ISPs Finally Shifting To IPv6 > Citizen identification

said by matrix3D:

However, I took the math somebody did on Wikipedia and reposted it here:

"The very large IPv6 address space supports a total of 2128 (about 3.4×10^38) addresses—or approximately 5×10^28 (roughly 295) addresses for each of the roughly 6.5 billion (6.5×109) people alive in 2006.[10] In a different perspective, this is 252 addresses for every observable star in the known universe."

»en.wikipedia.org/wiki/IPv6

said by GNH:

I remember when someone thought IPv4 would NEVER be exhausted...

said by GNH:

I remember when someone thought IPv4 would NEVER be exhausted...

said by Gbcue:

Do you also remember when 256k of ram would be all you ever need?

said by elios:

when the first class As were given out
there WAS no such thing as NAT.... sooo yea

said by PapaMidnight:

I remember when the 1GB HDD came out and no one knew what the heck to do with all that space, lol.

said by nixen:

There's very few companies, then or now, that would need an entire /8 of public IP addresses. ...

said by matrix3D:

I'm assuming your post is meant to be sarcastic... but if it's not, why would a process or a thread even NEED an IP address? Only individual nodes or devices need an IP address.

said by patcat88:

We can eliminate port numbers by giving each process an IP address. With the future (jk) cloud computing/internet appliance/X Windows style dumb terminals, all programs will be RPC enabled, and have unique IPs. Handles and objects transparently work over the internet. GUIDs are history, we have IP ranges instead of GUIDs.

said by koitsu:

Please explain your logic here on a technical level, and verbosely at that. I'll save my flames for *after* you're given the chance to explain how this would work on a technical level.

said by patcat88:

We replace TCP and UDP and use a protocol without port numbers. Lets allocate 32 bits of the 128bit IP for all possible RPC/IPC objects an application might use (4 billion). Then we allocate 24 bits of the 128 bit IP address to label each protocol/application, 65000 (16 bits) wasn't enough with IPv4. The 32 RPC/IPC object bits and the 24 application bits combined remove the need for a loopback range. Let allocate 16 bits of the IP address for all the devices one person might own (65000). Now we are down to 56 bits left. In the next couple decades we are going to have atleast 40 billion people on earth since population growths exponentially. Lets allocate 44 bits of our remaining 56 to individual humans (a device might wind up with multiple IPs, one for each of its owners), that gives us the ability to give an IP range/address to 17592 billion humans, we might go intergalactic and spread all over the universe, IPv6 needs to follow us. The remaining 12 bits we give away to actual ISPs, thats gives us 4096 ISPs in the world. We just exhausted the IPv6 address space, time to hastily upgrade to IPv7.

Also there must be some external mechanism to replace broadcast addresses and multicast. Maybe we can allocate one of the 4096 ISPs to be a multicast range.

said by nixen:

And it wouldn't be, had they been used, properly. I mean, does it make any kind of sense for an enterprise to have any public IPs on non Internet-facing systems? Were NAT more widely used, we'd have a lot less IP consumption going on. Hell: ISPs that really wanted to prevent customers from running servers would put their customers on private address spaces.

said by patcat88:

But we already gave half the address space away to MAC addresses.

said by brad:

Try working in the real world and you'll find NAT is not really solving anything. Its just a problem waiting for more problems. If we had more NAT more shit would be broken. yay. that's what I want.. NOT.