Notes that can actually be found

A note nobody can find is a note nobody wrote. For most of this year the library had a real problem underneath it: a note page was the end of the road. You could arrive at one from a search result, read it, and then have nowhere to go — no link to the subject it belonged to, no link to the other notes on the same course, nothing for a search engine to follow. Several hundred pages sat in the index with no path between them.

This release joined them up. Every subject now has a page of its own that lists what exists for it, so a note belongs somewhere rather than floating on its own. Notes and discussion threads resolve by a readable name instead of a database identifier, which means a link you paste into a group chat says what it is before anybody taps it. The old identifiers still work — an address that has been shared should not stop working because we changed our minds about it — they simply redirect to the readable one, and every page now names a single address as its own.

The other half of the release was measurement. The site carries an automated check that grades every page it publishes, and that check was quietly grading only the first two hundred pages of a sitemap that had grown past three hundred and seventy. It reported everything as passing because it had never looked at the rest. Raising the limit was a two-character change; finding it took considerably longer, and it is the reason the number on this page is a number we believe.

We also fixed the one measurable thing a visitor on a mid-range Android phone would have felt. On a throttled connection the homepage was shifting under the reader as it loaded — the footer painted first, landed in the middle of an empty screen, then got shoved down the page when the content arrived. It scored well outside the threshold Google treats as acceptable. It now scores zero.