Danny,

On Nov 17, 2009, at 22:07 MST, Danny McPherson wrote:
> On Nov 17, 2009, at 9:57 PM, Shane Amante wrote:
>> One way to avoid this is to number the DNS and/or mapping system components 
>> [exclusively] inside Locator space, thus there's no circular dependence on 
>> getting _to_ a "mapping resolver".  Thus, the routing system containing the 
>> set of LOC's does not have a circular dependency ...
> 
> I understood the distinction..  Perhaps not referring to it as "the DNS" 
> would alleviate a great deal of confusion

Apologies for the confusion; however, in the past on the list, some people have 
pointed out that there is a difference between "the DNS", the protocol, and 
(separately) DNS as currently deployed for hostname to IP address resolution.  
I would agree that sub-optimal /implementations/ of the DNS have lead to poor 
experiences and/or operational practices.  OTOH, it's impossible to deny the 
existence proof that many multi-billion dollar organizations (e.g.: content 
providers of news, information, music, movies, etc.) depend on the successful 
functioning of the current DNS -- even with its flaws, (lack of DNSSEC probably 
being the most notable example).  Said differently, the organizations that 
understand the critical role that DNS plays in their day-to-day business *do* 
figure out the best practices to engineer and operate it, (or, even, outsource 
it to companies that *can*).  Those that don't, well I guess you know the 
answer there.  :-)

Ultimately, the point I would make (not necessarily to you, but others on the 
list) is that DNS as currently deployed gets a bad rap (perhaps, unjustifiably 
so); however, a lot of thought, engineering, etc. has gone into (and, is 
continuing to go into) "the DNS".  Thus, IMHO, it would be a mistake to dismiss 
"the DNS" as a potential ID-to-LOC mapping system out-of-hand, given the 
experience we as a community have with it and the dependence we already have on 
it.  Certainly there are things that can be done to improve "the DNS", but 
those are likely very well understood at this point vs. imagining and 
engineering a whole new mapping resolution system.

 

> - and might even focus folks on 
> more optimal mapping resolution mechanisms in the future (i.e., the current 
> DNS is at least as broken as the routing system, methinks :-)

See above.  :-)

-shane
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to