lol no https://www.ietf.org/archive/id/draft-thain-ipv8-00.html

EDIT: I'm 99.9% sure this is a shitpost, not a serious proposal.

@soatok this is not an april fools rfc

@soatok fwiw i think the problem with ipv6 adoption is

• network admins delaying infra upgrades by 30 years
• folks treating it as “a more inconvenient ipv4”

@soatok This part sends me:

"IPv8 resolves address exhaustion as a consequence of its addressing architecture, not as a primary design goal."

@soatok we don't even have universal rollout of IPv6, why do we need another new one? 🤦

@soatok They missed April Fools' by 2 weeks.

@charlotte @soatok IPv6 would be fine if there was education about it that did not rely on "it's like IPv4 but…" because that's confusing & hard to remember for everyone.

@jackemled @soatok yeah education about it is terrible. i don’t think i was taught about it academically ever

@jackemled @soatok but really that is more due to academic computer education being broken and outdated and slow moving

@soatok people in 2080:
— We did it, we drop IPv4 off at all and move on to IPv6!
— Bro IPv10 just came out, so we can start moving on to IPv8

@soatok Oh, god, I snorted coffee out through my sinuses when I got to the JWT part

@soatok >Every manageable element in an IPv8 network is authorised via OAuth2 JWT tokens...

*closes tab*

@theorangetheme @soatok Right? What the hell? Are they serious?

@orsinov @theorangetheme No, I don't think they are

@soatok "systemd-allurnetwork"-ass proposal

@soatok I hope they accounted for ML-DSA's signature size when planning to add all these JWTs.

@sophieschmieg And I insist on them accounting for ML-DSA-87, because CNSA 2.0 :P

@sophieschmieg @soatok Suddenly reminded of why a major higher ed vendor tells people "please don't define too many role groups, our own stack won't balk at the size of your token but most will" because it stuffs all of the valid user roles into the generated JWTs for external service auth @.@

@soatok that summarizes our position as well

@ireneista @soatok Oh thank goodness, I saw it and thought the same thing but was concerned I might be too stupid. Glad to know I’m in good company.

@samir @soatok someone we know who has a bit more context said it's a serious proposal from an unserious person, which does explain it

@soatok could someone ELI5 why people hate on IPv6?

It seems very serviceable and has some cool features. We already use it and have solutions to I think most of the problems that arose over the years. It has way more future proof than IPv8 and doesn't introduce the privacy issues that hardware and firmware authenticated within your ASN introduces.

I've been asking this question over the years and nobody was able to provide an answer beyond the silly

• NAT gives you security (solution: block SYN packets from the WAN, get the same effect)
• Hexadecimal addresses are hard to remember (must be played in your head in a whiny voice, solution: use DNS)

@pq1r @soatok
It could have been backwards compatible with IPv4 but wasn't which went even worse than the Python 2->3 update.

@kirtai @pq1r idk, I like IPv6 shrug

@soatok @pq1r
So do I but the transition has been handled extremely badly.

@kirtai @soatok There can't be an organised transition, since there's just no way to force the Internet at large to transition to another protocol.

But today, finally, after 27 years since its drafting and almost 9 years of it being a proper standard, it's ubiquitous at the OS software stack level, at least for anything that's not a cheapo IoT device (which is a bit ironic since IPv6 was supposed to solve the emergence of a multitude of IoT devices).

My ISP provides IPv6 out of the box and when I check my currently established connections on this box, I only have IPv6 connections to addresses external to my local network. Pretty cool.

I think the only reason IPv6 is not adapted sooner is because we keep inventing ways to avoid doing it because of some perceived need for backward compatibility. I think that this is a self perpetuating problem - hosts won't configure for IPv6 because IPv4 still exists, and that keeps the demand for IPv4 going.