in response to Barry Shein on ASRG

this I was editing in response to http://www.ietf.org/mail-archive/web/asrg/current/msg14061.html but decided it would be better here as a blog entry than as a big list fart.

Running several risks, I'm going to try to counter some of the assumptions and implied proposals I see being made here with aspects of the current vision of the Advenge sender-pays clearinghouse system, which may be coming back up very soon, like next week. I hope to strengthen the vision in response to comments made in response. I am not an anti-spam kook.

On Fri, Nov 14, 2008 at 5:25 PM, Barry Shein wrote:

> > > You create a header.

I believe that in order for a bearer header to work, the bearer header needs to get checked against a central system, which means that a bearer header, although accurately mapping "postage stamp" to SMTP, is inferior to central accounting system. (Credit cards are a central accounting system, so are paper checks.)

> [the system may be ignored.]

Yes, the system may be ignored. This is only a problem when you have Universal Adoption as a design goal. I do not.

> you couldn't (at least in theory) generate your own postage
> and have it accepted as valid by anyone interested. You're right that
> if the recipient hasn't any interest there's little one can do but so
> what?

My idea is that the recipient sets their own price, including the currency that price is to be paid in. The central system manages the bookkeeping for everyone's accounts in all defined currencies, and defining a currency is easy to do. That is, if Barry Shein wishes to define the "Shein Nickel" as the currency required to deliver messages to Barry Shein, and create images of coins with bees on them (Simpsons reference) that's great. If it takes one Shein Nickel to deliver a message to Barry, and he auctions off a million of them, and issues them to whoever he wants to get messages from, that's what Advenge supports. The recipient charges what the recipient charges, the system is nonjudgemental about what makes sense for each recipient.

> > > Monies collected for such postage could be divided up, I won't belabor
> > > this detail.

Draft division:
70% to recipient, 20% to ESP, 10% to central system.

Central system auctions currencies in terms of each other, operating essentially a forex market.

> It's a new medium, we're inventing some things.

Agreed.

> You want to play, fine, here's how it's done...(RFCs etc.)
>
> You don't? Fine! Accept what you like by your own algorithms, or
> accept everything, write your own mail system, etc. none of my
> business.

Should be available as a QPSMTPD plugin next week, adoption to grow at whatever rate it occurs.

[Barry said:]
> > > Anyhow, obviously the entire idea would comprise many pages of details
> > > but I hope that outlines an overview of a plausible system which
> > > doesn't require micropayments, monopolies, or all the other straw man
> > > nonsense which gets thrown at the stage like rotten tomatoes when the
> > > idea comes up.

but, "add a bearer header" is exactly well within the class you dismiss as "straw man nonsense."

> > The difficultly is nobody likes spam, but nobody wants to pay to send an
> > email beyond what they already pay for Internet service. Pay per use
> > for anything trivial (like email) tends to disappear into flat
> > subscriptions, which is what ISPs do now when they charge us for
> > access. And this mailing list would die if we had to pay to send.

The recipient, in this case a mailing list, would set its own delivery policy, such
as "free for subscribers; varying rate in dollars donated to Oxfam (e.g.) depending
on sender reputation for non-subscribers" or might even have for-subscriber rates
depending on a slashdot-like karma score. Come to think of it, having a more subtle
rate curve than subscribe/don't subscribe would help a lot with a whole class of
flamewars seen on mailing lists. But that's for the week after next :)

> I guess no matter how many times I explain this each person gets to
> come to the plate and bat out the same straw men...

you could create (or edit an existing) taxonomy of argument document

> THIS INVOLVES BULK COMMERCIAL EMAIL.
>
> There is absolutely no reason why you would pay anything.
>
> Real commercial emailers might send out millions of emails per day or
> month anyhow.

Real commercial e-mailers, such as the customers keeping Adknowledge.com in their swank office space, are already paying prettily to see their messages delivered. I want to divert some of this income stream to the recipients.

> I get some sort of boring come-on from Amazon almost every day because
> I once bought a book from them. Why should they be able to do that for
> free? They must be sending out tens of millions of these per day,
> maybe more.

> So, without trying to detail to the last bit of minutiae doesn't that
> seem to lead towards a system which might charge large bulk emailers
> while preserving the current status quo for most others?

> Say it gave you, YOU, don't go running for other hypothetical
> examples, 100,000 free messages per day for non-commercial use. Would
> that cover it? Would it work for amazon or some spammer?

Barry, you objected to this off this list, but your on-list counterproposal seems to me inferior. I propose that the default currency, which can be switched by recipient configuration, is an accounted quantity that is awarded for completing Captcha puzzles. One completed reCaptcha gives a sender more than ten deliveries to system participants who have not changed their delivery price (not burdensome), makes it reasonably priced for legitimate direct marketers to want to purchase the credits to deliver their direct marketing messages, giving the Bangladeshi guy who wants a job completing Captcha puzzles all day a customer, and setting the price too high for the people who do direct contact marketing via botnet broadcast.

> Or they could send w/o postage and people who wanted to receive their
> stuff could put an exception rule in their checker if they have a
> checker:

yes, everyone gets a checker.

> if(from(IETF)) then acceptWithoutPostage().
>
> (whatever "from(IETF)" means, SPF, DKIM, etc.)

yes.

> Similar for end-user ISPs, non-commercial usage postage license, 1B
> msgs/month, $1,000/year. It'd be in the noise, mainly a processing
> fee. And they would do well to pass on to their customers the idea
> that if any are doing bulk commercial emailing they will need their
> own license. Etc.

I see no reason to try to involve ISPs and end-users together in the scheme. This suggestion brings pointless complexity in my estimation.

> > [David Wall's] company offers a secure web email platform in which zero spam and
> > [...]
> > [supporting] [B2B] communications situations that really
> > need our service and they are willing to pay for those benefits [...]

Sounds great! Well done! Is the new Tesla everything it's supposed to be?

> > most people would not pay when regular email with all its warts, viruses
> > and scams is free. So while we're totally for making regular email
> > expensive so our product is more universally attractive, it's just not
> > likely going to happen.
>
> Ok, straw man, right?

That looked like "Voice of Experience" not straw man. My company, with its products still in the design and implementation phase, although I had a whitelisting e-mail fowarding service up briefly in 03-04, offers a long tail marketing plan rather than a B2B exclusive channel plan. An exclusive channel certainly works, and it is easy to see how membership in such a club could become a cost-of-doing-business necessity. As someone pointed out, invite-only social network messaging is growing and e-mail is shrinking. Viewed from one angle, I want to bring the invite-only social network model to SMTP.

> > There are too many email users, too many email servers and too many
> > email clients around the world to coordinate for a such a new standard.

A new standard is not required; a new service operating within the current standards is adequate.

> Then I guess we better disband the entire IETF on that basis! Why
> doesn't this reasoning apply to everything that goes on here?

The answer to this question is, because many list participants view the purpose of the dialogue to be, "create a new standard" and therefore proposals, no matter how workable, that do not involve creation of new standards, are seen as off-topic.

> > It would require world-wide government cooperation and mandates -- and
> > if you're holding your breath for that, good luck to you. It's not a
> > technical issue. Just because you can devise a scheme won't make it
> > happen. And if you can make it happen, do so and become filthy rich
> > rather than complaining we all don't get it. Just do it!
>
> Hey look, I'm not complaining, I'm injecting some ideas.

Barry, your ideas are not new. Do you think they are?

> You're the one whining about them, and mostly nonsensically I may add.
>
> Have you ever considered re-reading your own posts with a little
> self-criticism? A skeptical eye? Imagining how someone who doesn't
> agree with you 100% a priori might read them?

Did that remark need to be made in a public forum? I would like to again offer my rule of thumb for reducing the heat level of mailing list discussions, and that rule is, "Complaints that something is off-topic are nearly universally off-topic and should be made off-list."

I'm going to make this a blog entry instead of a list post.

--
"Slingshot delivers all mail shipments to the post office by bicycle!"

previous - next