New blog post! I recently did a deep dive on "Class E" space for a talk at RIPE, and figured I should write up the talk in a more searchable fashion :) https://blog.benjojo.co.uk/post/class-e-addresses-in-the-real-world
Reclaiming IPv4 Class E’s 240.0.0.0/4
antiphase@cosocial.c..
replied 27 May 2024 16:58 +0000
in reply to: https://benjojo.co.uk/u/benjojo/h/5L2sJ2C6f963Tmr1N9
@benjojo Does 240/4 not contain 15 usable /8s (240/8-254/8 inclusive)?
That's a lot easier to divide between 5 RIRs than 8 would be, not that it makes it any likelier to happen.
benjojo
replied 27 May 2024 17:05 +0000
in reply to: https://cosocial.ca/users/antiphase/statuses/112514111891928571
@antiphase good catch! but that last 255/8 (or nearly a full /8) would be too tempting that I'm sure some fight would break out if it ever came to be
famfo@chaos.social
replied 27 May 2024 14:09 +0000
in reply to: https://benjojo.co.uk/u/benjojo/h/5L2sJ2C6f963Tmr1N9
benjojo
replied 27 May 2024 15:55 +0000
in reply to: https://mstdn.social/users/HopelessDemigod/statuses/112513684551612797
@HopelessDemigod An age old question! I think the core answer is "lack of customer demand", networks are there to provide services to their customers in exchange for money. IPv6 is rarely a "big ticket feature" (side story: I know when I worked at Cloudflare that we forced many ISPs to establish IPv6 peering/support in order to have CF PoPs in their network) My suspicion is that as time goes on, IPv4 performance will get more crummy as more and more CGNAT is added, and at that point adding IPv6 support to your content (aka websites) is a obvious pick. The eyeball side will add IPv6 to take the load off their expensive CGNAT boxes. "We" are seeing some of this playing out already, But 90,000+ different networks have to do it before it makes sense, and many of those networks are not growing and so see no pressure to deploy CGNAT, and thus IPv6.