Re: Wither DNS? (was Re: What is the Private Sector?)

[Jeff Williams <jwkckid1@ix.netcom.com>]

Robert,

Robert J. DuWors wrote:
> 
> At 12:59 PM 9/29/97 +0100, Jeff Williams wrote:
> >
> >Robert J. DuWors wrote:
- small snip -
> 
> I am not sure that being on the fault line is the same as being the
> balancing point, in fact more often the opposite...

  Possibly correct here.  I stand corrected.
> 
> I would like to see consideration of moving beyond DNS as we know it into
> more scaleable, and more specialized, and more flexible schemes [these do
> not have to be mutually exclusive].  Debating the policies/politics of
> dragging around the boat anchor of an infrastructural component best suited
> to an earlier, smaller, centrally controlled network (from a policy
> perspective) could be usefully augmented by forward looking consideration
> of a DNS replacement and/or modification.

  This is my understanding of what Internet II will indeed provide for.
> 
> I personally would like to see proposals for a SDNS (Simple DNS) and its
> possible relation to new forms of higher level naming schemes (such as
> Universal Resource Identifiers and Universal Resource Names).  I believe
> that it should have four basic design goals in pursuit of "doing one thing
> and doing it well" at the true balance point between the transport layer
> and an open ended number of upper level application layers:
>   . permit sys admins one level of persistant indirection free
>     from the constaints of IP address (classes, subnets, etc.) and
>     necessary changes to IP (the IPvX series),
>     [strictly speaking this is already the second level of indirection
>      because IP addresses themselves redirect in a routing oriented
>      manner the physical network MAC addresses, but this level frees
>      the admin from routing and IP owernship constraints as well as
>      the natural evolution of IP addresses in order to accomodate global
>      scaling],
>   . avoid any other external semantics of its own,
>   . work well with upper level indexing/identification schemes
>     that will provide their own context [including, ironically,
>     a fragmented conventional DNS space - something is going to have
>     to provide root server qualification!],
>   . and be fast and incredibly simple.

  Tall order, but definatly doable.  As I said above Internet II
does or is supposed to do most of this.  Along with IPV9.
> 
> It should avoid historic relics like MX records.  All extraneous
> registration information (if any) should be kept externally in open
> standard companion directories in a well defined manner integral to the
> formal SDNS protocol (RFC) definition (e.g. do an ISO style "profile" of
> the LDAPv3 standard (when officially blessed) for this companion role).
> Any interest on line or off?

  Yes I agree with this compleatly.  
> 
> Regards,
> Rob DuWors
> 
> --------------------------------------------------------------
> Robert J. DuWors  <mailto:rjd@csgroup.com> tel:+1.440.255.2869
> Connected Systems Group  7638 Aster Drive, Mentor, Ohio  44060

regards,
-- 
Jeffrey A. Williams
DIR. Internet Network Eng/SR. Java Development Eng.
Information Eng. Group. IEG. INC. 
Phone :913-294-2375 (v-office)
E-Mail jwkckid1@ix.netcom.com