Field note: May

May was a month of subtraction. I did less. The things I did were better for it. I'm still not sure if that's a win or a rationalization — probably both.

The big change: I audited my calendar on May 3rd and killed six recurring meetings. Not paused them, killed them. Two were status updates that could have been a Slack message. One was a "sync" with a partner I hadn't talked to about anything actionable in three months. Three were the slow-drip kind — useful once, calcified into habit. Combined, that freed up about four hours a week. I put three of those into the Tuesday writing block and kept one as slack.

The result was two long essays in May. In April, I wrote zero.

What I shipped

• "The half-life of a good tool" — the longest essay I've published on this site. Took three weeks to write, which is about two weeks longer than I thought it would take, and entirely worth it.
• Sage After Dark archive page — the thing I said I wanted in April. Seeing all the posts in a grid made the gaps obvious in a useful way.
• Trayd flag system refactor — moved our feature flags from a janky env-var pattern to a proper table with sunset dates and exposure logging. Took four days. Shipped without incident.

What I read

• The Timeless Way of Building — Christopher Alexander · The core idea (patterns as living structures, not templates) is directly applicable to how I think about product design. The fact that it's about architecture makes it read fresher than another software book.
• Working in Public — Nadia Eghbal · Third time through this year. The section on parasocial audiences maps onto newsletter building in uncomfortable ways.
• An Elegant Puzzle — Will Larson · Skimmed chapters 3–5 on engineering teams. The "org debt" framing is the most useful new lens I've added to how I read slow-moving problems.

What I noticed

The meeting I should have skipped was a product review I'd convened myself. I called it to share context with three people who didn't need it yet. The real reason I called it was anxiety — I wanted to feel like the work was progressing by talking about it. The meeting produced nothing. The work would have progressed faster if I'd spent the hour building.

The calendar audit is not a one-time thing. Within two weeks, two new recurring meetings had appeared. One I'd agreed to on a call without thinking. One materialized from a Slack thread that reached a natural resting point in "let's get everyone together." The calendar colonizes itself if you stop watching. Schedule a review. Put it on the calendar.

The archive page taught me something I didn't expect: I can see the month I stopped writing by looking at the gaps. January through March this year were thin. That's not visible in the work itself, but it's visible in the shape of the archive. The gaps are information. I want to close them.

May score: 7/10. Two essays, one good refactor, one meeting audit. Would have been an 8 if I'd gotten to 100 subscribers. I'm at 87. Close enough to feel the next one.