Dear Kurt, Many thanks for this response, I find it very helpful.
Kurt Hackenberg (2025/05/18 23:42 -0400): > On Sun, May 11, 2025 at 08:47:40AM +0100, Chris Green wrote: > > > > It sure would be a lot simpler to: > > > Eliminate commands "(un)subscribe" and "(un)lists" > > > Eliminate $auto_subscribe > > > Identify messages from lists by List-ID: > > > Make <list-reply> only send to List-Post:, and fail if it's absent > > > > > > There are more complex possibilities, too. > > > > > > > Relying on List-Id and using List-post would work for most mailing > > lists but not absolutely all of them. I have a couple where I have to > > use x-mailing-list to identiify it as a mailing list and I wouldn't be > > surprised if there are other exceptions. > > Mutt needs to know what addresses are mailing lists to generate the header > Mail-Followup-To:, which asks recipients not to send duplicate messages to > you. If <list-reply> gets destinations addresses only from List-Post:, it > can figure that out, but all the other ways to send messages -- <mail>, > <group-reply>, etc. -- can only generate Mail-Followup-To: correctly if they > know what destination addresses are lists, and Mutt gets that information > through "subscribe", "lists", and $auto_subscribe. It makes sense to me yes, thanks. > I don't see any way to eliminate that configuration of mailing lists and > still keep Mail-Followup-To: fully working. > > We could argue about whether it's worth keeping, but I'd hate to throw it > away. $auto_subscribe is easy to use, though maybe not complete. Not even mentionning the risks of backward-incompatibility. These commands should definitely stay in place, yes! > Separately from all that, I think it would be straightforward and useful to > make the $index_format expandos that identify a message as coming from a > list (%L, %T, %Z), also look at List-ID:. That would work even without > "subscribe" and such, and would also work for those Discourse messages that > aren't detected now, even with "subscribe", because they don't include the > list address in any destination address of the message. If this in indeed a reasonnably low-hangign fruit, then I can tell you it would make my everyday work at the computer much nicer. Being able to classify as reliably as possible messages just based on one caracter is, when you read in braille, a very cool feature. To push this even a bit further, I'd love to be able to use the expandos in patterns. So I could ask to see only those messages that do not have the L flag, which I would find not only infinitely cool, but also and more importantly infinitely useful. Seb.
