[JohnFen]

I think that it's fair to blame the specification. I don't think the problem is that the industry has changed, I think it's that there's a huge amount of friction and headache for shifting to it. I think that's because it was too large of a change all at once, combined with the initial specification having some real problems (that did eventually get mitigated in later specs).

I suspect that if IPv6 limited itself to just increasing the size of IP addresses, IPv4 would largely be a distant memory by now.

[WorldMaker]

"IPv4 with bigger addresses" would never have been backwards compatible and would always have been a compatibility break and would always have a slow rollout.

The proposals that seemed backwards compatible were just aggressive CGNAT consolidating even more power in the hands of IPv4 address owners. That doesn't seem like a sustainable fix in the long run.

[JohnFen]

> "IPv4 with bigger addresses" would never have been backwards compatible and would always have been a compatibility break

True, but it would limit that break to a single thing. That's much easier to deal with than the whole basket of things that IPv6 brings with it.

[WorldMaker]

Compatibility breaks are always headaches. There's not "just broke a 'single' thing" when it comes to compatibility breaks. That is why strict semver suggests a major bump no matter how "small" a compatibility break appears to the developer. There is no such thing as a "simple" compatibility break to downstream users.

In general, despite the complex vocabulary about most of it, in many ways IPv6 is simpler than IPv4. Its header has fewer fields. Its QoL/QoS fields aren't accidental hacks on top of old debugging fields but intentionally designed fields for that very purpose. SLAAC is a simpler protocol than DHCPv4, though the algorithm sounds more complex at first. (DHCPv6 is basically as complex, but fewer devices and fewer subnets should need DHCPv6 in the first place.) Much of the "basket of things" that IPv6 brings with it are designed to remove complexity that has concreted around IPv4.

They ripped the bandaid completely off with the backwards compatibility break that they made with IPv6, and apparently a lot of people loved the cute stickers they had applied on top of the bandaid. But at this point it is probably better for the skin below to heal without the bandaid than to continue to sticker and bandaid over that and let all that unnecessary glue fester in place. (To push such a metaphor almost to its breaking place.)

[JohnFen]

> in many ways IPv6 is simpler than IPv4.

It's not really about whether or not IPv6 is simpler than IPv4, though. It's about how painful moving from IPv4 to IPv6 is. And it's very painful. If the only thing that changed between the two was that the IP address space is bigger, it would reduce the pain of changing.

I'm certainly not going to claim that my experience is representative of anyone except for me, but the reason that I'm not going to shift to IPv6 until I literally have no other choice is because doing so is an enormous undertaking. Since IPv6 doesn't bring me any benefit that I care about, there is no reason to do so unless I simply can't get on the internet without it anymore.

Please note: I am not bashing IPv6 here, and I'm not saying that a change isn't needed. I'm just expressing some of the reasons I've seen why people resist changing to it, and that I think adopting it would have happened within a reasonable timeframe if it weren't as ambitious.

[WorldMaker]

> I think adopting it would have happened within a reasonable timeframe if it weren't as ambitious.

There's zero proof that an "extended IPv4" would have been adopted on a "more reasonable" timeframe, no matter how you define "reasonable" (faster, I guess is what you are arguing for?).

Exactly where and how do you expect "just add more address bits to IPv4" is an easier transition than IPv6?

The IPv4 header is a fixed size. You can't add more address bits without breaking existing routers. Period. End of technical story. You could embed the additional address bits in the next layer up (TCP/UDP) but you greatly increase the complexity of routing equipment by making it have to understand those layers, to what benefit? In the dual-stack real world we do that all the time with VPNs and STUN tunnels and other gateways and tunnels. We have those exact same tools, already, and those haven't made the transition any more "reasonable", have they?

But it's worse that while routers don't understand the extra bits, the parts of the addresses that get used (the prefixes small enough to fit in IPv4 headers) have to become massive NAT gateways and become massive gatekeepers of huge parts of the IP address space. We know from deep experience that IPv4 address allocation wasn't "equitable" (ARIN got way more space than RIPE and both got more space than AFRINIC and so on; companies like Microsoft and GE got /8 allocations just for asking in the right years).

Does it make that much sense to establish existing IPv4 holders as the forever "landlords" of the internet? That seems to me to only add more incentives to make the transition more unreasonable than IPv6: why support router initiatives that understand the additional address space when the IPv4 address holders can get "extra rent" if they don't, presumably charging all their "downstream" traffic for their gateway usage? We're in a time where IPv4 addresses have noticeable rental costs, I can't imagine what that would be like in a world where large parts of address space have to be on VPNs controlled by IPv4 owners. That doesn't sound to me like a good present or future for the IP protocol, no matter what.

[Dagger2]

It's honestly not that hard. Looking at your other posts, you think it's hard because you're unfamiliar with it, because you're trying to overcomplicate it, and because you're trying to do everything all at once rather than gradually.

None of these things are IPv6's fault.

Hell, give me remote access to your network and I'll set it up for you -- at least enough to get you started on it if not 100% on every single thing. I don't expect you'll take me up on that offer, but it'd be easier to just do it than tell you about it, since you can't tell people anything: http://habitatchronicles.com/2004/04/you-cant-tell-people-an...

[bauruine]

Why is it so painful for you? As someone that has been running everything dualstack for over a decade I'm seriously interested where people are struggling with it.

The only pain I've ever seen is in corporate networks where all the tooling around the network management are IPv4 only but those would break even if you add a single bit to an IPv4 address.

[magicalhippo]

ISP doesn't provide much in terms of support. They had a IPv6 page for some years announcing it was available for the eager ones, which has been since been taken down. Searching for IPv6 in their documentation comes up empty.

Software isn't ready or was written back in the early days before they figured out how IPv6 would actually be deployed. I had an Asus router back in the days, it had primitive and unstable IPv6 support, which just grabbed a single /64 and that was it.

I then ran pfSense for some years and it was unusable for IPv6 due to everything being geared around fixed prefix. Even if my ISP had given me a fixed prefix it would have been a pain, because I got two new cable modems in that period (one due to the previous being too old, other one due to a move), so would have gotten a new prefixes anyway and would have had to rewrite all my firewall rules. Would have been major PITA each time. No such need with IPv4.

I switched to OpenWRT some years ago and it's mostly worked since. Mostly.

Android phones in the homes don't respect my DHCPv6 settings, and it took me some time to figure out it ignored my DNS settings because I had only IPv4 configured as my local DNS server. Other boxes were fine with that, but not Android, which silently used Googles DNS instead.

Running dual stack isn't ideal either, since it can lead to inconsistencies. When some things work and some things don't, and it can be difficult to figure out why because of the non-trivial interaction between IPv4 and IPv6.

It's also painful because just about everything is different. So very little of what I know of how to configure my IPv4 network carries over. It's very much not just IPv4 with more bits. I'm old enough now that I'm not terribly excited about digging around in the IPv6 technical weeds, and I haven't found a good guide or reference I can fall back on.

[JohnFen]

> Why is it so painful for you?

Mostly because of the number of machines that I have to fix up and the fact that updating each of them involves a fair bit of time. My estimation is I'm looking at at least a week's worth of work, during which my network isn't fully functioning.

It's also complicated by the number of devices I have that aren't possible to make work with IPv6 at all, which means I have to maintain some IPv4 segment and deal with making the two work together in a seamless way.

I'm also very concerned about security. I'm not confident that I know enough about how to secure an IPv6 network adequately, so I need to set aside a fair bit of time for study before I even start.

All in all, it's a large project with a lot of friction that wouldn't be as large if I didn't have to rethink everything.