Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Good question. The difference is in scale and tolerance.

Crashes used to be localized, one app, one machine. Now a missing field in a config file can take down 8.5 million Windows systems globally. Spotify leaking 79GB of RAM isn’t a “bug,” it’s normalized waste.

The signal isn’t that bugs exist, it’s that catastrophic ones no longer trigger process change. We’ve accepted systemic failure as normal because hardware and cloud budgets hide the cost.





Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: