Loop Problem

From
Date
Priority
Reply-To
Once again, let me apologize for the loop problem PaleoNet suffered over the 
weekend.  The loop was caused by a subscriber's full mailbox bouncing 
incoming messages back to the list.  It was further compounded by the fact that 
her local system was bouncing 2 messages back to PaleoNet for every 
message it received from PaleoNet.  Early in the loop's cycle James Maffrey and 
Una Smith both attempted to contact the subscriber's local postmaster and 
asked him/her to address the problem. However, the local postmaster failed to 
take any action.  Perhaps he/she was away for the weekend.  On our end, Sean 
Steppie (NHM Network Manager, and one of PaleoNet's unsung heros) noticed 
the problem on Saturday afternoon and had stopped the bounces from going 
out to everyone by 6:00PM Saturday evening.  Although deleting 400 unwanted 
messages isn't any fun (I know because, as paleonet-owner, I was greeted by 
5000 unwanted messages this morning), if Sean had not caught the problem on 
Saturday it could have been much worse.  He deserves a lot of credit for this 
and many other PaleoNet-related programming projects he's done.

I'm sure that none of us want to go through this again so, at least for the time 
being, I've asked Sean to change PaleoNet's default reply to the message 
originator's e-mail address rather than PaleoNet's address.  This loop was 
generated because the subscriber's server was sending the bounces back 
from where (it though) they were coming from - paleonet.  Now, the worst that 
can happen is that a few rogue bounces might appear in your mailbox from time 
to time (standard bounces will, as before, be routed to me).  The cost of this 
extra protection, however, is that now in order to post a reply to an item to the 
entire list you will have to type PaleoNet's address into the messages "To:" line. 
Many e-mail programs (e.g., Eudora, NCS Mail) possses an automatic address 
book to make this re-routing easy so this shouldn't be too much of a problem for 
most of us.  Until we can work something more elegant out, I think this is a 
workable solution.  As Jim Mahaffrey noted, when you use computers to 
automate events, sometimes interesting, though unanticipated, things happen.

Once again, PaleoNet apologises for the unconvenience and assures it's 
subscribership that we will continue to take steps to make sure that this loop 
problem doesn't reoccur.


Norm MacLeod