more on anti-*AOL plan* coalition

[David Farber <dave () farber net>]

Begin forwarded message:

From: Dave Crocker <dcrocker () bbiw net>
Date: March 26, 2006 2:22:54 PM EST
To: David Farber <dave () farber net>
Subject: Re: Fwd: [IP] anti-*AOL plan* coalition

Dave,

I appreciate Cindy's response.  However pursuing any sort of  
constructive
dialogue is going to require that we be considerably more careful  
about facts
and fears, and especially be careful to distinguish them.


> First, the rejection.  Dave asserts that it's improper for me,
> personally, and apparently EFF, to voice any objection to any
> proposal to try to respond to spam or phishing unless we also come up
> with a complete solution to spam and phishing.

1. I referred to Cindy, not EFF.

2. I'll guess that Cindy's perception of my "requirement" is in  
response to:
> If it is so easy, Cindy, why haven't you promoted one and gotten it  
> used?

This was not about making criticisms, but about making overly-facile  
assertions.
 Cindy claims something is easy when it has been demonstrated not to  
be as
evidenced by its absence, after some years of effort and very strong  
market need.


>     I have no problem with criticism of what I say; I object  
> strongly to a suggestion that I have no right to comment in the  
> first instance.

So do I.  Glad we agree.

My only question is who made such an inappropriate suggestion,  
because it was
most certainly not me.


> many overlays, there should be lots of inputs. This is, in fact, one
> of my strongest objections to what AOL is doing here, and part of
> what I meant when I said that there are "plenty of ways to do
> 'certified' or 'digitally signed' email without having ISPs choose
> winners and charge per message."

1. AOL already has multiple vetting mechanisms.  So the agreement was  
merely
adding one more.

2. OK.  So the transgression was AOL's failing to make multiple  
additional
agreements at the same time?  Unfortunately, organizations have limited
resources, and adding a single new mechanism is extremely difficult  
and costly.
 No competent operations manager would agree to the addition of  
multiple
services at once, for (legitimate) fear of hurting system  
reliability. I wonder
how AOL stockholders would feel about AOL taking on that much  
operational risk
to their email service?


> AOL is here choosing one input and one proprietary "trust overlay"
> system -- Goodmail's -- and taking a financial cut of the money
> earned from that overlay and input.  My argument is that by doing so,
> AOL is undermining the markets for both overlays and inputs, not
> fostering it. AOL is also reducing their own incentives to rely on
> other inputs, including their own whitelist and enhanced whitelist.

Until the AOL and Yahoo announcements, the email reputation market  
has, at best,
shown modest acceptance and growth. By some measures, it could  
reasonably have
been characterized as "languishing". Since then, it has gained center  
stage.
This does not sound like undermining a market. It sounds like  
kickstarting it.

The question of incentives is always a good one to ask.  Again,  
however, the
topic is rather more difficult and subtle than seems to be getting  
acknowledged.

As a small example, I'll note that AOL's deriving revenue from a  
third-party
trust service might be expected to encourage AOL to conform to the  
quality
assurance requirements of that trust mechanism.  This is a Good Thing.

So, it is important to look for ways an agreement can misalign  
incentives, but
it is equally important to look for the ways it can align them.


> I believe that this will result in real pressure on senders to
> join Goodmail's system or face not getting their messages through.

They already face not getting their messages through.

Use of a reputation service provides a new option for improved delivery.


> And if other ISPs follow suit, legitimate senders may be faced with  
> having to
> navigate (and pay for) several different systems, further increasing
> their costs and hurting their ability to speak to a willing audience.

This appears to be a concern about the impact of having multiple  
reputation
services.  Yet earlier the concern was for having only one.

I'm now confused about what is being required.

Certainly there is a need for standardizing the various interaction  
mechanisms
used by different services.  However the bricks-and-mortar world  
flourishes with
multiple reputation services within a sector, so we should expect  
that same
benefit in the Internet.

That said, there is always a concern that competing services will create
confusion or other problems.  I hope, however, no one is suggesting  
some sort of
legislated, single assessment service for any part of the community.  
If they
are, they should cite some real-world history of such an approach  
being viable.


> My concern is based in part on the observation that
> AOL will reasonably seek to maximize its profits and minimize it's
> costs. Pushing senders into Goodmail does both and that's why I'm
> skeptical of AOL's claims that nothing will change for senders who
> don't choose Goodmail.

I don't recall seeing AOL claim that nothing would change for  
senders.  I'd like
to review their statement, if someone could point me to it.

Given the nature of the announcement that AOL made, I would certainly  
*hope*
that quite a bit will change for senders.

Better delivery.  Fewer complaints.  Big changes.


> My concern is also based on my very real
> experience with Bonded Sender. Here's a similar report from Bennett
> Hazelton about his Peacefire mailing list:

Unfortunately, Mr. Hazelton's article contains serious factual  
errors.  As a
small example:  Hotmail never had a whitelisting service, although it  
did have a
problem-fixing mechanism.

So the Hotmail/Bonded-Sender example is quite different from the  
current situation.


> My main concern here is about the risk that AOL
> will undermine the market for some of the very things that Dave
> suggests by choosing one proprietary system and pushing senders  
> into it on pain of not having their messages delivered to AOL  
> subscribers. I hope
> this point isn't lost in the sturm and drang.

We agree that this is an important concern.

The counter-question, to any expression of certainty that things will  
play out
in this unfortunate way, is why anyone believes that AOL's  
subscribers would
tolerate that outcome, since it means that they would not be getting  
email they
wanted.

d/

--

Dave Crocker
Brandenburg InternetWorking
<http://bbiw.net>





-------------------------------------
You are subscribed as lists-ip () insecure org
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/