headers
Valdis.Kletnieks | 18 Jul 2000 20:32
X-Face
Picon
Favicon

Re: mail-abuse.org tests, and weird addresses...

(Am cc'ing the IETF-SMTP list, hopefully it's alive and somebody there
will have guidance on this... For the IETF-SMTP list, the discussion is
what to properly do if handed this during an SMTP transaction:

MAIL FROM:<user <at> domain.name <at> other.domain>

Sendmail currently issues a '250 OK'.

On Tue, 18 Jul 2000 10:46:13 PDT, Kyle Jones <kyle_jones <at> wonderworks.com>  said:
> The attitude I took when I ran a relaying server was that stuff
> like a <at> b <at> c was better off bounced immediately. 

I can deal with that attitude.

However, RFC821 also lists the following codes on page 56:

         550 Requested action not taken: mailbox unavailable
            [E.g., mailbox not found, no access]
         551 User not local; please try <forward-path>
         552 Requested mail action aborted: exceeded storage allocation
         553 Requested action not taken: mailbox name not allowed
            [E.g., mailbox syntax incorrect]
         554 Transaction failed

But then continues to say:

            MAIL
               S: 250
               F: 552, 451, 452
               E: 500, 501, 421
(Continue reading)

Permalink | Reply | 9 comments |
headers
Gregory J. Woodhouse | 18 Jul 2000 21:29

Re: mail-abuse.org tests, and weird addresses...

In this case, I think the correct response would be 501, since it is a
syntax error in the argument to MAIL. 

On Tue, 18 Jul 2000 Valdis.Kletnieks <at> vt.edu wrote:

> (Am cc'ing the IETF-SMTP list, hopefully it's alive and somebody there
> will have guidance on this... For the IETF-SMTP list, the discussion is
> what to properly do if handed this during an SMTP transaction:
> 
> MAIL FROM:<user <at> domain.name <at> other.domain>
> 
> Sendmail currently issues a '250 OK'.
> 
> On Tue, 18 Jul 2000 10:46:13 PDT, Kyle Jones <kyle_jones <at> wonderworks.com>  said:
> > The attitude I took when I ran a relaying server was that stuff
> > like a <at> b <at> c was better off bounced immediately. 
> 
> I can deal with that attitude.
> 
> However, RFC821 also lists the following codes on page 56:
> 
>          550 Requested action not taken: mailbox unavailable
>             [E.g., mailbox not found, no access]
>          551 User not local; please try <forward-path>
>          552 Requested mail action aborted: exceeded storage allocation
>          553 Requested action not taken: mailbox name not allowed
>             [E.g., mailbox syntax incorrect]
>          554 Transaction failed
> 
> But then continues to say:
(Continue reading)

Permalink | Reply | 0 comments |
headers
Tony Hansen | 18 Jul 2000 22:34
Picon
Favicon

Re: mail-abuse.org tests, and weird addresses...

FWIW, we respond with a 501 code.

	Tony Hansen
	tony <at> att.com

Valdis.Kletnieks <at> vt.edu wrote:
> 
> (Am cc'ing the IETF-SMTP list, hopefully it's alive and somebody there
> will have guidance on this... For the IETF-SMTP list, the discussion is
> what to properly do if handed this during an SMTP transaction:
> 
> MAIL FROM:<user <at> domain.name <at> other.domain>
> 
> Sendmail currently issues a '250 OK'.
> 
> On Tue, 18 Jul 2000 10:46:13 PDT, Kyle Jones <kyle_jones <at> wonderworks.com>  said:
> > The attitude I took when I ran a relaying server was that stuff
> > like a <at> b <at> c was better off bounced immediately.
> 
> I can deal with that attitude.
> 
> However, RFC821 also lists the following codes on page 56:
> 
>          550 Requested action not taken: mailbox unavailable
>             [E.g., mailbox not found, no access]
>          551 User not local; please try <forward-path>
>          552 Requested mail action aborted: exceeded storage allocation
>          553 Requested action not taken: mailbox name not allowed
>             [E.g., mailbox syntax incorrect]
>          554 Transaction failed
(Continue reading)

Permalink | Reply | 0 comments |
headers
John C Klensin | 19 Jul 2000 06:07

Re: mail-abuse.org tests, and weird addresses...


--On Tuesday, July 18, 2000 2:32 PM -0400 Valdis.Kletnieks <at> vt.edu
wrote:

> (Am cc'ing the IETF-SMTP list, hopefully it's alive and
> somebody there will have guidance on this... For the IETF-SMTP
> list, the discussion is what to properly do if handed this
> during an SMTP transaction:
> 
> MAIL FROM:<user <at> domain.name <at> other.domain>
> 
> Sendmail currently issues a '250 OK'.
>...   [more of original message  below]...

Hi.

If my memory is correct, 501 (syntax error) is intended, as was
the exclusion of 553 from the MAIL FROM responses.  I agree that
the 1123 text is a little muddy (and it is probably at least
partially my fault).  Please take a look at
draft-ietf-drums-smtpupd-12.txt and see if you think it is clear
and good enough; if not, comments should probably go to the DRUMS
list (copied on this response).

I would think that 553 would be used for a situation in which,
e.g., the left-hand-side (local-part) of an address was valid
syntax in 821 (etc) terms, but invalid for processing/ delivery
on the target host while 501 would be used for things that were
lexically unacceptable given the 821 grammar.  If that were the
theory, then a receiving SMTP wouldn't be able to comment on the
(Continue reading)

Permalink | Reply | 1 comment |
headers
Doug Royer | 19 Jul 2000 19:07

Re: mail-abuse.org tests, and weird addresses...

Valdis.Kletnieks <at> vt.edu wrote:
> 
> (Am cc'ing the IETF-SMTP list, hopefully it's alive and somebody there
> will have guidance on this... For the IETF-SMTP list, the discussion is
> what to properly do if handed this during an SMTP transaction:
> 
> MAIL FROM:<user <at> domain.name <at> other.domain>
> 
> Sendmail currently issues a '250 OK'.

I think it is a valid address. It was manually routed from one
domain to another. I used that feature a few years ago to
stop spam as a result of people scanning net-news postings.
(I posted with x <at> y <at> z). Any email that came to that address
at domain 'y' would get extra filters before relaying to domain 'z'.

It is another way of doing mail relaying.

I did an exhaustive search a few years ago. The ability to
route with multiple ' <at> ' was in an RFC, so it is valid.
(If it has not been superseded).

Set mail relay to off, if you can. (I don't know the exact keyword).
It should stop working.

-Doug
Permalink | Reply | 4 comments |
headers
Keith Moore | 19 Jul 2000 21:22
Picon

Re: mail-abuse.org tests, and weird addresses...

> > > MAIL FROM:<user <at> domain.name <at> other.domain>
> > 
> > Sendmail currently issues a '250 OK'.
> 
> I think it is a valid address.

it is invalid syntax in both RFC821 or RFC822.

Keith
Permalink | Reply | 3 comments |
headers
Gregory J. Woodhouse | 19 Jul 2000 22:32

Re: mail-abuse.org tests, and weird addresses...

It's been a long time, but I once had to use the so-called "%-hack" in my
e-mail address, but I believe you are correct that user <at> host1 <at> host2 is
incorrect syntax. I'm curious, does sendmail treat this as equivalent to
user%host1 <at> host2?

On Wed, 19 Jul 2000, Keith Moore wrote:

> > > > MAIL FROM:<user <at> domain.name <at> other.domain>
> > > 
> > > Sendmail currently issues a '250 OK'.
> > 
> > I think it is a valid address.
> 
> it is invalid syntax in both RFC821 or RFC822.
> 
> Keith
> 

---
Gregory Woodhouse
gjw <at> wnetc.com    /    http://www.wnetc.com/home.html
"An atheist staring from his attic window is often nearer to God than the
believer caught up in his own false image of God."
--Martin Buber
Permalink | Reply | 1 comment |
headers
Gregory Neil Shapiro | 19 Jul 2000 22:37
Picon

Re: mail-abuse.org tests, and weird addresses...

gjw> It's been a long time, but I once had to use the so-called "%-hack" in my
gjw> e-mail address, but I believe you are correct that user <at> host1 <at> host2 is
gjw> incorrect syntax. I'm curious, does sendmail treat this as equivalent to
gjw> user%host1 <at> host2?

Yes.
Permalink | Reply | 0 comments |
headers
Valdis.Kletnieks | 19 Jul 2000 23:33
X-Face
Picon
Favicon

Re: mail-abuse.org tests, and weird addresses...

On Wed, 19 Jul 2000 00:07:40 EDT, John C Klensin said:
> If my memory is correct, 501 (syntax error) is intended, as was
> the exclusion of 553 from the MAIL FROM responses.  I agree that
> the 1123 text is a little muddy (and it is probably at least
> partially my fault).  Please take a look at
> draft-ietf-drums-smtpupd-12.txt and see if you think it is clear
> and good enough; if not, comments should probably go to the DRUMS
> list (copied on this response).

Yes, 501 *is* the correct error, and your example of an embedded blank
getting a 553 was better.

smtpupd section 4.3.2 specifically allows any/all of 501, 550, and 553
on the MAIL FROM: which gives future releases of Sendmail sufficient room
to work.  That text looks good and reasonable.

Is there a target date for smtpupd to move out of Draft status and into
RFC, where it will be more cite-able without being a moving target?

--

-- 
				Valdis Kletnieks
				Operating Systems Analyst
				Virginia Tech
Permalink | Reply | 0 comments |
headers
Soegi Hartono | 19 Jul 2000 23:12

Re: mail-abuse.org tests, and weird addresses...

I've tried to send e-mail like this syntax user%host1 <at> host2, it successfully
sent but it's not getting to the final destination, anyone know?

-----Original Message-----
From: Gregory J. Woodhouse <gjw <at> wnetc.com>
To: Keith Moore <moore <at> cs.utk.edu>
Cc: Doug Royer <Doug.Royer <at> software.com>; Valdis.Kletnieks <at> vt.edu
<Valdis.Kletnieks <at> vt.edu>; Kyle Jones <kyle_jones <at> wonderworks.com>;
sendmail-beta <at> Sendmail.ORG <sendmail-beta <at> Sendmail.ORG>; ietf-smtp <at> imc.org
<ietf-smtp <at> imc.org>
Date: Wednesday, July 19, 2000 1:40 PM
Subject: Re: mail-abuse.org tests, and weird addresses...

It's been a long time, but I once had to use the so-called "%-hack" in my
e-mail address, but I believe you are correct that user <at> host1 <at> host2 is
incorrect syntax. I'm curious, does sendmail treat this as equivalent to
user%host1 <at> host2?

On Wed, 19 Jul 2000, Keith Moore wrote:

> > > > MAIL FROM:<user <at> domain.name <at> other.domain>
> > >
> > > Sendmail currently issues a '250 OK'.
> >
> > I think it is a valid address.
>
> it is invalid syntax in both RFC821 or RFC822.
>
> Keith
>
(Continue reading)

Permalink | Reply | 0 comments |

Gmane