laurensdewaele / xkcd

An xkcd reader

xkcd reader

https://xkcd-reader.netlify.app/

An xkcd reader, built as an interview assignment for Embark Studios.

How it works:

1. The scraper.ts scrapes all the comics, with it's additional metadata, off of the xkcd site and saves it into a giant comics.json. It also fetches the comic image, resizes it and converts it to the .webp format.
2. The static html site generator, generator.ts, loops over the comics.json and builds placeholder <div>'s with the width and height of each resized .webp image.
3. We can't load all the images at once, so only the images that are in the viewport are being fetched. This takes place in main.js.
4. For the detail page, we pass along the comicNo on the queryparams like so: /detail.html?21. detail.js parses it, and retrieves the comic in the comics.json file and fetches the original, non-resized image.
5. For deployment there's a Github actions cron job, that runs a Netlify hook and redeploys every 4 hours to accommodate new comics.

Keep in mind that, in order to update the index.html, once needs to do it in the generator.ts file!

Issues

Comics

Comic no. 404 is an obvious, but quite hilarious prank.

Comics no. 1608 and 1663 have been omitted. They are interactive comics.

Styling

Slider

Styling the slider is non-trivial in webkit. This is due to conflicting css properties. For a vertical slider we need:

-webkit-appearance: slider-vertical;

To custom style it we need:

-webkit-appearance: none;

A workaround is to position it horizontally, transform it, and apply the calculations needed in order to reposition it correctly. See https://css-tricks.com/custom-interactive-range-inputs.

For now, we've styled it for Firefox and left the default webkit styling for what it is.