Re: [EXIM] bug in exim when talking to non-ESMTP host?

Kezdőlap
Üzenet törlése
Válasz az üzenetre
Szerző: Philip Hazel
Dátum:  
Címzett: Paul Mansfield
CC: Exim users mailing list
Tárgy: Re: [EXIM] bug in exim when talking to non-ESMTP host?
On Thu, 5 Nov 1998, Paul Mansfield wrote:

> The customer who is running "David Postman" (some sort Novell mail gateway,
> http://www.tobit.bit) talked to their tech support at Tobit UK, and they say
> that Exim is WRONG to batch mail down the same socket when the receiver can't
> handle it and quotes the RFC s4.5.1:
>
>   >      4.5.1.  MINIMUM IMPLEMENTATION
>   >          In order to make SMTP workable, the following minimum
>   >          implementation is required for all receivers:
>   >             COMMANDS -- HELO                        MAIL
>   >                         RCPT                        DATA
>   >                         RSET                        NOOP
>   > As you can see from this quote the SEDN, SOML, SAML are not a part of the
>   > minimum requirements.


This is totally off-beam. Nobody said anything about SEND, SOML, or
SAML. What is he/she raving about?

<example proving SEND, SOML, SAML rejected cut>

> They go on to say that their system is thus a minimal RFC821 compliant box in
> that it does handle the minimum... but when Exim tries multiple mails they
> reply with 500, but Exim ignores it:


Rubbish. Exim does not send SEND, SOML, or SAML. These are obsolete
RFC821 stuff which nobody supports. They are concerned with writing to
the user's terminal. Nothing to do with multiple messages.

>   > E:   DATA   ;E-mail contents...
>   > P:   354    ;Start mail input; end with <CRLF>.<CRLF>

> >
> > ... and then the data... (Parameters to the commands are not shown.)


... terminated with <CRLF>.<CRLF> after which the server should be ready
to accept a new "transaction", that is, a new MAIL FROM. *If* they then
give 5xx to the MAIL FROM, the sending Exim will take this as a hard
error, and bounce the message.

However, you said that they got one gigantic message with everything
concatenated. That looks to me as if they are not recognizing
<CRLF>.<CRLF> correctly. Can you tell me their domain? I can then do
some tests by hand to see what happens. Alternatively, what you can do
is queue up a message form them with

exim -odqs postmaster@???
This is a test message.
.

and then send another with debugging turned on to show the SMTP
dialogue:

exim -d postmaster@???
This is another test message.
.

This will show what their responses to the various SMTP commands are.

> > The order of the SEND, SOML and SAML commands could be in any order - or
> not used at all. The
> > main point is
> > that the reply to these three commands will always, under the current
> implementation of
> > Postman, be 500.


Totally irrelevant. What has caused this blether?

> Maybe Tobit's software is really primitive, but if they're technically
> right, does Exim need to have an extremely dumb mode of operation.


They are talking rubbish. However, as I said, you *can* stop Exim
sending multiple messages down the same connection, by setting batch_max
in the smtp transport. But that will apply to all sites, unless you
arrange for a special transport for this one domain.

-- 
Philip Hazel            University of Cambridge Computing Service,
ph10@???      Cambridge, England. Phone: +44 1223 334714.



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