12 Programming Myths Devs STILL Believe in 2025! скачать в хорошем качестве

12 Programming Myths Devs STILL Believe in 2025!

On this episode of The Programming Podcast, Danny Thompson and Leon Noel unpack the biggest programming myths that confuse developers at every level. From “AI will take all dev jobs” to “DRY at all costs,” they separate hype from reality and share hard-won lessons from real teams in production. Listen to The Programming Podcast on Spotify or wherever else you listen to podcasts! https://open.spotify.com/episode/74qo... You will hear why paper Big-O is not the whole story, how cache behavior and data size impact real performance, and why map/reduce vs for loops is a wash on modern engines. We get into testing culture too: why E2E does not replace unit tests, how to use AI for test scaffolding without losing your engineering brain, and what actually improves product reliability. Danny also tackles the myths that Java is slow and GC is always bad, and both hosts talk about the cost curve where cloud is not cheaper than on-prem. The conversation closes with an “Ask Danny and Leon” mailbag on what really separates junior, mid, and senior engineers: independence, guardrails, impact, and the quality of questions you ask. If this helped, drop a comment with a myth you want us to tackle next, and subscribe for more practical, no-fluff engineering talk. Topics include: AI as a productivity tool vs one-click magic Big-O vs real-world performance and memory behavior jQuery, Deno, Bun, and the hype cycle JavaScript the language vs browser APIs map/reduce vs for loops on modern engines Unit tests, integration tests, E2E, and using AI wisely Java performance and garbage collection tuning DRY vs duplication and over-abstraction Accessibility as a default Cloud costs vs on-prem at scale Career ladder: junior, mid, senior traits Hosts: Danny Thompson (Director of Technology, This Dot Labs; Commit Your Code Conference) and Leon Noel (Managing Director of Engineering, Resilient Coders; instructor at 100Devs) Chapters 00:00 Intro and why myths still persist 00:58 Host intros and setup for “Gem City” episode 02:00 Myth 1: “AI is taking all dev jobs” 03:33 When AI image gen goes sideways and why it is a tool, not a replacement 06:03 Leon’s motion-blur trick for more believable AI images 07:08 Myth 2: Big-O vs real performance in the wild 09:52 Cache misses, allocation, data size, and why paper math can mislead 11:30 Myth 3: “jQuery is dead” and the reality in legacy estates 12:23 Deno and Bun hype vs actual employer adoption 13:32 Why jQuery still ships and what we lost chasing complexity 15:30 JavaScript the language vs DOM and host environment APIs 16:58 Myth 4: “map/reduce are slower than for loops” on modern engines 18:11 Myth 5: “E2E replaces unit tests” 20:57 When testing cultures go wrong and how to course-correct 22:46 Using AI for tests without losing critical thinking 25:03 The 80-20 way to use AI on tickets and test suites 27:01 Danny gets baited, Leon laughs 28:01 Myth 6: “Java is slow” and “GC is always bad” 30:30 Region and concurrent collectors, and why allocation patterns matter 31:02 Engine differences and mental models across stacks 31:58 Myth 7: “Everything must be DRY” vs useful duplication 33:01 Strong opinions held weekly and leaving dogma behind 34:22 How dev opinions evolve with experience 34:52 Accessibility as a default, not a later task 36:29 Myth 8: “Cloud is always cheaper than on-prem” 37:17 Real-world cost surprises and pulling workloads back 38:30 Hype cycles, Jamstack memories, and maintenance pain 41:56 On-prem done right and budget realities 43:49 Mailbag: junior vs mid vs senior, company variance 45:25 Danny’s framework for levels: guardrails, impact, and ownership 51:10 The power of high-quality questions at senior and staff levels 52:10 Leveling up from mid: own initiatives and become the firefighter 53:31 Wrap-up and sign-off