I think that really what is going on here is that a very small number
of people who talk a lot have prevented forward progress fixing an
issue that significantly affects many IETF participants who aren't
subscribed to ietf@ because of the noise factor and hence haven't seen
the discussion.
The ability to send replies off-list is something I would personally
like to see made harder, because in most cases these replies could
have been skipped with no damage, and in practice if someone really
wants to send an off-list reply it's pretty easy. I would just as
soon not get duplicates when people don't trim replies. FWIW, when I
reply to this list and Cc you, John, the mail bounces due to a DMARC
failure, so in practice that feature is already broken.
On Wed, Nov 2, 2016 at 1:43 PM, John Levine <johnl(_at_)taugh(_dot_)com> wrote:
In article
<CAPt1N1=_jvrNbhxDyWXpJszUtqRZEEouRibwgWD1aY5wfhsX_Q(_at_)mail(_dot_)gmail(_dot_)com>
you write:
There's a pretty clear ops problem here that could be solved by simply
detecting addresses with DMARC and rewriting the From: headers on
those messages. This would eliminate all problems immediately.
Assuming you mean replacing the actual author's address with the
list's address, that has the cost of breaking the way mailing lists
have worked for 30 years, and in particular making replies to the
author unworkable. Many people would strenously disagree that this
"eliminates all problems", but merely replaces one problem with
another.
There are other workarounds with different costs and benefits, e.g.,
the one I use that rewrites DMARC'ed addresses into local temporary
forwarding addresses, in my case in the trendy dmarc.fail domain.
That lets people keep using lists the way they have but requires more
mail system hackery than most list managers are able or willing to do.
Again, this has been discussed at great length here and on many
mail-related lists. Please see the archives.
R's,
John