On Wed, 28 Jun 95 16:37:29 TZ Benjamin Holz <t-benho@microsoft.com> said:
>I figured this would be as easy as placing the address of my local list
>on the external list however this would result in the same message being
>constantly sent back and forth. How is the accomplished? Do you need to
>count on the mailing list software to know about circular references?
It depends on which software you are using but in general you do need to
do something to prevent a loop and avoid unwanted error messages. For
instance, if you use LISTSERV and you don't do anything, you won't get a
loop but for each posting you'll get an error message saying this is a
duplicate of a previous message and maybe something is wrong. To make
things work more seriously you would need to "peer" the lists. Then
LISTSERV knows that the two lists are really the same thing and it can
forward subscriptions to the closest available peer and do the same with
commands like REVIEW, DELETE, etc.
The real issue though is why do you need to do this? Peering was used
heavily in 1986-88 to do load balancing. Computers were slower and the
US-Europe backbone was a mesh of two 9.6k lines, of which one was
statistically always down at any given time :-( It could easily take a
day for mail to cross the ocean. Then DISTRIBUTE was introduced to
balance mail delivery whether you have a peer or not, and the network got
more reliable so you didn't have to worry about being unable to access
the archives over the weekend because both 9.6k lines were down and the
people who could fix them were taking the kids to the park. Anyway,
nowadays most people just run one big list. The lists that you see with
many peers were mostly created back when things were done this way, and
people just left it as is.
Eric
|
|