Privacy

What dash-mui-charts records when you read the documentation, and what it does not.

Last updated 2026-09-05.

The short version

This is a documentation site. It records which pages are read, by roughly what kind of client, from roughly where. It does not store your network address, it does not send anything about you to a third party, and there is nothing here to log into.

What is recorded when you read a page

Every page request writes one row containing:

• when — the time of the request;
• which page — the path, and nothing about what you did on it;
• device type — desktop, mobile, tablet or crawler, derived from your

browser identification string;

• browser identification — the User-Agent your browser sends, stored

as sent;

• visitor key — a salted, one-way hash of your network address together

with that identification string. It exists so that ten pages read in one sitting count as one reader rather than ten. It cannot be turned back into your address: the salt is generated per deployment, is never committed to this project's source, and is not sent anywhere;

• approximate location — a country, and a city where the network in

front of this site supplies one. This is read from the headers that network attaches to your request. Nothing is looked up.

What is not recorded

• Your network address is not stored. It is used to compute the visitor

key described above and then discarded. (An operator running their own copy of this site can switch address storage on; on this deployment it is off, and the switch is off by default in the source.)

• No third party is told anything about you. This site makes no

outbound request about your visit — no analytics service, no geolocation service, no advertising identifier. An earlier version of this software sent each reader's address to a third-party geolocation API; that was removed on 2026-09-05 and the code that did it is gone.

• No cookies are set for analytics. There is no cross-site tracking,

no fingerprinting beyond what is described above, and no profile.

• Nothing about what you type. There are no forms on the

documentation pages that submit anything anywhere.

Crawlers

A request from an identified crawler additionally records which crawler it was, whether that identity could be verified against the ranges its operator publishes, and which document was served. This is how the site answers "did this AI crawler actually read the documentation?" — a question about machines, not about people.

How long it is kept

Rows are kept for 45 days on this host and then deleted. A daily summary — counts, not rows — is sent to Pip Install Python LLC's network dashboard so several documentation sites can be compared side by side. The summary contains no visitor keys and no addresses.

Internal traffic

Requests made by this network's own machinery — health checks, link audits, deployment tests — are identified by their User-Agent and are not recorded at all.

Contact

Questions about this policy: raise an issue on the project's repository.