On Fri, 2023-07-28 at 16:29 -0400, Konstantin Ryabitsev wrote:

On Thu, Jul 27, 2023 at 01:00:15PM +0200, Andrew Lunn wrote:

quoted

quoted

Think as if instead of being Cc'd on patches, they got Bcc'd on them.

And how does reply work? I assume it would only go to those in To: or
Cc: ? Is there enough context in the headers in a reply for the system
to figure out who to Bcc: the reply to?

I have actually solved a similar problem already as part of a different
project (bugbot). We associate a set of additional addresses with a thread and
can send any thread updates to those addresses.

It would require a bit more effort to adapt it so we properly handle bounces,
but effectively this does what you're asking about -- replies sent to a thread
will be sent out to all addresses we've associated with that thread (via
get_maintainer.pl). In a sense, this will create a miniature pseudo-mailing
list per each thread with its own set of subscribers.

I just need to make sure this doesn't fall over once we are hitting
LKML-levels of activity.

How about whenever a single mailing list like
	linux-patches@vger.kernel.org
gets new 0/n without an in-reply-to header and m/n patches with
only the single in-reply-to header of an 0/n patch or simply a
single patch without an in-reply-to header, the cc list is
automatically generated from a tool like get_maintainer and a
From: <sender> line is added if necessary to the message body
and the email forwarded to all cc's and linux-patches is removed
from the email?

I believe that would help solve most correctness of recipient
list issues and then the linux-patches list would not need
further involvement.