On 9/10/2012 10:34 PM, Pars Mutaf wrote:
A single common internetworking layer helps reach other nodes regardless of their underlying technology. Current situation shows that, now we need to reach other nodes regardless of their underlying technology, and regardless of their IP version. This is what Discrete IP does.
The Internet asks each new technology to support IP, and thus leverage everyone else who already speaks IP. In that case, ONE node (the new node) needs to be modified.
Your proposal asks others (potentially many others) to adjust to you. The Internet was created because that doesn't scale.
Not allowing different IP versions is a technology blocker. IPv6 is ready, some entities need it, but they cannot use it fully. This is because it is required that other entities also transition to IPv6.
We already use the kind of translators you're thinking of to support allowing IPv4 to talk to IPv6. It works, but it's cumbersome and has a lot of opportunity for error.
That's why we try to minimize the number of IP versions - at most, to the "preferred" one and sometimes (and hopefully briefly) a "legacy" one.
They do not have to because they do not see the economical incentive. We clearly have a paradox here. Some entities need IPv6, others do not. It is also assumed that IP research is done. Someone may invent a totally new and wonderful concept of IP and call it IPv7. Some entities may use it, while others still use IPv6 and IPv4. Discrete IP solves this paradox: Everybody do what they wish.
People can do that today, but in the paragraph above this one you explain why that didn't work - in the 1960's and today.
Joe _______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
