"HAS" == Hans-Albert Schneider <Hans-Albert.Schneider@mchp.siemens.de> writes:
"NK" == Kapoor, Nishikant X <Nishikant.X.Kapoor@xcelenergy.com> writes: NK> All, I am trying to use rc.local.s10.mhonarc from NK> http://www.ha-schneider.de/software/smartlist/ to set up NK> an archive for a mailing list. I have renamed NK> rc.local.s10.mhonarc to rc.local.s30 and added following NK> line to rc.custom
NK> RC_LOCAL_SUBMIT_30 = rc.local.s30 HAS> RC_LOCAL_SUBMIT_xx are included in rc.submit (look for HAS> "INCLUDERC=$RC_LOCAL_SUBMIT_xx"). RC_LOCAL_SUBMIT_30 is HAS> not used, so your setting has no effect. Sorry, I forgot to give you the overall picture: After loading rc.init (global initializations) and rc.custom (per-list customization), $RC_LOCAL_SUBMIT_00 is processed. This means that *all* mails are seen by this file, even misdirected subscribe messages and other stuff you don't want to distribute. (Subscribe and unsubscribe mails should go to list-request@your.dom.ain, but it is a common mistake to send them to list@your.dom.ain.) Then, some sanity checks are done: It may not be too large, may not come from a daemon (that might be a misdirected bounce message), may not look like an administrative request (misdirected un/subscribe etc), and may not bear the lists stamp (to avoid mail loops; see below). If the mail passes all these tests and the sender may post to the list (the sender is on the accept list, or foreign_submit is off), the $RC_LOCAL_SUBMIT_10 gets its hands on the mail. Next come some special purpose handling: - If the list is moderated and the mail has not yet been approved by a moderator, it is forwarded to the moderators for approval. Finished. - If desired, the sender is forcibly subscribed (unless on the reject list). - If this list is digested, do digest processing. Finished. Now the mail gets archived, and the header manipulated (e.g., "Received:" headers are removed (as are some others), list specific headers are added (including the anti-loop stamp mentioned above), and a dummy "Subject:" is added if missing). Now comes $RC_LOCAL_SUBMIT_20, and immediately sfter that the message is distributed. It is therefore important to think about which rc.local.* file is appropriate for which recipes you add. In case of MHonArc archiving the messages, it is mostly a matter of taste whether to call it in rc.local.s10 or rc.local.s20. However, if your list is moderated, you will prefer the latter (after approval); if it is digested, the former seems mandatory, since rc.local.s20 is never reached in this case (if I interpret rc.submit correctly). Hans-Albert -- Hans-Albert Schneider <Hans-Albert.Schneider@infineon.com> Infineon Technologies AG phone: (+49) 89 234 45445 Corporate Logic; Verification fax: (+49) 89 636 42284 (CL DAT DF LD V) Munich, Germany