Databricks renamed it.

A single static page that answers one question: "What happened to the thing Databricks used to call X?" It lists Databricks product and feature renames, deprecations, and new features - sourced, dated, and searchable - dressed up as the Databricks console.

rebricked = renamed or deprecated.

Not affiliated with Databricks. The console chrome is an homage. This site is for educational purposes only.

What it does

• Look up an old name and see what it became - full lineage, with dates (e.g. Delta Live Tables → Lakeflow Declarative Pipelines).
• Three kinds of entry, each held to the same bar of real, sourced changes only:
  • Renames - a new name for the same thing. The name in use now shows as Active (green); each superseded former name shows as Renamed (slate).
  • Deprecations - retired or replaced, pointing at the successor (Deprecated, amber).
  • New features - genuinely new capabilities worth tracking (Active, green).
• Filter and search - a status filter (Active / Renamed / Deprecated), keyed on the badge each card shows, plus category chips, a console-style sidebar, and instant search.
• Every entry has an official source (Databricks or Microsoft Learn docs) and a verified date.

The fun bits

• Quiz mode - Associate ("guess the current name," 5 questions) or Professional ("name the feature from its definition," 10 questions), with score + streak and a results screen. Share your score on LinkedIn via a link that encodes the result, so whoever opens it gets a "beat this score" challenge banner.
• Deep links - #<entry-id> opens a single entry; ?q=<term> reflects the search box. Each card can copy its own link or a tidy blurb for pasting into Slack.
• Year timeline - a stacked bar chart of changes per year, colour-coded by status (Active / Renamed / Deprecated) and synced to the filter; click a bar to filter by year.
• "On this month" spotlight, a random-entry roulette (with brick confetti 🧱), a light/dark theme toggle, and a deadpan "days since the last change" counter.

Running locally

No framework, no backend, no JS toolchain - it's a static site, and the whole site lives in www/ (that folder is what GitHub Pages publishes). The only build step assembles the data: entries are authored one YAML file per entry in kb/databricks/, and www/databricks.features.json is generated from them. The page fetches that JSON, so it must be served over http (opening the file directly is blocked by the browser):

pip install pyyaml                    # one-time: the only dependency, and only for the build
python scripts/build_features.py      # kb/databricks/*.yaml -> www/databricks.features.json
python -m http.server 8777 -d www

Then open http://localhost:8777/.

Layout

File	What it is
www/	The deployed site root - everything GitHub Pages publishes.
kb/databricks/	The data. Source of truth. One YAML file per entry, named <id>.yaml - a rename, deprecation, or feature.
www/databricks.features.json	Generated from kb/ by scripts/build_features.py (gitignored). What the page fetches at runtime; don't hand-edit.
www/index.html	App shell: Databricks-style sidebar rail + content area.
www/app.js	Vanilla JS (single IIFE, no deps). Fetches the data, renders everything, wires search / filters / quiz / roulette / theme.
www/styles.css	All styling. CSS variables; light default, data-theme="dark" toggle.
scripts/build_features.py	Assembles kb/<vendor>/*.yaml into www/<vendor>.features.json. Run before validating or previewing; CI runs it before deploy.
scripts/validate.py	Schema / format gate for the built databricks.features.json.
docs/	The documentation, organised with Diátaxis: tutorials, how-to guides, reference, explanation.
CONTRIBUTING.md	The entry schema and field rules.
AGENTS.md	Guidance for AI agents (and humans) working in the repo.

Documentation

docs/ is the full documentation set, organised with Diátaxis:

• Tutorials - start here: get the site running, add your first entry, publish your first guide.
• How-to guides - one task per page: add a rename, add a deprecation, insert a name into a rename chain, check citations, fix a failing build.
• Reference - the entry schema, guide front matter, every script and its flags, every generated file, the frontend contracts, the tracked analytics events.
• Explanation - why status is the sole discriminator, why the data is one file per entry, why there is no framework, why citation rot has its own checker.

Analytics & privacy

Visits are counted with GoatCounter - cookieless, no personal data, so no consent banner is required. A few anonymous custom events (filter toggles, quiz opens, searches, shares) help show what people use. LinkedIn share links carry UTM tags so shared traffic is attributed. Everything is guarded: if the analytics script is blocked or absent, the app behaves identically.

Authors

• Ilya Aniskovets (LinkedIn) - original idea and co-author.
• Maksim Pachkouski (LinkedIn) - co-author.

The ideas the site is built on came from both of them; the project itself started with Ilya.

Contributing

Spotted an error, an out-of-date name, or a change we're missing? Contributions welcome: https://github.com/aig/rebricked.

The one rule: real, sourced changes only. Never be confidently wrong. Every entry needs an official source and a verified date; if you can't verify a claim against a live doc, flag it rather than adding it. Before opening a PR, run the schema gate (CI runs the same one):

python scripts/validate.py

See CONTRIBUTING.md for the full field rules.

Disclaimer

This site is for educational purposes only and is not affiliated with, endorsed by, or sponsored by Databricks. Product names and trademarks belong to their respective owners; the console-style chrome is an homage. Every entry is sourced and dated, but - given the subject - may already be out of date.