This week on jakelawrence.xyz — 2026-W28

30 merges this week. The number I care about isn't 30, it's the count of changes that survived contact with the drift guards I wrote to keep old bugs from crawling back. That's the part of shipping fast that nobody photographs: the fixes that stay fixed. Written from the merge log, sent by a human.

30 changes landed this week. The honest metric isn't the count, it's how few of them were re-fixes of things that broke before.

The full log, written from the merge history and sent by a human: https://www.jakelawrence.xyz/changelog#2026-W28

30 merges this week. The interesting number isn't 30, it's how few of them were fixing something that had already been fixed once. Written straight from the merge log, sent by a human, not a digest bot.

30 changes landed on the site this week. The count is the least interesting part. What I actually track is regressions: how many of those merges were re-fixing something that quietly broke again. This week, close to none, which means the guards are earning their keep. Written from the merge log, sent by a human: https://www.jakelawrence.xyz/changelog#2026-W28

30 merges this week. The number worth bragging about isn't 30, it's how few were re-fixes of things that already broke once.

Full log, written from the merge history: https://www.jakelawrence.xyz/changelog#2026-W28

30 changes shipped this week. The honest metric is regressions, not merges: how many of these were fixing something that quietly broke again. Written from the merge log, sent by a human.