Doesn’t it also use case and esacand a few others? Such a weird language^^

That’s true, but the syntax is different then. In this blog post I cover fish, and I didn’t intent to say that this cannot be done in other shells (and I think I never said so).

Sure, web applications have different requirements and might warrant the use of more JavaScript than a website does. But one of the biggest problems nowadays IMO is that many developers choose these fancy technologies also for websites, just because they like them, without thinking too much about how that affects the user, and it does so in a mostly negative way. If you are building a website for the local bakery HTML and CSS backed by any CMS probably suffice, and there is no need to add the complexity of client-side JavaScript and SSR (or whatever) to it.

I think the point of the article (and I agree to that) is that “modern” websites (i.e. use heavy javascript frameworks) are having real issues that websites being built without loads of client-side javascript do not have. I guess some websites built in 2005 are performing better and are more accessible than websites being built today.

I have even mentioned stacking contexts in the article, and the thing is that they are not only introduce with z-index, which makes them even more complex :-/ So yeah, it certainly helps if you understand them, but I think it does not make the problem less complex.

I mean it is not really inline styles, with inline styles only it is e.g. not possible to implement a hover style AFAIK. I think the inventor has written a blog post explaining the steps, is that what you are referring to? I also read that, and it kinda makes sense, but basically giving up on development tools to work properly is kind of a high trade IMO.

I would also be interested in seeing a performance benchmark. As the article says, gzip will probably make the difference in terms of network traffic negligible, but it would be interesting to see the impact it has on parsing HTML.