← Back to home

Usage latency

I’m thinking more about software performance. Mostly about user-facing client applications.

Here’s something to chew on: it’s better for an app to achieve user’s goal in 3 steps that are 100ms each than in 10 steps that are 10ms each. While the second app is ten times faster in every interaction, it requires more cognitive load from a user, and user themselves will introduce more latency than a computation will.

Last edited on Jul 25, 2019