On Fri, Mar 28, 2003 at 02:49:10AM -0500, John R Levine wrote:
People can ask for any e-mail they want. That's a real and useful choice.
Indeed, but not one in this protocol proposed.
But in the end, I see this banner as effectively equivalent to writing
into SMTP, "Senders MUST NOT send UBE"
Which might indeed be what most people want, but it's not for the IETF
to do, and as you point out, it's not that meaningful without a law behind
it, which is also not for the IETF to do. The IETF defines protocols that
lets users specify their choices.
The goal here is not to give users a choice, the reality is it is to
have everybody specify the same choice, for who wouldn't want to say they
don't want spam? I am concerned by defining a technically meaningless option.
It's like "Mail-type: Electronic"
There's an entirely sensible reason for definining it this way, which I
suspect you already know. Server banners allow laws to be phrased in
terms of property rights, as opposed to laws that would completely ban
classes of mail which get tied up in knots with arguments about free
speech. I happen to think the free speech arguments are misguided,
Well, you know I have many thoughts on these issues, but I will refrain from
going onto them on this list, where I prefer to discuss the merits of
engineering solutions.
The engineering question here is, "Do users need a way to refect their
UBE preferences during the SMTP dialog, and if so, how should they be able
to do it in a way that meets their desires." I'm not sure a basic banner
would meet that test, but the concept can be explored more. However, it's
only really meaningful in the concept of law.
Any sort of opt out or tagging rule or any other limit on spam is
meaningless without a legal stick to whack violators, so I don't see much
point in defining constructs that have no chance of having the necessary
stick attached.
Well, that's certainly true, though in fact this does not require the
congress. For example, ISP TOS can and do have the ability to cut off
users for sending traffic that is in violation of protocols. In many ways
it can be clearer than just the no spam rule.
However, they can't apply this sort of test if the ISP has its own relay
server since the proposed protocol does not allow the user of a relay server
to even see the policy.
On the nearly-technical side, I am also concerned about how much success there
will be in defining the terms. Some what Solicited to be implied on a
pre-existing
relationship (possibly with a "respect unsubscribe requests" rule) and others
want solicited to mean explicit subscription. I would hope they can be
reconciled. And there are arguments on bulk in either case and where the
cutoff should be, though in truth all low cutoffs are probably enough to solve
the spam problem -- if they were obeyed. Alas, I have poor faith on any
of this being obeyed. As you know, a significant fraction of the spam
sent today is already illegal under non-spam laws --- fraud, illegal medical
sales and all the rest -- so I see little hope in those spammers obeying
the No-UBE sign we might hope to put up on our mailbox.
Back to the technical, I would still recommend that any protocol defined
in this matter allow individual choice. Thus an ability to express desires
after a RCPT command, in addition to at the banner, with a requirement that
new mail tools SHOULD support this while older ones may only support the banner.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg