Tailwind is such a terrible option to teach newcomers. It’s basically a vast CSS dialect, at this point it would be much more valuable to learn CSS itself.
With Tailwind, you're not caught up in understanding the relationship between CSS and HTML which includes challenges like inheritance. Another benefit of Tailwind for new learners is the exhaustive documentation which captures all of the rules and their relationship in one place: it's hard to overstate the value of being able to look at a single page to fully understand how you can pad an element.
Certainly, there's limits to what you can achieve with Tailwind, but given it does expose a learner to the key principles of styling, it can then lead into an expanded learning of CSS if a user is interested in learning more -- you don't have to unlearn anything from Tailwind.
I can absolutely understand why someone would dislike Tailwind, especially those with pre-existing knowledge of CSS for whom this is a totally different way of styling, but to describe it as problematic for newcomers seems ignorant of where the real challenges of learning CSS are.
Can you provide more insight into the specifics of the issue? I am happy to try and diagnose the issue -- even if you've moved on from Tailwind, you might get value out of understanding the issue! A GitHub repository link would be ideal, or some code in a Gist :)
My gut feeling is that the issue you're describing has nothing to do with Tailwind. One of the challenges of learning any library or technology: you can experience an issue outside of the bounds of your understanding and misattribute the issue to the tool in question, even after decade(s) of experience, this'll still happen!
I’m sorry but to really learn CSS, you need to get this. The fact the TW abstracts this away may be easier to learn, but it does the learner a huge disservice if the don’t get what’s going on. Learning from first principles is always the best approach, even if the learning curve is slightly steeper, because it reduces a lot of mystery down the line when something isn’t behaving as expected.
I, personally, started writing HTML and CSS long before any of the tools we are discussing existed, CSS Zen Garden was state of the art back when I was learning… and nowadays I am very thankful that I don’t need to keep all of the esoteric aspects of CSS in my head when styling my HTML.
The most effective method for learning is the method that keeps the learner interested, and for most people, that’s the method that gets them to results the quickest. If you want to start learning how to make webpages by doing a deep dive on the DOM and WebKit, great, but that’s not the path for most people.
I'm a hardliner against CSS frameworks. Any time you learn a CSS framework, you're picking up a tool that is going to break or be made obsolete in a couple years. Just learn CSS right the first time and you're good forever.
That's why CSS frameworks are so popular. I can install bootstrap and make something not terrible pretty quickly. As someone new to programming to do the same in raw HTML and CSS is much much harder.
CSS is like global mutable state — it is too flexible to properly work when multiple people work together on it. Sure, I’m never against learning something from fundamentals, but I am on the opinion that it is increasingly the wrong default.
I would argue there's something to that.
Maybe hello world level Basic or Pascal or beginner Python first, but then assembly before C or Rust or even JavaScript. Something's been lost for the set of devs who don't know what higher languages are doing for them under the hood.
I'd similarly agree with ricardobeat that CSS frameworks should come after CSS, and JS frameworks after JS. Rapidly ascending abstractions is great, if you take long enough to understand what they're built with, and some things will always be solved more simply by dropping back down a layer or three.
* lvl102's comment was deleted: https://news.ycombinator.com/item?id=34764573
People that never go low level usually don't feel it in their guts how computers work. Which is OK most of the time, but it bites you back often enough that it's annoying having to explain the actual fundamentals when your colleagues haven't had that experience
Though to add a more relevant point, I read a great saying on HN that a good developer knows their stack. A great one knows a layer beneath as well. (I usually butcher up quotes, sorry. It may have been more layers actually)