Re: Reverse dns checking for local machine

Page principale
Supprimer ce message
Répondre à ce message
Auteur: Greg A. Woods
Date:  
À: Piete Brooks
CC: rich, exim-users
Sujet: Re: Reverse dns checking for local machine
[ On Wed, September 3, 1997 at 11:27:04 (+0100), Piete Brooks wrote: ]
> Subject: Re: Reverse dns checking for local machine
>
> > Even more important than diagnosis though is direct end-to-end delivery
> > of mail without intermediaries wherever possible for minimal latency and
> > greatest security and reliability of delivery.
>
> as well as improving control (it sits in *your* queue, so you can process
> it as you want), saves bandwidth (e.g. if the final MTA rejects the message
> before the DATA stage -- particularly for "message too large"), and ensures
> that if the message is rejected, the user gets the whole message (rather than
> a truncated error report), allowing retransmission, etc.


Come on guys! I wasn't talking about anyone who has half a clue. I'm
talking only of dial-up IP users who have dynamic IP# assignments and
who don't even know how to spell SMTP or VRFY, never mind know what a
queue is!!! 99.9% of the common users have no business making direct
SMTP connections to anyone and they should not be permitted to do so.

I work with a number of large ISPs in the Toronto area and have seen
every kind of abuse possible. A very small percentage of dial-up IP
users will purposefully abuse their connection. The rest won't even
realize the firewall is in their way and wouldn't even know how to
detect it if they were told about it. They run PCs with Netscape,
Eudora, or Microsoft IE and may not even have a telnet client installed,
never mind any sophisticated debugging tools like the vrfy(1) tool.

There are, of course, the "want-to-be" folks who might run Linux,
FreeBSD, or even a commercial unix on a PC at home and who might try to
run something like Exim as their mailer on a dial-up IP link. However I
think we've all agreed long ago that Exim and SMTP are not suitable for
such users and that they either need a full-time connection or to forget
about running their own mailer.

I'm also not talking about any kind of censorship -- the only intent is
to enforce accountability and responsibility.

Talking about latency, security, reliability, and private queue control
as advantages to direct SMTP delivery from dial-up clients is
meaningless. Dial-up clients are inherently high-latency, low-security,
low-reliability, and queue-less (and clue-less!). They are usually tied
directly to the user interface and sport only minimal half-baked
implementations.

Remember way back when some folks considered the use of SMTP for MUA
client mail delivery to be nearly blasphemous? I do. These arguments
were all raised before in the other direction and the proponents of
using SMTP for MUA client delivery all said not to worry -- they'd never
do direct SMTP delivery to the end recipient.... How quickly things turn.

-- 
                            Greg A. Woods


+1 416 443-1734      VE3TCP      <gwoods@???>      <robohack!woods>
Planix, Inc. <woods@???>; Secrets of the Weird <woods@???>


--
* This is sent by the exim-users mailing list.  To unsubscribe send a
    mail with subject "unsubscribe" to exim-users-request@???
* Exim information can be found at http://www.exim.org/