ietf-asrg
[Top] [All Lists]

Re: [Asrg] define spam

2003-03-29 00:18:54
On Fri, Mar 28, 2003 at 09:32:42PM -0800, Dave Crocker wrote:
Folks,

This groups needs to settle on a single, precise, operational definition
of "spam".  A definition that permits objective, reliable tests for
qualification.

Actually, our charge is to investigate ways to solve an engineering problem --
servers, pipes and mailboxes overloaded to bursting with unwanted messages,
costing resources and damaging the utility of mail.

We may or may not need to define spam to do that, but so far I haven't
seen much luck.

I had always thought the best solution was to take all the major definitions,
and then take the _intersection_ of those definitions.  Then look at that
intersection and say, "how much of the problem does this cover?"  And more
to the point, "Does this cover enough of it to take back our mailboxes?"

The reason to go with an intersection, is that if a message is in this
subset, everybody agrees it is spam.  They may think that there are other
things which are spam of course, but in theory if those are just a tiny
fraction of the intersection set, they should be happy.

This seems like the logical approach, but it has the curious consequence
that the definition it produces is possibly not identical to anybody's specific
definition. 

Remarkably, some people fight a lesser definition than what they prefer, even
if they will agree that the spams not in that definition are well under 1%
of the total.   They feel it is particularly important to get every last one.
I'm not sure why, I don't see an engineering goal.  I suspect it's because
the spammers make us so angry many are not satisfied until they are all
crushed.

My current impression of the intersection definition is mail that is:

                    Sent as part of a mass mailing, which is to say some
                        person commanded that mail be sent to more than
                        a few hundred people

                    The recipient never solicited the mailing, nor has
                        had voluntary contact with the sender in the past

                    Is commercial (Many people request this in the definition,
                        but usually only if not including the mass mailing
                        component, so this might not stay)

                    The sender has not been asked to stop mailing the
                        recipient

                    If was not directly solicited (ie. is sent due to
                        an existing relationship) the sender adheres to
                        a reasonable and operational system to allow the
                        recipient to ask not to be mailed again (See #4)



                    


        
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg



<Prev in Thread] Current Thread [Next in Thread>