Hexadecimal is used because a network is designated by an N-bit prefix, and it's *much* harder to manipulate bits in decimal, especially when each number is 16 or 32 bits long.

Once you've gone to hexadecimal, using dots to separate the address leads to ambiguity. Is a.b.c.d.e.f.beef.de an IP address or a hostname?

With IPv6, your network will have its own 48 to 64-bit prefix. Once you remember that prefix, you can choose your suffixes to be as simple as you'd like.

You can only omit one run of zeros, because otherwise the length of each run would be ambiguous.

But small children will naturally become taller over time. Why are we mandating a technical solution for a problem that solves itself in 10 years?

When IPv4 was first developed, subnets were only allocated on 8-bit boundaries. Since CIDR rolled out in the 90's, the subnet length is now arbitrary, and working with subnets requires doing tedious decimal-binary conversions in your head. It's a mess.

IPv6 uses hexadecimal, because a hex digit represents exactly 4 bits. This makes the CIDR math really simple.

Now, what if we used hex with dots? Is a.b.c.d.e.f.beef.de an IP address or a hostname? Hell if I know. That's why we use colons now.

Precisely. Now whose side are you on?

And the second approach (a reverse SSH tunnel) only works if you have that server at work with a public IP. What happens when you want to connect to your friend's computer, which happens to be behind a different ISP's NAT, or even an overlapping RFC1918 space within the same ISP?

You basically have to lease a public IP on a server somewhere, to accomplish things that should be trivial.

Manufactoria - http://pleasingfungus.com/

Best programming-related Flash game ever.

How would you propose to make IPv6 backwards compatible? How can a device that only understands 32-bit addresses send a packet to a 128-bit address, and what problem would it solve that NAT64 doesn't?

T-Mobile disagrees. They're deploying an IPv6-only NAT64 network as we speak:
http://groups.google.com/group/tmoipv6beta/web/t-mobiles-technical-architecture

The actual number of addresses is a red herring, since every user will be assigned between 2^64 and 2^80 addresses.

For all practical purposes, IPv6 supports about 64K times as many users, but each user can have an unlimited number of devices.

So does anyone have an actual link to an example of the player? That seems like a rather blatant omission from the article.

The raw 2**128 number sounds impressive, but it's mostly meaningless, since each customer is assigned somewhere between 2**64 and 2**80 addresses. In practice, IPv6 supports maybe 64K times as many users, but each user can now have an ~unlimited number of devices.

[citation needed]

You're technically correct, but the mob has spoken. Just call your 3D "holographic" and get on with your life.

You can deploy an IPv6-only network where the clients can still connect to IPv4-only servers.

It's called NAT64:
http://tools.ietf.org/html/draft-ietf-behave-v6v4-xlate-stateful

Isn't a carbon nanotube just a rolled-up strip of graphene?

Are you sure about that? Their website says $40 for 1GB, $60 for 5GB:

http://www.virginmobileusa.com/mobile-broadband