IP: RE: Let you decide where the truth is -- AOL and Harvard

[David Farber <dave () farber net>]

Boy people are leery about saying things with their name attached djf

> Date: Mon, 7 Jan 2002 13:16:40 -0800
> From:
> To: farber () cis upenn edu
> Subject: Re: IP: RE: Let you decide where the truth is -- AOL and Harvard
> User-Agent: Mutt/1.2.5i
>
> also for publication, but not for attribution:
>
> the previous sender is correct that aol SILENTLY DISPOSES of incoming
> email which it believes is spam, without even generating a bounce
> message to the sender!
>
> but it's even worse than the situation he described, in which aol
> attempted to distinguish a spammer from a large opt-in mailer.
>
> it also does this, particularly, for machines which it believes are
> in the dialup or dynamically assigned address pools of various isps,
> and who send mail directly to aol, rather than relaying through their
> isp's mail relays.  (many spammers fall in this category.)
>
>         obviously dynamic addresses cannot be put on a whitelist.
>
> but this is true also for small business customers who have inverse
> dns map host names that *resemble* dynamically assigned addresses,
> even if they happen to be statically assigned!
>
>         aol seems unwilling or unequipped to put ranges or lists of such
>         addresses on a whitelist, so anyone who uses NAT with a range of
>         outgoing addresses is out of luck.
>
> victims include many smtp-capable senders (a small business, in this
> case) who use dsl or cable modems.
>
> aside from silently s2tcanning mail, there is no way for you, the sender,
> to tell that your machine is on their blacklist!  your recipients at aol
> simply don't receive your mail, despite its acceptance without error
> at their incoming mx.
>
> On Mon, Jan 07, 2002 at 02:44:19PM -0500, David Farber wrote:
> >
> > >From:
> > >To: "David Farber" <dave () farber net>
> > >
> > >FOR publication, not FOR attribution.  Thanks.
> > >
> > >Dave,
> > >
> > >I work for a company that delivers large amounts of e-mail for 
> Fortune-1000
> > >companies.
> > >
> > >We had a very similar problem with AOL: we started to see many of our
> > >messages fail to arrive in AOL mailboxes even though AOL reported they 
> were
> > >successfully accepted for delivery.
> > >
> > >Sending one or two messages at a time would work flawlessly, but when
> > >sending anything larger, some might arrive while many would not.
> > >
> > >After a lot of testing, we discovered AOL has some very interesting 
> policies
> > >regarding email delivery.
> > >
> > >In a nutshell, if you deliver more than some number messages to AOL within
> > >an a certain timeframe, AOL will accept them for delivery but in 
> actuality,
> > >delete them without notice.  The metrics AOL uses to decide what should or
> > >should not be delivered are not published.
> > >
> > >We solved this when, after several months of spotty delivery, we 
> finally got
> > >to the right person within AOL who informed us they maintain a "white 
> list"
> > >of companies like ours that are allowed to deliver messages "in bulk" 
> to AOL
> > >recipients.
> > >
> > >Once we found out about its existence, a few emails and a faxed affidavit
> > >later, we saw all our messages delivered without incident.
> > >
> > >I would imagine Harvard fell afoul of these policies.
> >
> > For archives see:
> > http://www.interesting-people.org/archives/interesting-people/
>
> --
> mark seiden, mis () seiden com

For archives see:
http://www.interesting-people.org/archives/interesting-people/