On 03/11 02:00 , Dave Sherohman wrote: > On Wed, Mar 10, 2010 at 04:14:17PM -0600, Carl Wilhelm Soderstrom wrote: > > On 03/10 04:02 , Mike Miller wrote: > > > The other messages had only the list address in the Reply-To field, but > > > this message (that I am now replying to) had both the sender address and > > > the list address in the Reply-To field. I guess if the sender sets > > > Reply-To himself, then the list appends to that. > > > > Thanks for noticing that Mike. > > So if you want to have replies go directly to you; put your address in the > > Reply-To: header. Mutt will respond to both; not sure of the behavior of > > other MUAs. > > Carl, did you deliberately set it up that way or is that the default > behavior of the list management software? The doco for mailman seemed to indicate that when the reply-to header was set to point to the mailing list, that it would strip other reply-to headers. That said, there is also a setting available that will strip preexisting reply-to headers. > Either way, that doesn't seem right to me. Normally, lists which set > Reply-To will only add it if it is not present and leave it alone if the > user has already inserted their own Reply-To header. The problem with this is that the behavior is inconsistent from message to message. When replying to some people's messages, the 'reply' function will go to the list; and with other messages the same keystrokes will reply to the sender. This violates the principle of least surprise. I personally use mutt and when I'm composing a message I don't have the 'To:' field visible. Combine that with how fast and easy it is to use mutt; and it's really easy to pay absolutely no attention to where the message is going. I *expect* the message to return from whence it came -- and if it came from a mailing list (which is a proxy for the original sender) then it should return along the same path. Really, how many people send you messages with a different Reply-To: header? It's valid; but unusual in practice and would tend to make me suspicious. I'm reasonably happy with the list behavior as it is. Nothing will satisfy everyone tho. -- Carl Soderstrom Systems Administrator Real-Time Enterprises www.real-time.com