Stewart Stremler wrote:
IPv6 over IPv4?
That's a hack, too. Clever one, indeed... but irrelevent with regards
to NAT.
It is a great way to enable transition without having to do a complete
cutover.
An all IPv6 network will still have IPv6 network addresses to be bought.
Your ISP ain't gonna give 'em to you for free. And if the 'Net is all
IPv6 and no IPv4, well, you can still be made to pay per address.
Is it so _wrong_ to _pay_ for what you use? At least a *little* bit?
Nothing wrong at all with paying for what you use. But with IPv4 IP's
are relatively scarce which makes them expensive. And it is an
artificial scarcity caused by the limitations of IPv4. When we go to
IPv6 the number of IP's available will increase to such a degree that
they will be free to a very good approximation. For what it's worth I
currently pay a lot more than the typical DSL subscriber so that I can
have 5 static IP's on my DSL line. But I wish I didn't have to pay this
much, especially since it is due to an artificial restriction which we
have a solution for but people are dragging their heels implementing. I
guess we will just have to continue waiting until more of the supply of
IPv4 addresses gets used up. How expensive does your single NAT'd IP
have to get before you would consider something else? :)
I blame the programmers for being lazy. It's not the programmer's
right to claim that the user's policies are an "unnecessary adversity".
Remember, _security_ is claimed to be an "unnecessary adversity" by many
programmers. That doesn't mean we listen to 'em.
NAT is not a "user policy". It is a policy imposed on IPv4 users by
those who refuse to implement IPv6. Users in general would much rather
do without NAT because it breaks things and causes timeout issues like
Carl had and makes them have to forward ports and increases complexity
in general.
--
Tracy R Reed
http://ultraviolet.org
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
