Privacy

This site measures how its articles are read, so I can see which ones work and where readers stop. This page lists what is recorded, why, and how to opt out.

What is recorded

• Page views. The page, the time, the address of the page that linked here (without its query string), and any via or utm_ tag in the link you followed.
• Reading progress on articles. How far down you got, which sections you reached, how long each section was on screen while you were active, and whether you scrolled back up.
• A few interactions. Share buttons pressed, links to other sites followed (only the destination's domain), the length of text you copied (never the text itself), likes, clicks that had no effect or came in a rapid burst, whether interactive demos loaded, and script errors.
• Signs of a person. Whether the page received real mouse, touch or keyboard input, and whether the browser reports that it is automated. These separate readers from bots.
• Server requests. For each page request: the browser's user agent string, the referring page, the status, and the time. This counts search engines, AI crawlers, AI assistants fetching a page for someone, and the link previews chat apps make when a link is shared.

How repeat visits are counted

A random number kept in your browser's local storage (under _r.v) tells a returning reader from a new one. It is not a cookie. It lives in this browser on this device, and clearing this site's data removes it. Pressing Like stores a second random number (_r.l) so the site remembers that this browser liked the article.

My own browsers carry a mark (_r.me, set by opening a page with ?_r=me) so my visits can be left out of the counts; nothing like it is set for anyone else.

Your IP address is not stored. The server checks whether it belongs to one of my own networks, again so my visits can be left out, then replaces it with a code that changes every day, used to count distinct visitors within a day and to slow down abuse.

What stays out

• No cookies are set for any of this.
• No third-party analytics, advertising or tracking scripts run on this site.
• The measurements stay on a server I run. They are not shared with anyone or sold.

Opting out

If your browser sends Global Privacy Control or Do Not Track, the site keeps no visitor number: your visits are still counted, but not linked to each other. Blocking JavaScript stops the reading measurements entirely; the server still sees the page request itself, as every web server does.

How long it is kept

Measurements are deleted after 13 months. Likes stay as long as the article does.

Why

Section by section reading data shows which parts of an article people read, skim or skip, which is where the writing needs work. The crawler and link preview counts show how often articles are fetched by AI tools and shared into chats, which page views alone would miss.

Questions are welcome through the contact links.

Last updated October 2026.