is reprehensible. I do my own email filtering on my home server. I find ISP's doing their own filtering and
modification of incoming user messages based
on their criteria to be a very bad situation.

        Also, someone mentioned safe-harbor provisions.
Those were designed for websites where material is
hosted for public view on the host's computers. The
wording in the US act applies to those who *publish
information by others* (like a website).  It doesn't
apply to pass-through services like ISP's and
email services.

        Email and ISP service has been more held
in line with services provided by the phone company --
that of common carrier status that is predicated
upon NOT inspecting content as it travels through
a carrier's equipment.

        That said, different countries will have
different laws, and the laws are changing as email-provider demonstrate their ability to monitor
and filter real-time communications.  Some countries
are moving to have email providers be responsible for
filtering or allowing discussion of illegal acts.
That's not a good direction to be going, IMO.




Reply via email to