Maybe -- I'm just asking for clarification on how this works. I guess
I should have put it into two parts:
How does Sieve reference the canonical message when it suppresses
duplicates? How well does this work?
and
When someone copies a message from one folder to another using IMAP
commands, does this create another reference or does it copy the
entire message?
Thanks!
Jen
Lawrence Greenfield wrote:
>
> I think you're confusing duplicate delivery suppression with "single
> instance store".
>
> Larry
>
> Date: Tue, 05 Jun 2001 10:58:03 -0700
> From: Jen-Mei Wu <[EMAIL PROTECTED]>
> CC: [EMAIL PROTECTED]
>
> Thanks for the suggestion. I read all the files and duplicate
> delivery suppression is barely mentioned. A lot of bugs relating to
> it were apparentlyl fixed, though. ;-)
>
> If there are other sources describing this feature, I'd love to hear
> about them. If anyone has experience with the delivery suppression
> feature, I'd love to hear about that, too (including how well it
> works).
>
> My understanding of it is that Sieve identifies duplicates (using
> message-id) and somehow makes it so there's only one reference (I
> assume by having database entries reference the canonical copy of the
> message). However, does that mean that if people copied messages from
> one folder to another (or one mailbox to another) that duplicate
> suppression doesn't happen? Of course, by that time, messages are all
> stored in a database, so maybe just the key is copied?
>
> Thanks!
>
> Jen
>
> Amos Gouaux wrote:
> >
> > >>>>> On Wed, 30 May 2001 11:11:41 -0400,
> > >>>>> Jen Wu <[EMAIL PROTECTED]> (jw) writes:
> >
> > jw> I read in "Managing IMAP" that Cyrus can be configured to keep
> > jw> only one copy of a message that would otherwise be duplicated
> > jw> many times. For example, if 10 people are on the same mailing
> > jw> list, only one copy of each mailing list message would be kept.
> >
> > jw> How does this work? And how well does it work? The book said
> > jw> that Sieve was involved, but I'm not sure why that would be the
> > jw> case. For example, if someone copies a message from one
> > jw> folder to another, that would result in a duplicate message, too,
> > jw> and Sieve wouldn't be able to handle that, would it?
> >
> > you might check in the doc directory:
> >
> > $ grep -i 'duplicate delivery' *
> > changes.html: <li> The duplicate delivery database expiration (deliver -E)
>was
> > changes.html: <li> Fixed bugs in deliver relating to duplicate delivery
> > changes.html: <li> deliver now will not use Sieve if duplicate delivery
> > changes.html: duplicate delivery database to be checked even if dupelim was
> > changes.html: <li> Duplicate delivery changes:
> > changes.html: Split out duplicate delivery elimination to multiple
> > changes.html: directory, nothing bad will happen (other than duplicate
>delivery
> > changes.html:<li>Redo duplicate delivery elimination in <tt>deliver</tt>.
> > install-perf.html:duplicate delivery suppression, each time a mail message is
>delivered
> > install-upgrade.html:<li>Create some extra directories and remove the duplicate
>delivery
> > install-upgrade.html:<p>Duplicate delivery suppression is now required for
>Sieve.
> > overview.html:<li><a href="#duplicate">Duplicate Delivery Suppression</a>
> > overview.html:<LI><TT>INFO</TT> - Mailbox openings, duplicate delivery
>suppression
> > overview.html:<h3>Duplicate Delivery Suppression</h3><a name="duplicate"></a>
> >
> > --
> > Amos
