Hallam-Baker, Phillip writes:
Alternate roots are bogus. The only case where they work is where people
do not want to connect to the rest of the world.
That's exactly what a lot of national governments would like to do.
Fragmentation of the root is a real threat, but only if people do
try to do something silly (e.g. Kyle's mom gets congress to exclude
.ca).
That's exactly what a lot of national governments would like to do.
Subsequently we have developed mechanisms such as MX and SRV that
try to change this but people continue to insist on the original
architecture as the only legitimate approach. Witness all the
shouting that has gon on around attempts to store policy information
in the DNS.
When every change must be propagated to a billion machines, a
conservative approach is best.
Arbitrary registration of top level domains would not have prevented
local delegation. The problem with monolithic DNS is that it forces
hierarchy where none exists.
But it does exist, just as it does for the telephone network.
If we were redesigning the DNS today the root would contain as much
information people cared to put in it.
If we were redesigning it today, it would never actually be up and
running. Instead, it would be continually revised in endless volumes
of specifications written by people with nothing better to do in life,
and nobody would implement more than a fraction of the spec, and
they'd always be several versions behind, and their implementations
would never be quite correct, and nothing would ever work together
very smoothly at all.
The reason the Internet is successful is that it was designed before
the bureaucrats took over. The reason X.400 failed is that it was
designed after the bureaucrats took over.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf