ietf
[Top] [All Lists]

Re: Change the mailing list protocol, not DMARC.

2014-06-12 07:47:59
On Wed, Jun 11, 2014 at 1:00 PM, Martin Rex <mrex(_at_)sap(_dot_)com> wrote:

Phillip Hallam-Baker wrote:
Hector Santos <hsantos(_at_)isdg(_dot_)net> wrote:

Let me ask, what if a fedex.com employee use this email domain for
subscribing to the IETF list?

Any subsequent problems are irrelevant unless FedEx, the owner of
fedex.com considers them to be relevant.

That is what folk complaining don't get: you don't have the right to
use your employers email or a public email provider's email any way
you want. The domain name owner makes the rules.

As Craster insists: My domain, my rules.

Strange concept!

Does your jurisdiction allow your landlord to interfere with
postal/snail mail that is delivered from or to your rented appartment?


Absent specific government action to protect the tenant, the landlord can
abuse them in almost any way they choose. Which is why we have voluminous
legislative protections for tenants. There is an entire field of law
concerning that.

The reason that is necessary is that the alternative to renting is very
expensive and requires a large amount of capital. But if houses were $6
then it would be perfectly reasonable for society to tell tenants that they
are on their own because they should become freeholders.



In the medium term, lets kill the stupidity of mailing lists with a
protocol that works. NNTP was originally designed to replace mailing
lists. It actually works quite well at that. The only problem was the
IT-Dictator mindset that underlies it: newsgroups have to be approved
by the Commune!

Set up your own mail2news gateway.

/usr/lib/aliases  http://www.tldp.org/LDP/nag/node213.html
main2news script  http://www.sirlab.de/linux/descr_m2n.html


My point is that mail is an old protocol and people who expect that it can
be kept going unaltered in its original form serving all the purposes that
it was never designed for but have emerged over time are going to be upset
no matter what.

Spam is an attack. The people who send spam are hardened criminals who have
murdered at least two people in the past five years. It is futile to expect
that the mail system can continue to operate without changes.

The major ISPs can and will and SHOULD consider the interests of the
majority of their customers as they move forward. If they don't, open email
systems based on SMTP will go the same way as USENET and die because people
don't want to put up with them any more.


The possibility that SAP might force you to subscribe to IETF lists through
a different address does not worry me in the slightest. That is not one of
the uses of email that I consider any sort of priority. I am quite happy
making you change your mail config.

But looking further ahead, it is becoming clear that maintaining mailing
list capabilities is going to become increasingly difficult in the face of
escalating anti-spam controls. Which is why I suggest that rather than the
IETF community reacting to DMARC by refusing to consider any change that
inconveniences members of its club, IETF instead designs a protocol that
addresses the actual needs.


In the SMTP/IMAP model the message is pushed to the sender outbound MTA,
pushed to the receiver outbound MTA and pulled by the client from the
outbound MTA.

The SMTP/IMAP model is thus PUSH/PUSH/PULL

The model of NNTP and Dropbox and many of the blogging comments apps is
PUSH/PULL/PULL. That is a message pattern that we can and should support in
IETF protocols. Because when someone is sending 10Gb of data, the PUSH/PUSH
model is not viable. Better to leave that data siting on the outbound
server rather than have it clog up hundreds of inbound servers and sit
unread.


US6192407 has a priority date of Oct 24, 1996, now would be a good time to
work on such a protocol.