Most metadata is powered by one `resume.json` file based on [🗂 JSON Resume](#-json-resume), and one `projects.yml` file to [define the displayed projects](#-one-data-file-to-rule-all-pages).
All displayed project content is powered by one YAML file where all the portfolio's projects are defined. The project description itself is transformed from Markdown written inside the YAML file into HTML on build time.
Gatsby automatically creates pages from each item in that file utilizing the [`{ProjectsYaml.slug}.jsx`](src/pages/{ProjectsYaml.slug}.jsx) template.
Most site metadata and social profiles are defined in [`content/resume.json`](content/resume.json) based on the [JSON Resume](https://jsonresume.org) standard and used throughout the site as a custom React hook. Additionally, a resume page is created under `/resume`.
If you want to know how, have a look at the respective components:
On build time, all my public repositories are fetched from GitHub, then filtered against the ones defined in `content/repos.yml`, sorted by the last push date, and provided via the `pageContext` of the front page.
On client-side, my current and, if known, my next physical location on a city level is fetched from my (private) [nomadlist.com](https://nomadlist.com) profile and displayed in the header.
Fetching is split up into a serverless function, a hook, and display component. Fetching is done with a serverless function as to not expose the whole profile response into the browser.
Includes a theme switcher which allows user to toggle between a light and a dark theme. Switching between them also happens automatically based on user's system preferences.
Includes a SEO component which automatically switches all required `meta` tags for search engines, Twitter Cards, and Facebook OpenGraph tags based on the browsed route/page.
The _Add to addressbook_ link in the footer automatically creates a downloadable vCard file on the client-side, based on data defined in `content/meta.yml`.
Site sends usage statistics to my own [Matomo](https://matomo.org) installation. To make this work in Gatsby, I created and open sourced a plugin, [gatsby-plugin-matomo](https://github.com/kremalicious/gatsby-plugin-matomo), which is in use on this site.
All project images live under `content/images` and are automatically attached to each project based on the inclusion of the project's `slug` in their filenames.
All project images make use of the excellent [gatsby-plugin-image](https://www.gatsbyjs.com/docs/reference/built-in-components/gatsby-plugin-image/) plugin, working in tandem with [gatsby-plugin-sharp](https://github.com/gatsbyjs/gatsby/tree/master/packages/gatsby-plugin-sharp) and [gatsby-transformer-sharp](https://github.com/gatsbyjs/gatsby/tree/master/packages/gatsby-transformer-sharp).
All together, Gatsby automatically generates all required image sizes for delivering responsible, responsive images to visitors, including lazy loading of all images. Also includes the [intersection-observer polyfill](https://github.com/w3c/IntersectionObserver) to make lazy loading work properly in Safari.
All SVG assets under `src/images/` will be converted to React components with the help of [gatsby-plugin-svgr](https://github.com/zabute/gatsby-plugin-svgr). Makes use of [SVGR](https://github.com/smooth-code/svgr) so SVG assets can be imported like so:
You can simply use [Docker](https://www.docker.com) & [Docker Compose](https://docs.docker.com/compose/) or install and run dependencies on your local system.
To add a new project, run the following command. This adds a new item to the top of the `projects.yml` file, creating the title & slug from the argument:
Every branch or Pull Request is automatically deployed by [Vercel](https://vercel.com) with their GitHub integration, where the `main` branch is automatically aliased to `matthiaskretschmann.com`. A link to a preview deployment will appear under each Pull Request.
A backup deployment is also happening to a S3 bucket, triggered by pushes to `main` and executed via GitHub Actions. The deploy command simply calls the [`scripts/deploy-s3.sh`](scripts/deploy-s3.sh) script, syncing the contents of the `public/` folder to S3:
Upon live deployment, deploy script also pings search engines. GitHub requires the following environment variables to be setup for successful deployments in the repository secrets:
All images and projects are plain ol' copyright, most displayed projects are subject to the copyright of their respective owners.
Don't care if you fork & play with it, but you're not allowed to publish anything from it as a whole without my written permission. Also please be aware, the combination of typography, colors & layout makes up my brand identity. So please don't just clone everything, but rather do a remix!
