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.
This hasn’t been true since shortly after the internet became commercially used. Microsoft rollouts have been crashing machines worldwide since the 2000s at least.
One of the main differences now is scale. You’d have to work much harder to show that quality is actually getting worse.
Comments
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.
This hasn’t been true since shortly after the internet became commercially used. Microsoft rollouts have been crashing machines worldwide since the 2000s at least.
One of the main differences now is scale. You’d have to work much harder to show that quality is actually getting worse.