Skip to content

Latest commit

 

History

History
153 lines (104 loc) · 5.56 KB

README.md

File metadata and controls

153 lines (104 loc) · 5.56 KB
Raw

Website

This website is built using Docusaurus, a modern static website generator.

Installation

$ yarn

Local Development

$ yarn start

This command starts a local development server and opens up a browser window. Most changes are reflected live without having to restart the server.

Build

$ yarn build

This command generates static content into the build directory and can be served using any static contents hosting service.

Deployment

Using SSH:

$ USE_SSH=true yarn deploy

Not using SSH:

$ GIT_USER=<Your GitHub username> yarn deploy

If you are using GitHub pages for hosting, this command is a convenient way to build the website and push to the gh-pages branch.

Search

The search functionality is provided using the docusaurus-lunr-search plugin. It automatically indexes all documentation pages and the blog, it does not index the normal pages like the main page or the about page. The search field is added at each page in the page title bar at the top.

It uses the lunr search engine internally. The engine supports using wildcards or usual + or - search terms, see more details in the documentation.

⚠️ The search index is generated only in a production build, not during development! To test the search functionality locally you need to run the yarn build && yarn serve commands!

Blog

To create a new blog post run the create_blog_post script with a short title as the argument, for example:

./create_blog_post "New blog post!"

The title is used in the created file name and in the post URL so use something short and meaningful. The blog post itself then can contain a longer title specified in the title front matter.

See more details about writing blog posts in the blog.md document.

Discussions

The Giscus application is used for discussing the blog posts. Internally it uses the GitHub discussions as the backend and storage. The advantage is that we can use the existing GitHub infrastructure and user authentication. Giscus is basically just a wrapper for embedding the GitHub discussions into another page.

By default discussions are enabled below all blog posts. To disable discussion under a specific blog post set discussion: false in the blog post front matter.

For blog posts discussions it was created a special Blog discussion category to easily distinguish the blog posts discussions from others.

Permissions

The new blog posts discussions can be created by Agama developers, the other users can create them only via Giscus. All interactions from the blog post page are done through the Giscus application. That means the users need to authorize the Giscus application to post on their behalf.

If you do not want to authorize the Giscus you can comment directly at the respective existing GitHub discussion without using the Giscus proxy.

Mapping

There two ways how to find the respective GitHub discussion for a blog post. Each approach has different pros and cons so the Agama blog supports both.

Search blog post URL path

The respective discussion at GitHub is found by searching the URL path in the discussion title. E.g. for the https://agama-project.github.io/blog/2023/02/16/d-installer-0-7 blog post URL it searches for a discussion with the blog/2023/02/16/d-installer-0-7 title (or similar, see below).

Advantages:

  • Giscus can create a new discussion automatically if it does not exist yet
  • Discussion are created only when needed, if nobody comments no discussion is created

Disadvantages:

  • The first user who comments needs to authorize the Giscus application, it is not possible to comment at GitHub because the discussion does not exist yet and external users cannot create a new blog post discussion, they can only comment in an existing one
  • Slightly slower loading, Giscus first needs to run a GitHub search query to find the URL of the respective discussion
  • The GitHub search query uses a fuzzy matching, if more blog posts use a similar URL path or title then GitHub might return a wrong discussion as the search result

Explicit discussion ID

The respective GitHub discussion ID is defined directly in the blog post front matter. E.g. for the https://github.com/agama-project/agama-project.github.io/discussions/42 discussion URL set the discussion_id: 42 value in the blog post front matter.

Advantages:

  • Faster loading, no need to run a GitHub query, the discussion URL is already known
  • Avoids using a wrong discussion because of the fuzzy GitHub search (see above)
  • The first user can add a comment directly to GitHub discussion without need to authorize the Giscus application

Disadvantages:

  • The discussion needs to be created manually in advance by the Agama developers, it cannot be created automatically by Giscus or any external user later

Usage

For the old imported blog posts we should use the first approach with URL path mapping to avoid creating more than a dozen of discussions manually where very likely nobody would comment.

For new blog posts we should prefer using a fixed discussion ID to allow commenting at GitHub right from the first comment.

Of course it is possible to change the mapping later, e.g. let the Giscus create a new discussion automatically and later pin it to the blog using its ID.