Hacker News Reader

This works unless the css classes or HTML structure change frequently.

I think e.g. Facebook does that and I’ve been unable to block annoying elements there.

Does anyone know how to make it work in that scenario?

Even worse with random CSS class names and random ids on many websites. Some of the JS dev crowd think it to be a good idea to have class names random generated. I consider it to be a kind of dark pattern by now.

Some CSS-in-JS libraries will create a hash of the underlying CSS and use that as the class name. So not quite random, and should remain consistent if the underlying CSS hasn't changed.

Yeah it's terrible. The JS framework we use hijacks the ID property with some internal ID which is quite annoying in some cases.

Reply on news.ycombinator.com