Don't put your server in a nuclear silo
summary
Deletes the production filesystem with sudo rm -rf / and rebuilds a server from scratch, reinstalls deps and repoints DNS; builds a Solr search that returns results from 1,000,000 articles in <15 ms but the startup fails for lack of monetization; chooses MongoDB then realizes most code is just forcing relational patterns so relational DB would have been fine; rejects a CEO plan to buy a decommissioned nuclear silo $800,000* to “secure” blockchain servers and uses metaphors to stop the idea.
AI-written summary · about
chapters (15)
- 0:00Welcome
- 1:19Linux admin mistakes
- 2:45Early dev mistakes
- 4:32Leading mixed teams
- 5:59Starting a social network startup
- 7:25Hype-driven dev
- 9:11Transition to consulting
- 10:55Innovative clients
- 12:21Questionable advice
- 13:48Nuclear silo blockchain
- 15:54Market fit focus
- 16:16Reinventing the wheel
- 17:01Design limits
- 18:47Work-life balance
- 20:12Closing
related talks
31mMy Server Runs Even Without Customers
32mYour manager’s greatest fear
50mStop Punching Yourself in the Face
43mFrom Developer to Development Manager
35mA Startup in Three Stages
28mBreaking Rules to Ship Products: A Beginner's Rails Journey
40mStop punching yourself in the face!
35mStop Punching Yourself in the Face