dovecot - Jan 2007 - Sieve and LDA

Hello !

I  have started  to use  Sieve support  in LDA.  From the  RFC  a same
message should not be delivered twice in the same folder :

,----
| 2.10.3. Message Uniqueness in a Mailbox
| 
| 
|    Implementations SHOULD NOT deliver a message to the same folder more
|    than once, even if a script explicitly asks for a message to be
|    written to a mailbox twice.
| 
|    The test for equality of two messages is implementation-defined.
| 
|    If a script asks for a message to be written to a mailbox twice, it
|    MUST NOT be treated as an error.
`----

Dovecot seems to ignore this "SHOULD  NOT" or am I missing something ?
At first,  I thought  they were  speaking of case  where there  is two
idententic "fileinto" directive in the ruleset. Something like :
 fileinto "box";
 fileinto "box";

But, I don't think that they  would they "the test for equality of two
messages" with such a trivial case.

Maybe dovecot  should maintain  a list of  recent message-id  seen for
each user (with the associated mailbox).
-- 
I WILL ONLY DO THIS ONCE A YEAR
I WILL ONLY DO THIS ONCE A YEAR
I WILL ONLY DO THIS ONCE A YEAR
-+- Bart Simpson on chalkboard in episode 3F31

On Mon, Jan 29, 2007 at 09:26:48PM +0100, Vincent Bernat
wrote:
> Hello !
> 
> I  have started  to use  Sieve support  in LDA.  From the  RFC  a same
> message should not be delivered twice in the same folder :
> 
> ,----
> | 2.10.3. Message Uniqueness in a Mailbox
> | 
> | 
> |    Implementations SHOULD NOT deliver a message to the same folder more
> |    than once, even if a script explicitly asks for a message to be
> |    written to a mailbox twice.
> | 
> |    The test for equality of two messages is implementation-defined.
> | 
> |    If a script asks for a message to be written to a mailbox twice, it
> |    MUST NOT be treated as an error.
> `----
> 
> Dovecot seems to ignore this "SHOULD  NOT" or am I missing
something ?
> At first,  I thought  they were  speaking of case  where there  is two
> idententic "fileinto" directive in the ruleset. Something like :
>  fileinto "box";
>  fileinto "box";
> 
> But, I don't think that they  would they "the test for equality of
two
> messages" with such a trivial case.
> 
> Maybe dovecot  should maintain  a list of  recent message-id  seen for
> each user (with the associated mailbox).
I have interpreted this to mean that within a single delivery (e.g. a
single invocation of a script for a particular message), the implementation
should take care to de-duplicate requests to file into folders, so that
attempts to file multiple times to a folder within that single script
execution should be thwarted.  Your last sentence makes it seem that you
are thinking that the prohibition spans multiple script invocations
(e.g. for different deliveries), but I don't believe that's the intent.

The last line you quoted:

   | If a script asks for a message to be written to a mailbox twice, it
   | MUST NOT be treated as an error.

makes that pretty clear to me, that it is about one single script
execution.

Note that I have no idea if dovecot deliver does that or not, I'm
just reacting to the implication in your last line.

mm