Send us Fan Mail

What if the biggest problem in networking is that we’re solving the wrong one?

In this episode of The Art of Network Engineering, Andy and Lexie sit down with Radia Perlman, one of the most influential figures in networking history and the inventor of Spanning Tree Protocol.

This conversation goes far beyond protocols and configurations. Radia shares how networking evolved, through constraints, tradeoffs, and human decisions, and why so much of what we learn today is incomplete without understanding the problems those technologies were trying to solve.

You’ll hear:

• Why Ethernet was never designed to be used the way we use it today
• How Spanning Tree came to be, and why it wasn’t meant to be permanent
• The blurred lines between Layer 2 and Layer 3
• The real reason BGP exists (and its limitations)
• Why engineers often jump into solutions before understanding the problem
• What people get wrong about quantum computing and AI

This is one of those conversations that doesn’t just teach you what networking is, it changes how you think about it.

About AONE
The Art of Network Engineering explores the human side of IT—real stories, real lessons, and practical insights from people shaping the industry.

This episode has been sponsored by Meter.

Go to meter.com/aone to book a demo now!

Support the show

Find everything AONE right here: https://linktr.ee/artofneteng

Send a text

In this episode, Andy sits down with Gregory Grimes to unpack the world of Wi-Fi 7 and what it means for network engineers.

If wireless has ever felt like magic compared to the predictability of route/switch, this conversation is for you. Andy and Greg walk through the evolution of wireless networking, from the early days of 802.11 to the latest innovations in Wi-Fi 7, including wider channels, better spectrum use, resource units, and multi-link operation (MLO).

They also explore the real-world question every engineer asks: who actually needs Wi-Fi 7? Is it a game changer for the average home user, or does it really shine in high-density and high-performance environments like classrooms, auditoriums, healthcare, and immersive AR/VR use cases?

Along the way, they translate complex wireless concepts into practical networking language that route/switch engineers can relate to, making this a great episode for anyone who wants to better understand modern wireless without needing a CWNA-level deep dive.

In this episode:

• A quick history of Wi-Fi and the 802.11 standard
• Why wireless feels so different from wired networking
• How contention, collisions, and airtime shape wireless performance
• What OFDMA and resource units actually do
• What makes Wi-Fi 7 different from Wi-Fi 6/6E
• How MLO changes the wireless conversation
• Why deterministic wireless matters
• Where Wi-Fi 7 fits in the enterprise
• When it makes sense to upgrade — and when it doesn’t

The episode also closes with a great reminder that networking is about more than protocols and throughput. Greg shares why the Art of Network Engineering community has mattered to him from the beginning, and why finding your people in this industry makes all the difference.

This episode has been sponsored by Meter.

Go to meter.com/aone to book a demo now!

Support the show

Find everything AONE right here: https://linktr.ee/artofneteng

Send a text

“AI Won’t Replace You, But Someone Using AI Might”

AI is everywhere; stickers, marketing, hype. Network engineers are understandably skeptical.

In this episode, Andy Lapteff is joined by longtime friend of the show John Capobianco (now Head of AI & DevRel at Itencho) and Mike Bushong for a practical, optimistic “ABCs of AI” discussion designed for working network engineers.

We start with a blunt reality: automation adoption is still low, and the old “automate or die” narrative hasn’t helped. Then we pivot into what’s changed: modern models are strong enough to be useful, but only if you stop treating them like a search bar and start connecting them to real tooling and real data.

John explains the core building blocks—LLMs, RAG, agentic workflows, and especially Model Context Protocol (MCP)—and why MCP may be the protocol that finally makes AI feel operationally real.

Finally, we land on a concrete “Hello World” for neteng: connect an AI client to a source of truth like NetBox or Nautobot (in a sandbox), start with read-only workflows (logs, config deltas, compliance), and build from there—safely.

If you’ve been curious but overwhelmed, this is your on-ramp.

This episode has been sponsored by Meter.

Go to meter.com/aone to book a demo now!

Support the show

Find everything AONE right here: https://linktr.ee/artofneteng