On Fri November 19 2004 15:31, Keith Moore wrote:
true. but I suspect it doesn't even fit _your_ definition of the right
thing.
for instance, if the same message is sent to both ietf-822 and ietf-smtp,
do you really want them both filed into the same mailbox?
Given that there is a single (logical) message, there are several
possibilities that one might consider:
A. Whether or not to suppress received duplicates
1. before filing
2. per-"folder", after filing
B. Whether or not to make copies (or logical links, saving
storage space) available via multiple "folders".
C. A hierarchy of filing for such cases.
D. Whether to program, test, debug, repeat for a script to
handle the various possibilities, or to handle simple cases
and deal with unusual situations by hand.
I use C and handle unusual cases manually. If better tools
were available for testing and managing sieve scripts, I
might entertain adding complexity to the script to handle
unusual conditions such as you describe. Given the state
of such tools, I (intentionally) have not done so [indeed,
until relatively recently, filing was handled by MUA (IMAP
client) filters which moved messages from the catch-all
INBOX to specific folders; I changed to a server-based
Sieve script to avoid dependence on a particular UA (or
having to replicate the filters in all UAs), but only with
some trepidation and after an unsuccessful search for
sieve script validation/testing/management tools].