The relevant few lines are these: if ("IntersectionObserver" in window) { const observer = new IntersectionObserver(entries => { entrie

The relevant few lines are these: if ("IntersectionObserver" in window) { const observer = new IntersectionObserver(entries => { entries.forEach(entry => { entry.target.drawer.set_visible(entry.isIntersecting); }) }, { rootMargin: "300px" }) all_containers.forEach(container => observer.observe(container)); } else { all_containers.forEach(container => container.drawer.set_visible(true)); } You can read more about it here: https://developer.mozilla.org/en-US/docs/Web/API/Intersection_Observer_API https://developer.mozilla.org/en-US/docs/Web/API/Intersectio... As a side note, all of the code is very pleasant to read. It's interesting that it's seemingly (almost) all hand-written, given that there are commented-out rules in the CSS, for example. I'm not sure what's going on with the HTML though, probably some custom generator.