IPv6 for Programmers
summary
IPv6 uses 128-bit addresses with a default /64 subnet so hosts auto-derive interface IDs from MACs (and often use privacy extensions to randomize them). Real-world deployment is small but growing — ~1% of global traffic, ~3.5% to Google, roughly doubling each year — and dual-stack will persist. Practical guidance: stop hardcoding IPv4, use getaddrinfo and iterate candidates, bind separate v4/v6 sockets and set IPV6_V6ONLY, store addresses in binary or use Postgres inet, test and enable IPv6 end-to-end.
AI-written summary · about
chapters (14)
- 0:00Start
- 1:25Address Shortage
- 2:13Advantages of IP
- 4:34Programming
- 5:45Solution
- 8:56Language Support
- 9:27Python
- 11:03Addressing
- 13:07Comparison
- 13:55Matching
- 14:07Parsing
- 14:20Storage
- 16:27Open Source Advantage
- 16:51Open Internet
related talks
55mIPv6 for Embedded Developers used to IPv4
36mIPv6 in 2025: Quo Vadis?
48mIPv6 for Server Admins and Client Developers
28mHow and Why to Publish Your Site on IPv6
18mWorld IPv6 Day: What We Learned
23mAIN: A Blueprint for an All-IP Data Center Network
17mThe Long Road to IPv4/IPv6 Dual-stack Kubernetes
32mIPv6 / Dual-Stack in Kubernetes - Why, When, Where and How?