Re: [dd] [Ext] starting charter text for the DELEG BOF discussion

[Dave Lawrence <[e-mail]>]

Havard Eidnes writes:
> > DELEG is a replacement mechanism for delegating the name space.
> 
> That ("replacement mechanism") may be the intended goal, and might
> have been feasible to acheive if it was invented in the early days
> of the DNS.

For what it's worth, I haven't heard any of the advocates of changing
the delegation mechanism even hint at believing that it would
completely supplant NS records.  I won't be so bold as to claim no one
has never made that remark, but it is not the common belief.

Yet the "replacement mechanism" language is still apt.  A resolver
that follows DELEG instead of NS has, in all practical senses,
replaced NS during that part of its process.

> However, we are quite far from that situation, and I have not seen
> any discussion about "deployment considerations" for DELEG in the
> current environment.

That you haven't seen it does not mean it has not been happening.  The
conversation has been present among software developers, operators,
registrars, and registries.  More of it can be happening on the list,
yes, and will be happening in the working group, but rest assured that
deployment considerations have not been ignored.

> Or is the plan to exert pressure on folks who don't upgrade their
> resolvers to versions supporting DELEG?  How?

As with my earlier observation, in none of my discussions have I
encountered anyone who has suggested that there be a plan to exert
explicit pressure on any part of the ecosystem, be that to resolvers
or authorities or registries or registries.

> Or is that going to happen "automatically" because of the built-in
> benefits of DELEG?

Yes, that's my own view, that features that can be enabled through an
improved delegation mechanism would lead market forces to gravitate to
it.  (I wouldn't use "automatically" to describe it though.)