Re: [dbound] On (not) moving forward

["John Levine" <[e-mail]>]

>At this point, we would like the working group to consider and discuss why we appear to be entrenched, and what might get us moving.

I've been scratching my head, too.  As best I can tell, this is one of
those problems where the more you think about it, the more twisty
passages there are.  So perhaps a revised goal would be to ask what is
a minimal design that would still be better than downloading the PSL?

I shamelessly argue that my draft is probably it.  The lookups are
fast, one plus the number of boundaries in a name (not the number of
name components), the names can be self-published within the domain's
tree, or in a third party tree, and it can handle different
applications with different boundaries.

The Decchio draft requires a new _odup top-level domain which I think
makes it a non-starter, and the lookup process is so complex I don't
entirely understand it but it appears I think it can involve tree
walking.  (Casey said that he had another version without a separate
TLD, but if he's posted it I can't find it.)  The Yao draft still has
problems with wildcard semantics.

I'm not arguing mine is perfect; Deccio can identify both boundary
domains (co.uk) and organizational domains (example.com) which I
don't, my organizational domains are implicit, so we might want to add
some of his qualifiers.  

Andrew's always wanted to do cross-tree relationships, e.g.,
example.org is the same as example.com, but as far as I know nobody's
come up with a way to do that without putting boundary records on
every name in a subtree that might be related to something else.

So how about if we pick a draft, preferably mine, and say either it's
good enough to move ahead, or else say why it isn't.

R's,
John