Hacker News Reader

Comment by zero_shift | Hacker News Reader

We tried this 20 years ago with "CSS zen" and the Bad Old Days of "semantic CSS". It failed because you cannot completely couple HTML structure and presentation.

Usually you have more complex styles and visuals than your HTML can express and trying to invent selectors / HTML heirarchies to describe them gets difficult very quickly

In addition you often need to support multiple style systems on a single web application whilst the design evolves, or else it becomes an all or nothing rollout when you change stuff.

It's common for engineers who weren't around back then to look at the abstractions that modern FE systems provide, and question if they're even necessary. That's healthy. What's not healthy is assuming they were only invented for fake or historical reasons, and telling everyone to just abandon them for commercial projects.

View on HN

You've got the horse before the cart, and completely misunderstand Semantic CSS.

The styling should simply serve to improve the content. The content should generally stand on its own without any styling.

In semantic CSS the goal is to write your HTML with as little regard to presentation as possible. Your HTML is simply structuring human readable data. At this step you give it no thought of appearance.

Styling that structured data comes second. You write the CSS to make the structured data look nice, and ideally do not touch the HTML in the process, at all. The styling is but an affectation.

This fell out of favor because people cared more about looking flashy than quality content, they didn't want to put any thought into developing the actual page and wanted to start with the design. It's a symptom of the overall disease of the modern web.

This is more achievable with modern CSS than ever before and I pray for it to make a comeback.

Note for carriage owners: “Horse before the cart” is actually the correct order.

The standard idiom is "cart before the horse"...

https://en.wikipedia.org/wiki/Cart_before_the_horse

Looking flashy is a _significant_ signal that users rely on to determine if the content is quality or not. It's a sad fact that, if your site does not ape the style consensus it'll look outdated or worse like a scam. Its the same reason used car salesmen wear suits.

You’re probably right about most users, but when I see a website that doesn’t bother with CSS at all, for me that’s a strong signal of higher quality content.

We tried all that stuff in the early 2010s. I was there. It didn't work.

People ditched element-based CSS for good reasons: it forces you to maintain increasingly complex stylesheets, because you end up having to come up with increasingly arcane rules describing how the design links back to content.

But some design isn't about content. Sometimes you need a button that's centered because it just looks better. There's no semantics or document structure you can point to, to justify that.

(So then you end up with post-hoc mental gymnastics like, "hmm, well, it's the second last button on the page in this one instance, so let's write an :n-selector that only targets the penultimate <button>...")

So having classes lets you keep the look and the meaning orthogonal. Which is especially helpful when you don't have total control over the markup (e.g. most content management systems)

I don't do a ton of web development but I occasionally need a little webview or form, and this is exactly how I write them.

That's how I was taught, you know, progressive enhancement. I didn't even know there was any other way lol

20 years ago CSS didn’t have the power it does now. There was no :has, :where, subgrid, and a whole bunch of other tools that let us effectively decouple structure from styling. Times have now changed, and it’s worth evaluating whether there are opportunities to roll back some of the changes we were previously forced to make.

Reply on news.ycombinator.com