On 28/Jan/10 20:03, Steve Atkins wrote:
On 01/28/2010 09:31 AM, John Levine wrote:
Even worse, users will learn what the button means by the effect (they
think) they obtain by hitting it, which may vary.
Web mail has had spam buttons for years, and the users seem to have
figured out how to use them. Can you explain exactly how the issues
with a spam button in a MUA would be different?
Because some ISPs' sizes result in vanishing likelihoods that Alice
and Bob share the same server (see the "ideal" FP example). Nor we can
expect smaller ISPs to implement FBLs on their own initiative...
Most of the people I see arguing that the "this is spam" button isn't
a good user interface for users to provide their thoughts are spammers
or grey area bulk mailers.
There's a smattering of operationally inexperienced anti-spam nuts,
and dilettantes who have nothing more productive to do than argue about the
wording of AOLs user interface and the colour the AOL bikeshed should be
painted, but it's mostly spammers and dirty bulk mailers.
These statements are highly enlightening! They implicitly contain an
explanation of why we cannot define the term /spam/, and what kind of
hypocritical equilibrium results from coloring bulk mailers with
various shades of gray. Since marketers want to send ads and users
don't want to receive them, MTA operators can pretend there is no
conflict and provide a direct-to-junk channel. Is everybody happy with
this [non-]solution?
IMHO it doesn't work very well: First, because ambiguity undermines
reliability. Second, because users have no direct control.
_______________________________________________
Asrg mailing list
Asrg(_at_)irtf(_dot_)org
http://www.irtf.org/mailman/listinfo/asrg