I’ve heard that the internet is moving to IPv6, which will change IP addresses from 32bit to 48bit. I also heard that they’re getting rid of the Classes A,B,C,and D. I don’t really know too much else. My biggest questions are:
Who is going to make this change?
How is it going to be done?–Sounds like a pretty big project to me.
When will this take effect?
What will take the place of the Classes a,b…
Will this change the current IP addresses?
and
Where can I find more infomation on this? I don’t even know where to begin.
Oh, and supposedly: “IPv6 gives us slightly over 300 trillion trillion trillion ip addresses”
(from an email I got from a list serve)
Many operating systems already support IPv6 (I’m talking about “real” operating systems here, not Windows.) Since the IPv6 address space is backwards compatible with IPv4, the transition does not need to be immediate. Once a large majority of people (say 95% or so, can’t wait for everyone) ICANN will start allowing use of IPv6 specific address space on the public Internet.
Considering what we know of beaurocracies, probably never.
Not entirely sure, it’s been a while since I was reading about IPv6. AFAIK, they’re doing away with network classes altogether, and will probably designate IP ranges on a first-come, first-serve basis.
No, since IPv4 addresses will be forwards-compatible with IPv6.
IPv6 is 128 bits. 2[sup]128[/sup] is more than 10[sup]38[/sup] addresses (you can figure this quickly by dividing 128 by the number 3.32, because there are 3.32 octaves per decade).
Well… The fact that a mapping is defined between IPv4 and IPv6 addresses does not in any way imply backwards-compatibility. As it turns out, IPv6 is decidedly not backwards-compatible with IPv4. Not even a little bit. Transitioning to IPv6 across the Internet will be as difficult as transitioning to AppleTalk or IPX. It’s a completely different protocol. The mapping of addresses is a miniscule part of the transition problem.
The real problem is that 3 classes of vendors must have reason to transition: clients, servers and routers. The clients aren’t going to transition until the majority of Internet services are accessible via IPv6 and all routers between said client and said server route said protocol. Servers aren’t going to transition until Clients and Routers exist. Routers aren’t going to transition until Clients and Servers exist.
Of course, the dig at Microsoft is kind of silly, since a huge majority of clients on the Internet are MS ones, and as you point out, huge acceptance (you said 95% and I concur) is necessary.
Some 8 years ago, I was writing routing protocols for a major router firm, and many were freaking out worrying about running out of IP addresses, insisting that IPv6 must be put into place immediately. I said it then, and I say it now: IPv6 won’t happen for another 5 years, and even then I’m sceptical that it will at all.
