Today’s edition comes courtesy of my brother, who asked for my thoughts.

I learned about this years ago in my networking classes at uni. I already had a kind of peripheral awareness before that, but learning about it made it a bit clearer. You can see from what they say about that one guy in there – he’s been trying to encourage this for 10 years. But the analogy from the other guy about trying to change the road and tyres while you’re still driving is pretty sound, too. One of the things that people came up with as a stop-gap was NAT (where local networks look like one address to the outside, but map different ports to different local IPs on the internal network).

Lots of web applications will break, no question about that. It’s going to be a messy transition, since NAT was only meant to be a temporary fix until someone came up with something better. At the very least it’s going to involve a lot of upgrades to infrastructure, although PCs have supported IPv6 for a very long time, so I don’t think it’ll be the PCs that’ll be the problem. It’ll largely be configuring the infrastructure, and making sure applications don’t break during transition. The non-backwards compatibility issue is the real problem – in some ways, it’s kind of like the problem of people still using phone lines and dial-up modems for internet, rather than using a better infrastructure that’s designed for modern internet use instead of tacked on (like internet over phone lines was). A lot of devices are probably going to become obsolete.

What’s probably more likely is that a dual-system will emerge, and then the “Internet 2″ initiative is going to become a commercial divide. There will be people who operate on devices connected to the older network, and then the majority will slowly move to the newer network because the price will reach some critical point at which they can’t afford to be on the old network and need to be on the new one. Most users won’t really understand the implications, since they never see the underlying network address – it will be up to the domain and web server administrators largely to make sure that the domain names point to the new addresses, and figure out a way to pass along the maintenance costs to the customers by advertising it as premium/better/whatever. Devices and applications will emerge that are only operable on the new network, which will drive adoption, not dissimilar to the transition in phone networks from CDMA to 3G. Eventually the old network will fade out.

All the PC end-users really need to do is have the new protocol enabled, which can be rolled into any of the system updates for Windows/Mac users if they don’t do it themselves, and Linux users probably already have it installed due to the fairly high technical threshold required to operate meaningfully with Linux systems. After that, it’s buying new devices that are compatible with the new protocol, or installing any firmware updates that manufacturers care to provide (if they are able to, there are some hardware limitations after all).

This has been another edition of Technology Soapbox(tm). ^_~