> Minified code requires the complexity of a transpiler and makes the site less open for inspection, learning, accessibility, and archive.
The fun of "early" websites was the fact you could just read the source of something you thought was neat and then copy and paste it to your own website. Now you also got this nice animation.
Neat lacks a header and navigation; this by design, and may be enough for simple sites. If you want more capability, Pure.css is good to try too https://purecss.io/
... by it being on a sample page for the framework - the 'simple shop' samples - I thought the 'widget' was just a placeholder, with an AI generated image, perhaps.-
Ah this project name brought back memories… https://github.com/thoughtbot/neat + bourbon.io was my Sass toolkit of choice for many years back in the early 2010s. For a moment I thought it was getting a pure-CSS revival now that CSS can accomplish most grid/layout stuff natively.
More or less agree. Other comments talking about how we used to be able read source. My old school behavior is that I’ve built up my own CSS framework that I start any new site with. Nothing ambitious, just reflecting the way I want my sites to be quick to code and responsive. It is satisfying to make your tools.
Granted, this is for my own small static sites, and I’ll still use Tailwind though I disagree with its philosophy, just because it’s good for prototyping (and then it sticks around) or is the basis for some other small theme I want to build on (like DaisyUI).
Sharing CSS projects like this isn't just about getting a bunch of people to use it. Sharing ideas and patterns can be very helpful, especially to someone newer to CSS. You may pick up a minimal CSS stylesheet to get started and pretty quickly digging into it to see how it's actually built.
There are also interesting patterns that can be learned from and adapted to your own code. PicoCSS[1] had a new release recently and there's some really interesting stuff in there. The biggest standout to me is how forms are handled and leverage semantic HTML ideas to automatically add form validation styles.
I love PicoCSS, but they go a bit beyond semanticism to build some of their components. For example 'article' becomes a card. Dropdowns use details/summary. Dialog close buttons use 'button rel="prev"' which is... maybe, I guess? If you consider a modal as some sort of shallow navigation...
Anyway, they make relatively few compromises on semanticity, and as a result, they can deliver a very polished-looking final product with more features than a purely semantic framework could deliver.
Again, I love it, just sometimes, you have to hold your nose a bit.
Yep, totally agree here. I ran into exactly those same use cases, wanting a standard summary/details and wanting to wrap a blog post in an <article> element. I've been chewing on whether scoping the card style only to articles that are children of a list item would make sense, could be worth an issue or PR to get their take on it.
I've definitely learned a few interesting tricks poking around in their CSS though. I've been coding professionally for 15 years and doing mostly web for around 6 years. I'm always interested to see these minimal frameworks popping up, I feel like I learn at least a few nest tricks from each.
The fun of "early" websites was the fact you could just read the source of something you thought was neat and then copy and paste it to your own website. Now you also got this nice animation.