On Wed 20/Nov/2024 03:09:19 +0100 Steven M Jones wrote:
On 11/20/24 06:43, Murray S. Kucherawy wrote:
On Mon, Nov 18, 2024 at 9:11 AM Dave Crocker <d...@dcrocker.net> wrote:
On 11/6/2024 6:54 PM, Wei Chuang wrote:
"message algebra
This topic has been a point of fascination for some years. It is,
indeed, attractive. >>>
It is also a research topic, absent a body of experience showing it
works on essentially all email traffic. >>>
That's not meant to argue against it, but rather to place the construct
in an area of unknown reliability, efficacy and usability, so that there
is effort to move to a place of engineering knowns.
Indeed, I remain puzzled by the number of times the idea has been dismissed
as unworkable in contrast to the excitement it appears to be generating today.
That said, it's my assumption that this renewed effort will include a
significant period of running something like this at scale with appropriate
interoperability testing to verify its efficacy before sending it to the
Standards Track, much as we did with DKIM itself.
I called out this recurrence in my response to one of Bron's separate threads,
and it is part of why I want to see this project - with so much enthusiasm and
so many engineering resources promised - explore the topic thoroughly.
Where between 50 and 100% will the percentage of expressible message
alterations fall, and according to whose corpus? Will we have a great enough
diversity of contributors to make sure we aren't missing important use cases?
Just the project of cataloging common message alterations seems like a very
useful exercise. (Not because I can't draw up a list, but because I want to see
the ones I would miss that other people have observed.)
If the intent of DKIM2 is to preserve message authenticity and semantics,
rather than recording /any/ change it should concentrate at delimiting the
allowed changes. For example, an 80 chars subject-tag would completely hide
the original subject. Ditto for extra long HTML insertions. Not to mention
malicious changes, such as different bank account number.
If DKIM2 can forbid conversions to base64, it can as well impose safe limits to
message changes. Most mailing lists are already self-imposing such limits.
Best
Ale
--
_______________________________________________
Ietf-dkim mailing list -- ietf-dkim@ietf.org
To unsubscribe send an email to ietf-dkim-le...@ietf.org
