Re: [EXIM] Detailed Q on Exim's delivery method

Página superior
Eliminar este mensaje
Responder a este mensaje
Autor: Philip Hazel
Fecha:  
A: Lazlo Nibble
Cc: exim-users
Temas nuevos: [EXIM] eximon
Asunto: Re: [EXIM] Detailed Q on Exim's delivery method
On Thu, 19 Mar 1998, Lazlo Nibble wrote:

> The site where my majordomo mailing lists live runs exim as its MTA. From
> time to time, people who join a mailing list will suddenly get a collection of
> messages that were posted to the list shortly before they joined. This is
> *not* a majordomo feature -- in fact it's come up before on majordomo-users as
> a bug, and was explained away there as a problem with smail, to wit:
>
>   * Message A is posted to the list, delivered to most users, remains in
>     queue for further retrys to a handful of addresses.

>
>   * Subscriber Z joins the list, gets intro messages, starts to get new
>     messages as they're posted to the list.

>
>   * smail walks the queue to look for pending deliveries, finds message A, and
>     *checks the list subscriber file for addresses it has not yet delivered
>     that message to* (i.e., it keeps its own "successful deliveries" state but
>     relies on the list subscriber file (an :include: alias under sendmail,
>     dunno how it's handled under smail or exim) for the "intended recipients"
>     list).

>
>   * smail sees subscriber Z on the subscriber list but not on the
>     "successfully delivered" list, so it sends Z a copy of message A
>     even though Z wasn't actually on the list with A was posted.

>
> I don't know if that's really how smail works, but that's the story I heard.
>
> THE QUESTION: is it possible for this to happen under exim as currently
> (or recently) designed?


Yes. Exim works in exactly the same way if you let it do the list
expansion. There are benefits. If an incorrect address in the list (or,
more usually, a user's .forward file) gets corrected, Exim doesn't keep
on trying the old, bad version.

Is this a serious problem rather than a quirky oddity? It would be a
fair bit of work to provide an option to do otherwise.

-- 
Philip Hazel                   University Computing Service,
ph10@???             New Museums Site, Cambridge CB2 3QG,
P.Hazel@???          England.  Phone: +44 1223 334714



--
*** Exim information can be found at http://www.exim.org/ ***