Re: [dbound] The proposals before us

[Casey Deccio <[e-mail]>]

On Mon, Sep 12, 2016 at 10:17 AM, John R Levine <[e-mail]> wrote:

> Ah, of course.  Oops.  I mostly wanted to clarify that odup does a tree
>>> walk, so if a hostile sender used addresses like
>>> a.b.c.d.e.f.g.h.i.j.k.l.m.n.o.p.q.r.s.t.u.v.w.x.y.z.blah.example it
>>> could
>>> produce a lot of noise traffic.
>>>
>>
>> Yes - it all depends on the DNS setup on the other side.  That is common
>> to
>> both proposals.
>>
>
> Nope.  In my proposal the number of lookups depends on the number of
> boundaries, not the number of components.


The moment you said "hostile sender", all bets were off.  They make as many
"boundaries" or any other component as they want.

My point in the earlier email was that the algorithm should be well defined
>> and with a sane default, so once deployed, if nobody does anything
>> different, their behavior doesn't change.
>>
>
> That'd be fine with me, but I wouldn't count on implementers paying more
> attention to this default than to the one for the current PSL.


Yes, that's exactly the point of "sane default": don't expect anyone to do
anything different.

Casey