[Mailman-Users] Goodmail spells doom for mailing lists?

Dave Crocker dhc2 at dcrocker.net
Sat Mar 4 16:36:17 CET 2006

Brad Knowles wrote:
> At 5:53 PM -0800 2006-03-03, Dave Crocker wrote:
>>  AOL creates a specialized, rather expensive process that it provides for
>>  free, to ensure delivery of a class of mail.
>     Unless you mean "all mail" when you say "class of mail", then no -- 
> not really.  So far as I know, they did not have any "certified mail" 
> type of solution that they made available to the paying spammers.

1. The effort it takes to get on the EWL and stay on it is substantial. It 
therefore creates a significant division between "all mail" and "mail from folks 
on the EWL".

2. "paying spammers" is nicely inflammatory, but factually incorrect, according 
to most definitions of spam.  In terms of the actual claimed goal for these 
types of services, in fact, it is completely and totally factually incorrect.

The claim is that the mechanism is intended for mail that receivers really want 
to get.  "Mail I don't want" is the classic form of the most extreme definition 
of spam.

Whether actual operation matches actual goal is always a good question, of 
course.  For that, one needs to look at the criteria to qualify for the program, 
its enforcement at the admission phase, and its enforcement after admission.

>>                                                The operation of this
>>  mechanism is pure overhead for AOL.  Worse, it is distinct to AOL.  To
>>  the extent any other receive-side ISP operates such a service, it is
>>  entirely independent of AOL.  That is, anyone wanting on these special
>>  lists must to special things for each of these lists.
>     Very few other operators in the world have attempted to run an 
> enforced whitelist solution of the sort that AOL has developed.

That is one of my points.

> disagree that this was necessary on AOL's part, but to the extent that 

I don't recall saying it was "necessary" for AOL.

I DID say that the per-ISP EWL model does not scale across ISPs.

>>  So along comes a few companies who are trying to find ways to let
>>  receive-side ISPs outsource the job of assuring that trustable bulk
>>  mail is, in fact, trusted.  (That is, the receiver wants this stuff and
>>  these services are provding ways to assure that they get it.)
>     That's the key point I don't believe.  I don't trust net-nanny 
> companies to properly operate an IP-address based black list, nor do I 
> trust that these kinds of operations will remain secure even if they 
> could be properly operated. 

We all have different biases.  What we choose to trust, in the absence of 
empirical data, is a matter of personal choice.

> <http://www.boingboing.net/2006/03/01/saudi_arabia_joins_l.html> and 
> <http://www.boingboing.net/censorroute.html>.

It is very much in vogue, these days, to assert such choices and claim that they 
really are based on fact. However any reasonable analysis of the basis for the 
choice turns out to have nothing to do with legitimate empirical data that is 
directly relevant to the conclusion.

More often that not a) there is no data at all, or b) the data cited are 
irrelevant, though possibly catchy and inflammatory.

>     Goodmail is just going to sell out to the highest bidder(s).

Goodmail has published criteria.  That makes one phase of analysis 

They have no track-record.  That makes the second phase of analysis a matter of 
pure conjecture.

If one does not find that their published criteria automatically unacceptable, 
then your statement is a prediction based on no data.

Hence it really translates into: Nobody who tries to make a profit ever has any 
integrity. (Combined with:  The market has no corrective forces against 
companies that are in the integrity business but that, themselves, have no 
integrity except to seek profit.)

  Even if
> they don't do it today, they'll get bought by someone who will.  And 
> then everyone who built Goodmail into their system will have given 
> spammers a level of unequaled access.

An argument that entropy will eventually convert the entire universe into pure 
randomness does not mean that we cannot find and enjoy productive coherence in 
the intervening eons.  As for myself, I enjoyed my dinner last night, in spite 
of knowing the the universe will eventually end.

In other words: even if you are right, the key flaw in your analysis is that it 
pertains to the indeterminate future, rather than any possible immediate benefit.

>     There's an old saying about power and corruption.

True.  That is why a) it is important to have real competition, and b) it is 
important to make sure that there are other forces to protect important 
categories of mail.

Or rather:  that is why it is important to make sure that these new services 
provided *added* services, rather than that they become the basis for *all* service.

Ignoring nit-picking about the formal definition of goodmail's "certified" mail, 
versus the formal definition of the term "certified" for postal mail, it is 
worth noting that both provide an *incremental* service on top of a basic service.

Although I understand the concern for possible loss of the basic service, I do 
not understand why anyone would object to incremental value.  I especially do 
not understand claims that an incremental service must, inevitably, cause the 
elimination of the basic service.

   Through these
> types of deals, Goodmail will have unprecedented power over the 
> mailboxes of hundreds of millions or even billions of people, and that 
> kind of power is guaranteed to be absolutely corrupting.

That is why a) their rules and actual operation need to be reasonably 
transparent, and b) community vigilance is an on-going requirement.

But the latter is true for lots of things.

For example, the fact that 60% of the end-user email market lives within a tiny 
number of service providers is something I find far scarier than an incremental 
mechanism from a third-party supplier.


Dave Crocker
Brandenburg InternetWorking

More information about the Mailman-Users mailing list