Michael Thomas wrote in
 <[email protected]>:
 |On 1/27/25 12:38 PM, Brotman, Alex wrote:
 |> Sorry, support for which arguments?  I'm not trying to be obtuse, \
 |> just trying to see if we can point you toward the information you seek.
 |
 |Papers, reports, really anything from mailbox providers and/or filter 
 |providers that detail how DKIM is being used in production. As one of 
 |the original designers, it's been pretty frustrating to not know how 
 |it's being used and what difference it actually makes. Any revamp would 
 |benefit from everybody here knowing whether and how it's making a 
 |difference, and what its deficiencies are. And more to the point: why 
 |it's deficient. I've seen a lot of assertions on that front which are 
 |actually not true.
 |
 |About the only thing I've managed to find is:
 |
 |https://www.usenix.org/system/files/conference/usenixsecurity18/sec18-hu\
 |.pdf
 |
 |but it's just a couple of researchers who aren't attached to any mailbox 
 |provider as far as I know.

Are you talking about l=?  But l= does not survive just any
modification, even invisible ones that do not change content like
reencodings etc.  And these are unfortunately performed by many
mailing-list software(s) (configurations, due to local policy).
So with l= i cannot verify your DKIM signature dependent upon how
your email client sends the message; for example, the OpenGroup
one notoriously transforms to 8-bit, so l= .. only by sheer luck.

Also mail filters (milters) see the entire body as one continuous
data stream (maybe in pieces due to size restrictions, but other
than that) without structure.  This means that a message that was
turned to MIME, or which' MIME layout was changed, does no longer
verify with l= because of the changes in the outer MIME structure;
for that l= would have to cover "some sliding text window" for
example, but that would imply milters would need to actually
understand the content of data, which they normally do not.  (Ie,
they would have to apply MIME parsing.)

In my easy-to-say-as-a-late-one opinion this never really worked,
even twenty years ago?  And while i speak, i find quite some
things from DKIM very suspicious, and i would not know how it was
meant at first.  But regardless of all that i say everybody has to
bite the bullet, throw over board what one got used to, and think
that cow that was micky mouse, in David Bowie's terms, new.  Email
must become easy again, and all those many standards in use are
not really of such.

--steffen
|
|Der Kragenbaer,                The moon bear,
|der holt sich munter           he cheerfully and one by one
|einen nach dem anderen runter  wa.ks himself off
|(By Robert Gernhardt)
|
|In Fall and Winter, feel "The Dropbear Bard"s pint(er).
|
|The banded bear
|without a care,
|Banged on himself for e'er and e'er
|
|Farewell, dear collar bear

_______________________________________________
Ietf-dkim mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to