Hacker News Reader

Comment by EstanislaoStan | Hacker News Reader

From the code and a quick test it seems relatively resilient as it saves percentage of current viewport scroll, not absolute pixels.

View on HN

As discussed in the essay there are still some potential issues with dynamic pages. [1] Horizontal resizes might change the location of certain HTML elements, if the website is designed to be "responsive". This is a problem that I still need to address. Another issue arises with pages that can be dynamically appended (e.g., through comments) towards the bottom. In this case, simply making a calculation based on the previous percentage where the user marked the page fails. This is yet another problem that I need to address.

However, the current jumping method seemed to be working fine for most of the sites I use, so I decided to release the app first and improve it along the way based on the traction it gets.

1. https://www.rugu.dev/en/blog/mark-scroll-positions/#3-decidi...

In that case, what happens when you resize your browser window to be shorter or taller?

You are on point, the extension starts to slightly malfunction if the user starts to resize the browser window vertically. This is because of the assumption that `(window.pageYOffset + window.innerHeight) / document.body.scrollHeight` will approximately equal 1 when the user has scrolled to the bottom of the page (which apparently, is not necessarily the case).

Will address this issue ASAP. [1]

Thanks for finding an edge case which I did not notice :)

1. https://github.com/kugurerdem/mark-scroll-positions/issues/1

Reply on news.ycombinator.com