4 ms·
On the imminent phone number shortage
- lmm 13y agoThis is just dumb. It's on the same level as "why not make the whole plane out of black-box stuff".
- jacob019 13y agoThinly veiled opinion piece on ipv6. Like it or not, ipv6 is happening.
- goodcanadian 13y agoYou might not realize it, but there's an imminent phone number shortage. Citation needed.
- kiallmacinnes 13y agoHe's actually talking about IPv6 ;)
- gggggggg 13y agoWhat's the PBX fix equal for the ipv4 issue?
- kiallmacinnes 13y agoWell.. The fix would be entirely different, so I'm not sure. Everyday people type phone numbers.. Sysadmins type IP's. His argument that "people never type phone numbers anymore" is entirely without any backing in truth IMHO, which makes the whole article moot.
- jacob019 13y agoNAT & NAT penetration technologies.
- kiallmacinnes 13y ago"NAT penetration technologies" is exactly what PBX's do.. With IP and NAT, you call 1.2.3.4 and ask for port 5432 - which sends you to the right computer. With phones and PBX, you call 1234 and ask for extension 5432 - which sends you to the right phone.
- mproud 13y agoThe thing is, there are no good alternatives to IP addresses with the Internet, whereas with phone numbers, you can easily do without. Between SMS, e-mail, Facebook and Twitter, plus Skype/FaceTime, who really needs a phone number these days?
- rachelbythebay 13y agoI love the PBX analogy for alternative ideas. I mean, what if we were actually working towards a scheme which essentially put one 32 bit address inside another, and you explicitly specified gateways? It might work really well for real-world use cases. Say I'm a small business and I sign up with the local cable modem provider and get a static IP address from them: A.A.A.A. Then I have my machines configured inside my network as 10.0.0.1, 10.0.0.2, 10.0.0.3, and so on. Just like now, right? Well, now let's say I want those machines individually addressable from the outside world. We'd need some way to encode that in the headers, and by extension, into things like DNS. serv1.example.com. IN A+ A.A.A.A:10.0.0.1 serv2.example.com. IN A+ A.A.A.A:10.0.0.2 So things change and now I want a second ISP. They also give me a second static IP address: B.B.B.B. I add this as a second gateway and use it to extend the possible ways to reach my machines. serv1.example.com. IN A+ A.A.A.A:10.0.0.1 IN A+ B.B.B.B:10.0.0.1 serv2.example.com. IN A+ A.A.A.A:10.0.0.2 IN A+ B.B.B.B:10.0.0.2 Yep, my same systems are now accessible either way. I just "multi-homed" my network without getting an ASN, setting up BGP and peering, buying a huge router, or any of that. It didn't require any coordination between the ISPs, and I didn't need to get "portable" IP space. Now, obviously, the clients would need to know how to speak this. But hey, the backbone doesn't have to change at all. To the rest of the Internet, it looks like (slightly fatter) packets flying around to A.A.A.A and B.B.B.B. They might have some extra IP options on board (to encode the internal address), but who cares? I'd think this would start showing up in systems which are mass-produced and which nail up a bunch of connections to servers - game consoles, cell phones, that sort of thing. The fact that only the endpoints need to know about it makes it even more obvious. And hey, the servers don't even need to know about this scheme if the gateway manages that for them. (Sound familiar? It's because I wrote about it in April as a "half-baked idea": http://rachelbythebay.com/w/2013/04/22/ipspace/ http://rachelbythebay.com/w/2013/04/22/ipspace/)
- p1mrx 13y agoWhat you're describing is essentially 6to4, minus the anycast gateways. The problem with using 6to4 as the IPv4 upgrade path is that it solves the easy problem (pushing big addresses through the core of the network), but not the hard problem (updating all the edge devices, and every piece of software that stores or manipulates an IP address.)