I try to learn from past efforts - both negative and positive. You on the
other hand demand that we consider the 1983 design of the Internet as
sacrosanct, except of course when you are sneering at people for proposing
'1980s technology'.
Okay, fair enough. Actually the Internet design is several years older
than 1983, and I'll note that I've also recently said that IPv4 is a
dinosaur, so it's not as if I regard it as sacrosanct. And we have a
replacement for the IPv4 protocol that is based on experience with that
protocol, and which tries to improve upon at least its most obvious
deficiencies while preserving its best design points. The parts of IPv4
that I defend are those that were found to work well - like global
addressing, and a clean separation of function between the packet
routing layer and the layer that guarantees reliable message delivery.
One part of IPv4 that I believe really needs to be improved on is the
failure of IPv4 (due to a fixed-length address and a shortage of address
bits) to really implement the catenet model that lets any network be
connected to any point to the existing Internet. Another part is the
poor support for autoconfiguration, especially of small networks.
As for NAT, we have plenty of experience that shows that NAT was, at
least architecturally, a bad idea. NAT can be defended as a stopgap
measure, and it can be argued that NAT points out architectural
deficiencies in IPv4. But NAT doesn't solve those problems, but rather
moves them and makes things worse overall. It also creates a tussle
between the interests of application implementors and the interests of
enterprise network operators where one did not exist before.
DNS has been stretched far beyond its original design goals, and in the
future it seems likely to be stretched even more - there are no shortage
of things we'd like to do with this distributed database. Even with
respect to its original design goals several deficiencies are now
obvious. An orderly and carefully-engineered replacement of DNS is
common sense. What makes no sense is the notion that we have to make
DNS even more critical path than it was intended to be.
Keith
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf