Re: [dbound] The proposals before us

[Casey Deccio <[e-mail]>]

On Mon, Sep 12, 2016 at 9:07 AM, John R Levine <[e-mail]> wrote:

> Sure.  Please see Section 6, "Examples":
>>
>
> 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.


> Fair enough.  Admittedly, the principles behind the current PSL are
>> actually more robust than the current algorithm, data, and implementations.
>> ...
>>
>
> In practice, people will write libraries to wrap the calls, and they'll do
> whatever they do if they don't find the data they're looking for.  I agree
> there's no obvious default, since you can't tell by inspection whether a
> TLD is a private vanity TLD, a gTLD that delegates at the second level, or
> something else.
>

Yes - no obvious default of the nature of a name is the reason why the PSL
and DBOUND exist :)

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.

Casey