-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Tuesday, August 26 at 04:57 PM, quoth Rocco Rutte:
> First, at least one mailing lists considers setting Mail-Followup-To
> headers to achieve exactly that (no duplicate messages) being
> "rude", e.g. http://marc.info/?l=git&m=121218565402351&w=2 though I
> consider that way of handling a mailing list "rude", too since they
> still time from me by forcing me to deal with duplicate messages.
Ye gods...
1. It's not rude to others: it merely indicates a preference. If you
want to reply to someone privately, you may do so. If this moron
cannot configure his email program to do that, then that's his
problem for using a lousy email program. When *I* want to reply to
someone on a mailing list privately, I press 'L', and my mailer
(mutt) constructs a message with only the original sender as a
recipient. The MFT header is for LIST REPLIES (which, for me, is
'r' whenever I'm viewing a mailing list message).
2. Yes, placing SOMEONE ELSE in your MFT is kinda goofy (or rude, I
suppose). That's why the only things that should be allowed in
your MFT are your own address and the mailing list's address. I
can't imagine why mutt would include him in the MFT header; that
seems broken to me.
I agree with Junio that Stephan shouldn't have put HIS address in the
MFT, but the idea that an MFT header is rude in general is absurd.
Mutt's documentation says:
The [MFT] header will contain ONLY the list's address for
subscribed lists, and both the list address and your own email
address for unsubscribed lists.
(emphasis mine). I can only imagine that Stephan must have used an
overly generous regex (such as one ending in .*) to specify the
mailing list address, in order to get the MFT to include Junio's
address.
> Second, since mutt already knows about List-Post to reply to a list
> list even without a "subscribe" or "list" command, it should be
> taught how to detect mailing lists completely. That way only mailing
> lists without a given set of headers would require "list" commands.
That only applies when we are replying to a message. When I compose a
new message to [EMAIL PROTECTED], I want it to match the ~l
send-hooks that I've set up. The only way for mutt to know that that's
a mailing list address is for me to tell it.
> Though it still requires issuing "subscribe" commands as mutt cannot
> distinct between subscribed/unsubscribed lists easily (looking for
> Delivered-To: or the like would get too messy to work in general I
> guess).
Agreed, though I think we can make a stronger statement: there's
absolutely no way for mutt to ascertain the current state of a
person's list subscription status based on the message alone. For
example, I could unsubscribe to a mailing list, and then decide to
reply to a few of the last messages that I received. Those messages
were sent (and received) while I was subscribed, but nothing in them
will inform mutt that I *just* unsubscribed.
And, of course, when composing a new message, there is no prior
message for mutt to consult to even attempt to detect my subscription
status.
~Kyle
- --
Idealism increases in direct proportion to one's distance from the
problem.
-- John Galsworthy
-----BEGIN PGP SIGNATURE-----
Comment: Thank you for using encryption!
iEYEARECAAYFAki0KeYACgkQBkIOoMqOI15OAQCfQwVzqkBx/al5kMDETsx+2GyN
wUEAoMuHOo8sNWy3J5snNaGxUvgvb+8J
=2guy
-----END PGP SIGNATURE-----
