In the early 1990s, internetworking wonks realized the world was not many years away from running out of Internet Protocol version 4 (IPv4) addresses, the numbers needed to identify any device connected to the public internet. Noting booming interest in the internet, the internet community went looking for ways to avoid an IP address shortage that many feared would harm technology adoption and therefore the global economy.
A possible fix arrived in December 1995 in the form of RFC 1883, the first definition of IPv6, the planned successor to IPv4.
The most important change from IPv4 to IPv6 was moving from 32-bit to 128-bit addresses, a decision that increased the available pool of IP addresses from around 4.3 billion to over 340 undecillion – a 39-digit number. IPv6 was therefore thought to have future-proofed the internet, because nobody could imagine humanity would ever need more than a handful of undecillion IP addresses, never mind the entire range available under IPv6.
This is a hill I’m willing to die on: the reason nobody wants to use ipv6 is the fucking colons.
You don’t need to use shift to type in an ipv4 address.
It’s fucking stupid and annoying.
Nah, it’s that the tooling on the infrastructure side still sucks in current year. Why tf as an AT&T customer can I not get more than one PD from upstream without editing my dhcpc conf? And why can my gateway, once getting multiple PDs, not tell me which ones I’ve assigned out to subinterfaces? Not to mention there’s still infrastructure devices which can’t have v6 management interfaces, layer3 switches which can’t act as RAs, all sorts of shit like that.
Isn’t the recommended strategy to delegate a larger prefix to the gateway and then make smaller subnetworks from that for each interface? Then you don’t have to deal with separate prefixes.
It is. How At&t handles it is they hand out only 1 /64 of the delagted /60 (by default) per explicit IA-PD request, rather than the full /60 they allocate by default (which, note, is not on a nibble boundry like it’s supposed to be, and you only get half of that as usable).
But on every gateway device I’ve used, even if you get a full /56 prefix, you still have to explicitly assign out each /64 to sub interfaces. Really, ipv6 is a bunch of great ideas which were ruined by shitty implementations everywhere.
Wow that’s extremely annoying.
On openwrt, you just tell the interface to grab a /64 from any other interface that tags its delegation as shareable. And on the source interface you can specify with what priority those /64s are given out.
That seems reasonable to me as far as implementations go. The ones where they will autoassign always just overload pd index 0 which is worse than doing nothing imo lmao
Exactly. Most good ISPs will give you a /56 or /60 range if your router asks for it, and then you can subnet it into multiple /64 ranges (16 /64 networks for a /60, or 256 networks for a /56).
I have three VLANs with internet access (main, guest, and IoT), and each one gets its own /64 range.
Note that you shouldn’t use subnets smaller than a /64, as several features (such as SLAAC and privacy extensions) rely on it.
Note that you shouldn’t use subnets smaller than a /64, as several features (such as SLAAC and privacy extensions) rely on it.
it seems so silly an oversight of ipv6. sometimes you just can’t have /64 subnets because the ISP only gives you a single /64
That’s an ISP being dumb, not a problem with ipv6.
Especially annoying since colons are already used to specify port number. Sure you can put it all inside [], but that just makes it even uglier.
That does bother me a bit
Plus you can type a v4 address with one hand on the numeric keypad alone.
not in europe, we have a colon there
If caps lock changed + to : and made the rest of the stuff on the sides of the number pad into a through… (f?) by default, I no shitting, honestly believe we’d see a huge uptick in ipv6 adoption.
It seems like such a small thing but it’s the kind of inconvenience that, repeated SO MANY TIMES over the course of a long period will make you want to destroy things.
Why would you do this?
Sometimes you gotta access your router. Especially when it’s giving you headaches.
How often do you type IP addresses? That’s what DNS is for. The only time I use IPs is when I haven’t configured a DNS record yet (and in the DNS configuration, of course).
MORE TIMES THAN THERE ARE ATOMS IN THE UNIVERSE, CHILD.
IT IS THE ONLY WAY I CAN CUM IF I CANNOT CUM THEN I CANNOT SHUM AND IF I STOP SHUMMING ALL THE UNIVERSE CEASES TO EXIST.
someone get this guy a ten key and a box of kleenex
Just rely on DNS. After all, it’s not like there are memes on its legendary reliability.
There are some things you shouldn’t use DNS for such as DNS. However, IPv6 has build in shorthand notation that makes it much easier.
And harder to remember when trying to ssh into your servers
This issue can’t really be fixed, more addresses inherently means larger addresses.
If anything I’d say IPv6 works to mitigate this by making each byte 2 charactersYou can use short hands or dns
ssh aliases changed my life
It doesn’t seem like that big of deal
My observations around the lack of deployment of IPv6, and many other iterative technical refinements, is the majority of those who make technical decisions today are so thoroughly abstracted away from roles which understand the need and benefit of it.
‘IT’ has become mature to the point that technical competency has been neutered by economics being granted an unchallengeable veto over its design and implementation.
Until the world recalibrates and suppresses economics to an appropriately subservient status, we won’t see properly robust technical advances again.
What do you mean by economics in this context?
It’s a bit convoluted, but…
I consider economics to be the control mechanism which filled the void in society vacated by religion after the enlightenment.
- Finance replaced mythology.
- Banks and insurance replaced cathedrals and churches.
- Economists replaced the clergy.
- GDP replaced God.
Basically economics, as an intellectual discipline, is far closer to theology than physics. Given that, I don’t respect the positioning of economics as dictating the function of legitimate fields.
The major problem with the IPv6 migration was the lack of a clear plan. We are now much better off since translation mechanisms are becoming more standard.
It kinda amazes me that CLAT still isn’t universally supported, e.g. Windows is only just getting it now.
Mobile adoption is pretty good on the other hand, a lot of the big providers went v6 only years ago without much issue (e.g. Telstra, my provider, went single stack v6 back in 2020.
One of the largest ISP’s in the UK (virgin media) still doesn’t support IPv6.
Stop trying to make “wonks” happen.
