Welcome back to Break, a Fallthrough aftershow! It's just Kris and Steve for this one! After brief reflections on the Gastown discussion, the episode pivots into a deep dive on semantic versioning, breaking changes, and the history of package management. Kris shares research showing most Go modules change far less code between major versions than people assume. They trace the history from CPAN and dpkg through NPM/Yarn to Go's GOPATH era, discuss supply chain security implications of MVS, and Kris teases an alternative Go toolchain project for 2026.Enjoying the aftershow? Let us know on social media! If you prefer to watch instead of just listen, head over to YouTube where you can watch this episode of Break!Thanks for tuning in and happy listening!Chapters:Prologue (00:00:00)Chapter 1: Reflecting on the Main Episode (00:00:17)Chapter 2: Data Centers, Technology & the Butlerian Jihad (00:01:22)Chapter 3: Semantic Versioning & Breaking Changes in Go (00:02:42)Chapter 4: Breaking Changes Across Language Ecosystems (00:05:04)Chapter 5: History of Package Managers (00:08:48)Chapter 6: OS vs Language Package Managers (00:11:30)Chapter 7: NPM, Yarn & Lock Files (00:13:08)Chapter 8: Go's GOPATH Era & Dependency Freedom (00:14:43)Chapter 9: Rethinking Dependency Management (00:16:37)Chapter 10: Alternative Go Toolchain (00:19:53)Epilogue (00:21:23)Socials:WebsiteBlueskyThreadsX/TwitterLinkedInInstagram

This week we're talking about Gastown! Dylan and Steve join Kris to break down the viral project that spins up hundreds of Claude Code instances to build a software factory. Steve makes the case for why this is an inevitable evolution and the conversation digs into what it actually means to treat software development as a factory floor. The panel traces the cycle from mainframes to PCs to cloud to AI, debates whether data centers are really the environmental villain, and gets into the real economics of AI pricing.As always, we've got supporter content! This week that includes the crypto rug pull scheme targeting open source maintainers, why conEdison is actually good at their job, whether AI subscriptions are just Uber-style subsidization all over again, the gambling psychology of usage-based costs, and Steve's secret project Docket. Not a supporter yet? Fix that today by heading over to https://fallthrough.fm/subscribe where you'll get not only extra content but also higher quality audio. Sign up today!If you prefer to watch this episode, you can view it on YouTube.This week's episode of Break continues the conversation. Kris and Steve dive deep into semantic versioning, the real cost of "breaking changes" in Go, and a whirlwind history of package managers from CPAN to NPM to Go modules. Watch it on YouTube or listen with your favorite podcasting app! Learn more by going to https://break.show/26.Thanks for tuning in and happy listening!Table of Contents:Prologue (00:00:00)Chapter 1: Welcome Back, Dylan (00:00:46)Chapter 2: What is Gastown? (00:06:02)Chapter 3: Where Are the AI Factory Floor Managers? (00:07:53)Chapter 4: The Industry is Cyclical (00:19:43)Chapter 5: Rug Pull as a Service [Preview] (00:23:36)Chapter 7: Energy, Grids & Con Edison [Preview] (00:24:12)Chapter 9: Mainframes Are Still Holding It Together [Preview] (00:25:00)Chapter 11: API Pricing & the Race to the Bottom [Preview] (00:25:35)Chapter 12: Is AI Pricing Just Uber All Over Again? [Preview] (00:26:12)Chapter 13: The Gambling Psychology of API Costs [Preview] (00:26:47)Chapter 15: Beads & AI Dev Tools [Preview] (00:27:23)Chapter 16: Steve's Secret Project [Preview] (00:27:52)Chapter 6: Data Centers Aren't the Villain (00:28:45)Chapter 8: The Case for Local AI (00:35:01)Chapter 10: Claude Code Usage & Hitting the Limits (00:39:55)Chapter 14: The Upgrade Treadmill (00:47:15)Chapter 17: Claude Debugging War Stories (00:56:31)Epilogue (01:01:28)Socials:WebsiteBlueskyThreadsX/TwitterLinkedInInstagram

Welcome back to Break, a Fallthrough aftershow! In this episode, Annie and Michael Hedgepeth stick around for Break. The panel kicks off with Michael's anxiety about his "distinguished engineer" rant, which shifts into a discussion about broken career ladders and why companies need to hire librarians. Michael adds wisdom on context management being the key to getting value from AI, while the group explores why staff engineers who only know code might be in trouble. Matthew shares how Oxide onboards new hires, Annie flips the script on what juniors bring to the table, and Kris questions whether software engineers know what engineering is. The episode wraps with unpopular opinions: Michael argues AI will create Michelin-star software instead of endless McDonald's apps, and Annie takes a firm stand on top sheet usage that sparks surprisingly passionate debate.Enjoying the aftershow? Let us know on social media! If you prefer to watch instead of just listen, head over to YouTube where you can watch this episode of Break!Thanks for tuning in and happy listening!Chapters:Prologue (00:00:00)Chapter 1: The "Distinguished Engineer" Problem (00:01:11)Chapter 2: Career Tracks and Skill Mismatch (00:03:08)Chapter 3: Why Companies Need Librarians (00:05:57)Chapter 4: Context Management and AI Effectiveness (00:10:07)Chapter 5: How LLMs Actually Work (00:14:21)Chapter 6: The Revolution in What Makes Engineers Valuable (00:17:45)Chapter 7: Why Staff Engineers Who Only Code Won't Make It (00:19:12)Chapter 8: Onboarding New Hires at Oxide (00:24:26)Chapter 9: The Multi-Dimensional Value of People (00:28:34)Chapter 10: Is Software Engineering Really Engineering? (00:31:48)Chapter 11: Unpopular Opinions - AI Creates Michelin, Not McDonald's (00:37:05)Chapter 12: Unpopular Opinion - Top Sheet Supremacy (00:43:00)Epilogue (00:47:46)Socials:WebsiteBlueskyThreadsX/TwitterLinkedInInstagram

Annie and Michael Hedgpeth, founders of People Work, join Kris and Matt to unpack the junior hiring crisis and what's really broken about how we grow engineers. Annie's viral blog post sparked debate about whether senior engineers have abandoned their responsibility to mentor and whether our obsession with career ladders created the problem. The conversation moves from systemic dysfunction to solutions: People Work, their local-first app that helps engineers manage professional relationships with a systems thinking approach.As always, we've got supporter content! This week that includes Kris's unconventional career path, a rant about why we have too many engineering titles, deep dives into relational intelligence and privacy concerns around workplace surveillance, and the technical architecture behind People Work: Swift frontend, Rust backend, and a custom DSL inspired by HCL. Not a supporter yet? Fix that today by heading over to https://fallthrough.fm/subscribe where you'll get not only extra content but also higher quality audio. Sign up today!If you prefer to watch this episode, you can view it on YouTube.This week's episode of Break continues the conversation. Annie and Michael stick around as the panel digs into why early career engineers rush to prove themselves, the trap of becoming "the glue person" instead of building technical depth, and why your strength as a junior is that you don't know anything yet. The panel also discusses some spicy topics like why software engineers spew logical fallacies, and the future of computing as hardware gains slow down. They round out the episode with some Unpopular Opinions. Watch it on YouTube or listen with your favorite podcasting app! Learn more by going to https://break.show/25.Thanks for tuning in and happy listening!Table of Contents:Prologue (00:00:00)Chapter 1: Meet Annie & Michael Hedgpeth (00:00:49)Chapter 2: The Junior Hiring Crisis (00:01:15)Chapter 3: AI as an Amplifier (00:10:37)Chapter 4: The Broken Apprenticeship Model (00:12:05)Chapter 5: Scaling Yourself Through Others (00:20:00)Chapter 6: Kris Never Had a Mentor [Preview] (00:25:59)Chapter 7: Too Many Titles [Preview] (00:26:14)Chapter 10: Relational Intelligence [Preview] (00:26:33)Chapter 12: Why Engineers? [Preview] (00:27:04)Chapter 13: Privacy & Safe Spaces [Preview] (00:27:36)Chapter 14: AI Strategy & On-Device [Preview] (00:28:06)Chapter 15: Swift + Rust + Crux [Preview] (00:28:34)Chapter 8: Networking That Actually Works (00:29:17)Chapter 9: What is People Work? (00:33:56)Chapter 11: The Onboarding Use Case (00:43:18)Chapter 16: Data Ownership & The DSL (00:48:16)Epilogue (00:52:25)Socials:WebsiteBlueskyThreadsX/TwitterLinkedInInstagram

Welcome back to Break, a Fallthrough aftershow! In this episode, Nick Gerace sticks around for Break. The panel compare audio engineering backgrounds, discuss AI-powered podcast workflows, and Nick shares his journey from IC to engineering manager, including using AI for management tasks. Things take a hardware turn with Nick's dual 3090 Ti NVLink setup, Matthew's case for water cooling, and the reveal of Kris's home lab. The episode wraps with Framework Desktop dreams and Mac Studio temptation.Enjoying the aftershow? Let us know on social media! If you prefer to watch instead of just listen, head over to YouTube where you watch this episode of Break!Thanks for tuning in and happy listening!Chapters:Prologue (00:00:00)Chapter 1: Audio Engineering Backgrounds (00:02:42)Chapter 2: AI-Powered Podcast Production (00:05:26)Chapter 3: Government Efficiency (Some Are Good, Actually) (00:07:11)Chapter 4: IC to Manager: Nick's Journey (00:12:57)Chapter 5: Using AI for Management Work (00:18:57)Chapter 6: 3090 TIs & NVLink Adventures (00:22:31)Chapter 7: The Case for Water Cooling (00:24:30)Chapter 8: Kris's Home Lab (Pure Insanity) (00:33:09)Chapter 9: Framework Desktop & Thin Clients (00:35:56)Chapter 10: Mac Studio Temptation (00:40:45)Epilogue (00:42:43)Socials:WebsiteBlueskyThreadsX/TwitterLinkedInInstagram