Data sources

Imported series are never merged with platform activity. A tool page shows them in a separate, muted section below the platform stats, and the trends page keeps them behind their own lens - one reading at a time, never a blend. Every imported chart carries the source it came from and links back here.

Units do not compare across sources: a download count, a survey share, and a pageview total measure different things. So each imported series is scaled to its own peak within the window on screen, and the raw value with its unit stays available in the tooltip.

• Stack Overflow developer survey - https://survey.stackoverflow.co/. License: ODbL. Attribution required. Yearly share-of-respondents per tool, aggregated per tool and never republished as a database.
• Stack Exchange data explorer - https://data.stackexchange.com/. License: Content CC BY-SA 4.0, counts are facts. Attribution to Stack Exchange. We store monthly question counts per tag and never redistribute post content.
• Wikipedia pageviews - https://wikimedia.org/api/rest_v1/. License: Counts are facts. Courtesy attribution to the Wikimedia Foundation.
• Wikidata - https://www.wikidata.org/. License: CC0. No obligation. Credited anyway. Used for tool facts such as release dates and aliases, not for time series.
• npm downloads - https://api.npmjs.org/. License: Open API. Courtesy attribution to npm.
• Hacker News mentions - https://hn.algolia.com/api. License: Open API. Courtesy attribution, via Algolia HN Search.
• Homebrew analytics - https://formulae.brew.sh/analytics/. License: Open (BSD-2 project). Courtesy attribution to Homebrew. Trailing windows only, so these are present-day snapshots rather than history.
• Google Trends - https://trends.google.com/trends/. License: Terms prohibit automated scraping. Manual exports from the Google Trends interface only, shown as "data source: Google Trends" per Google's guidelines.
• GH Archive - https://www.gharchive.org/. License: Open dataset. Courtesy attribution to GH Archive.
• PyPI downloads - https://packaging.python.org/guides/analyzing-pypi-package-downloads/. License: Open dataset. Courtesy attribution to the Python Software Foundation.
• State of JS - https://stateofjs.com/. License: Public raw data, credit requested. Attribution required. Numbers only, never their charts.
• JetBrains developer ecosystem survey - https://www.jetbrains.com/lp/devecosystem/. License: Public raw data, credit requested. Attribution required, with the source credited on every chart.
• RedMonk language rankings - https://redmonk.com/. License: Blog content. Attribution required. Rank numbers only, never a reproduction of their charts.
• Internet Archive Wayback Machine - https://web.archive.org/. License: Archive of third-party pages. Attribution required to the Internet Archive. Facts extracted from archived pages, never the pages themselves, to recover history for discontinued tools such as Atom and Brackets.

These are catalogued and permitted, but no data from them is on the site yet. They are listed so the terms are settled before anything is imported.

• Manual entry - /about/data-sources. License: Case by case. Hand-entered datapoints, each carrying the source link it came from.

TIOBE - the index is proprietary and its reuse is restricted, so nothing from it is imported, quoted, or charted here.

Stack Exchange bulk dumps are excluded too. Their 2024 terms are login-gated and restrict commercial and model-training use, so tag counts come from the public data explorer aggregates instead, and post content is never redistributed.

If a number looks wrong, or a source needs different credit, tell us through the contact form and it gets fixed. See also the about page and the copyright policy.