2010 could be the last year for IPv4 as we know it

arstechnica.com

8 points·vaksel··5 comments

IPv4 address space exhaustion estimates are about as accurate as Duke Nukem Forever release date estimates.

Oh no; they're not nearly that accurate. After all, the most credible Duke Nukem Forever release date estimates were "Never" for years before it finally died.

I'm predicting a catastrophe on the scale of the transition from 32 bit to 64 bit. Oh wait...

Client-server applications such as the Web and e-mail will work just fine through CGNs and IPv6-to-IPv4 translators, but peer-to-peer applications such as VoIP and BitTorrent, not so much.
...

So enjoy your peer-to-peer applications while you can; their expiration date will be coming up not long after 2010.

Oh, come now. Every major ISP has a VoIP project to support competition with local telcos. The telcos can see the benefits of VoIP and are or will be offering VoIP soon as well.

With http://en.wikipedia.org/wiki/Server_Name_Indication a major driver of hosting IP demand should decrease.

ISPs can provide the vast majority of todays users NAT with no ill effects -- in fact, ISPs could even sell it as a security feature, as you're keeping all of the bad guys away from the end user device.

We haven't even started using mitigation strategies in the USA. We have a long way to go before there's any concern -- we might "run out" of public IPs to pass out to every customer, but we're a long way from not being able to sell some services to people because of a resource shortage.

Yup. IPv4 can be stretched and band-aided to form some kind of network, but it would require always some effort, and I would prefer if that effort would be directed to advance IPv6 proliferation. But as short-term planning is trendy these days, so ISPs probably will just keep holding onto the sinking boat also known as IPv4

Reply on news.ycombinator.com