What the long run is actually for

I solve more architecture problems on Sunday mornings than in any meeting. Not because running makes me smarter, but because it removes the option to look something up.

Without a screen, you're forced to hold the problem in memory. That constraint strips away everything incidental and leaves the actual shape of it.

I bring one question on every long run. No podcast for the first hour. By kilometer twelve I usually either have an answer or a much better question.