On 8 Feb 2021, at 21:20, Dave Warren via mailop wrote:
On 2021-02-08 16:14, Bill Cole via mailop wrote:
[...]
The "de-tagging" tactic that Al noted has existed, although I don't
see much evidence of it in recent years. I think it may be that
enough people who use tagged addresses give tagged addresses less
scrutiny that senders who paid attention noticed that de-tagging
hurts deliverability.
There is also the possibility to give anything missing a tag extreme
scrutiny
Everything is relative :)
(or outright reject it) if a user is careful to never give out
untagged addresses.
Yes. e.g.: I have a ridiculously strict local IP blocklist fed by
automated mechanisms which sometimes list whole RIR allocations. It is
used via scoring before SMTP and in content filtering, but for untagged
addresses it is also used as an absolute ban if nothing else known at
RCPT time (i.e. recipient address tagging or other stuff) exempts the
message.
Beyond that, many years ago I turned my nominally 'main' email address
with which I use local-part tagging into a virtual address that is not
an authentication identity on any system. OTOH, the "real" address to
which my tagged addresses deliver (and which is used for authentication)
is not actually mailable. That's an extremely useful secondary effect of
intensive tagged-address usage, since the credential-stuffers all try
tagged addresses or the de-tagged 'main address" and so are easy to
identify without depending on rate or volume of attempts. No address
that anyone knows legitimately is part of any credential set.
By definition there is no consent given to a sender who just makes up
their own addresses (by stripping or changing tags), which is
significant to any sender trying to operate on an opt-in basis.
Absolutely correct.
--
Bill Cole
b...@scconsult.com or billc...@apache.org
(AKA @grumpybozo and many *@billmail.scconsult.com addresses)
Not Currently Available For Hire
_______________________________________________
mailop mailing list
mailop@mailop.org
https://list.mailop.org/listinfo/mailop
