Re: (ngtrans) Joint DNSEXT & NGTRANS summary

["D. J. Bernstein" <djb@cr.yp.to>]

Paul A Vixie writes:
> so, "let's standardize both and let the market decide."

So you don't mind us deploying clients that look for only AAAA and A?
And you don't mind us deploying caches that don't convert A6 to AAAA?
The BIND company will stop saying things like

   I will note that djbdns does not support A6, DNAME, bitstring labels,
   and other modern features of the DNS. We did not have the luxury of
   ignoring IETF standards; we prefer to work and play well with others.

and start saying things like ``Apparently the market has decided to use
AAAA'' and ``Beware that clients won't look for your A6/DNAME records''?

> who in this mix is representing the interests of the average MIS manager?

Such as a manager wondering why his users are having sporadic problems
reaching www.i.am?

Or a manager who wants his site renumbered _now_, and can't wait several
years for universal client-side A6 support?

Or a manager unable to find DNS software with the features he wants,
because DNS implementors have to waste so much time on garbage like A6?

> if ipv6 is no easier to renumber than ipv4, then we're going to see large
> scale NAT.  much larger scale NAT than we see today.

Explain. If the renumbering costs are the same, why will NAT become more
popular? Why will users become ``even more captive''?

The NAT sites I've seen are using it for precisely one reason: to
conserve precious IPv4 address space. They won't use NAT with IPv6.

---Dan