Re: More sorbs blacklisting
On 2006-07-10 0958, Craig Sanders wrote:
> On Sun, Jul 09, 2006 at 04:42:24PM -0400, John Kelly wrote:
> > There's so much garbage coming from dsl connections nowadays, that
> > many ISP's block all mail originating from dsl and/or dynamic IPs.
> > It's 98% (or more) spam/virus/garbage anyway, so who cares if they
> > lose 1% or 2% which isn't?
>
> it's not even remotely close to 1 or 2%. not even 1 in a million.
>
> i get really sick and tired of hearing whines from losers who think that
> their email is so important that everyone else MUST accept millions of
> spams from dynamic addresses just on the off-chance that they might one
> day want to send you a mail.
Accepting the pragmatism that spam has forced the world into, and that
it in some of our minds justifies half-measures such as dropping all
smtp-connections from certain netblocks, I'll still post my IMHO.
It's a dilemma. In the one hand we have freedom of expression, in the
other the demands of customers. That is, sacrificing every-Joe's ability
to set up a mail server on his own domain on his high-end DSL line will
drastically bring down the amount of illegitimate mail in later (and
more resource-demanding) filtering and thereby the amount of
false-negatives in costumers' mailboxes.
>From a business standpoint, there's no doubt. You'll simply skim the
statistics, and conclude that outlawing DHCP'ed hosts is optimal -- and
therefore various sysadmins shouldn't be hunted for their pragmatic
choices.
However, ideologically (hold the flames for just a paragraph), it's a
compromise. Some sort of registration (at your ISP or by buying
netblocks yourself) is going to be needed if you want to send mail,
anyway -- and besides discouraging people from setting up their own mail
servers for fun and learning, it'll cause the Internet to become more
centralized around the big players.
That's reason enough for me (albeit, I wouldn't assume that an employer
would be as easily convinced) to continue the non-binary choices of
bayesian and otherwise content-oriented filtering methods, while
something fundamentally fixing the feasibility of spamming is made up.
Of course the routing should count heavily in that estimation.
\end{ideology}
And in a technical note; the meaning of having valid reverse DNS is
that the delivering MTA's HELO/EHLO is equal to the rDNS-query on its
address?
Example on the above: Mail from lists.debian.org is HELO'ed with
murphy.debian.org, here, and the connection comes from 70.103.162.31,
which has rDNS murphy.debian.org. Had that last domain been something
else, 70.103.162.31 wouldn't have been legitimate, in this meaning of
the word?
Had SPF been widely adopted, it'd be theoretically possible to achieve a
positive answer to the i-am-not-a-Windows-moron-on-a-DSL-question there,
right?
If so, could one not merely drop all incoming smtp-connections which
neither originated from rDNS-legitimate addresses or had valid SPF
records...? That'd give Joe a foot in the door.
Regards, skrewz.
Reply to: