[Mimedefang] AD: Roaring Penguin press release on Locked Addresses

[David F. Skoll]

Jan Pieter Cornet wrote:

> If you want more options, you might want to look at the options in
> TMDA http://tmda.net/ . What David's invented is kind of a TMDA
> "sender locked" email address, except that the sender is only locked
> upon first arrival of an email.

That's pretty much the innovation.  You don't need to know ahead of time
who will be the first (presumably legitimate) sender.  This makes it
really easy to fill in web forms:  You just say "Generate me a locked
address, please!" and cut-n-paste it into the Web form.

> David: your proposal will use the envelope return-path, which means
> that it will fall down in case of VERP <http://cr.yp.to/proto/verp.txt>

Actually, for "pure" VERP, it works fine, because the sender address
is constant for a given recipient.  However, some mailing lists use
per-message VERP, which changes the envelope sender for every posting.

> which is the case Matthew noticed, and in case of BATV
> <http://mipassoc.org/batv/draft-levine-mass-batv-00.txt>.

For now, if you expect any scheme that plays with the "local part",
then you should lock your address to the domain, which is slightly
weaker.  I do not expect any VERP-like schemes to arise that adjust
the domain part. :-)

> BATV will be somewhat more of a problem if it becomes widespread,
> but from what it looks now you can anticipate this, and extract
> the "original" address, or wildcard the variant bits.

Right.  BATV has a syntax for extracting the original address, so
we're ok.  If BATV ever becomes widespread, it should be reasonably
easy to adjust the code, assuming there are sane and widely-obeyed
standards.  (I'd never heard of BATV before; thanks!  It's certainly
a much better protection against Joe-jobs than SPF is.)

Regards,

David.