This project is provided for Godot Engine users and contributors to document in a complete and comprehensive manner changes that go into each release and developer preview of the engine. This project aims to largely remove the manual labor from creating and maintaining a project changelog, and provides tools to generate static information, such as CHANGELOG.md and release notes for the blog.
The benefit of this automated system is that it allows to track PRs as well as individual commits, preserving important context for the work that goes into each release. For example, it can also detect commits that were cherry-picked, and identify their original PRs.
Live website: https://godotengine.github.io/godot-interactive-changelog/
This project is written in JavaScript and is built using Node.JS. HTML and CSS are used for the presentation. The end result of the build process is completely static and can be server from any web server, no Node.JS required.
Front-end is designed in a reactive manner using industry standard Web Components
(powered by lit-element). This provides native browser support, and results in a
small overhead from the build process.
To build the project locally you need to have Node.JS installed (12.x and newer should work just fine).
This project uses GitHub's GraphQL API. To fetch live data you need to generate
a personal OAuth token.
You can supply your token to the scripts using the GRAPHQL_TOKEN environment
variable. Note, that if you don't have member access to the organization, you
may not be able to access all the information used when generating the database.
Tip
For convenience, it's recommended to store the GRAPHQL_TOKEN environment variable
in a .env file at the repository root.
- Clone or download the project.
 - From the project root run 
npm installoryarnto install dependencies. - Run 
npm run buildoryarn run buildto build the pages. - Run 
npm run publish-dboryarn run publish-dbto prepare and copy configs and data. - Serve the 
out/folder with your method of choice (e.g. using Python 3:python -m http.server 8080 -d ./out). 
rollup is used for browser packing of scripts and copying of static assets. The
data fetching script is plain JavaScript with node-fetch used to polyfill
fetch()-like API.
- Add a config file under the 
./configsfolder. 
- It must follow this format: 
<REPO_OWNER>.<REPO_NAME>.<VERSION>.json, e.g.godotengine.godot.1.2.3.json. 
- Run 
npm run compose-db version:<VERSION>oryarn run compose-db version:<VERSION>to fetch the data from GitHub. 
- This project defaults to 
godotengine/godot, but you can also customize it withowner:<REPO_OWNER>andrepo:<REPO_NAME>. - By default the repo is checked out to the 
./tempfolder, you can skip this step withskip-checkout. - GitHub API is very flaky and requests can fail from time to time due to the server-side issues; the compose script has a few tricks to mitigate that, but ultimately it can still randomly fail on bigger changelogs. Just run it again.
 
- Run 
npm run publish-dboryarn run publish-dbto prepare and copy configs and data. 
This project is provided under the MIT License.