The fundamentals pendulum keeps swinging
Every few years, the web development community swings back toward the basics. A growing pile of recent posts, podcasts, and essays suggests we are in the middle of one of those swings right now, centered on the value of HTML, CSS, and vanilla JavaScript skills alongside a renewed interest in doing more without build tooling.
Fundamentals make a comeback
The argument that developers should invest in the native web platform rather than any particular framework has been made repeatedly across different voices. Dave Rupert and Chris Coyier on the ShopTalk podcast discussed how the industry had shifted hard toward all-in JavaScript frameworks, leaving people who were strong at HTML, CSS, and JavaScript feeling disenfranchised. The market appears to be correcting that, they suggested, as developers who understand how a line of CSS can replace thousands of lines of JavaScript are once again valuable.
Alex Russell, in the final installment of a four-part series on the cost of modern JavaScript frameworks, made an even more direct claim:
Never, ever hire for JavaScript framework skills. Instead, interview and hire only for fundamentals like web standards, accessibility, modern CSS, semantic HTML, and Web Components.
That advice becomes doubly important, he said, for systems built on a framework.
HTML gets closer look
The first State of HTML survey marked a shift in attention toward the language that underpins every web page. As discussed on the “Just a Spec” podcast, the survey was partly designed by Lea Verou and covered far more than pure HTML, reaching into accessibility and web components. Its conclusion pointed to a real need for more extensibility of HTML.
Part of the renewed interest, the podcast observed, is that developers who come from a React-only background are often surprised by what the platform already provides. Standard HTML forms that submit data without any JavaScript, or web components that extend functionality cleanly, can be revelations for people who never saw them before.
Focus on the native platform
Adrian Roselli, in his semi-annual call to hire and learn web standards, described a pattern that repeats itself: developers tire of a technology often marketed as an “HTML killer,” circle back to a focus on the native platform, and then inevitably try to reinvent that platform poorly. The reason companies resist making deep HTML, CSS, ARIA, and SVG knowledge core requirements, he argued, is commoditization, partly fueled by framework developers looking down on the basics.
Heydon Pickering, in a series on HTML elements that starts with the humble anchor tag, made a point about link semantics that deserves repetition:
Web developers and content editors, the world over, make the mistake of not making text that describes a link actually go inside that link. This is colossally unfortunate, given it's the main thing to get right when writing hypertext.
The AI caveat
Chris Ferdinandi poured some cold water on the idea that AI tools let developers write code they could not otherwise produce:
If you don't know how to write some code, you probably shouldn't use code that Chat GPT et al write for you.
His case: StackOverflow provides opportunities to learn and understand code in a way that AI-generated code cannot, making it a more productive learning tool.
Frontend Masters echoed the fundamentals-first approach for early-career developers in 2024. The knowledge underlying the web is durable and transferable, it argued:
The point of the fundamentals is how long-lasting and transferable the knowledge is. It will serve you well no matter what other technologies a job might have you using, or when the abstractions over them change.
Building without the build step
A parallel thread has emerged around deployment workflows, with developers exploring what it means to ship sites without heavy build chains. Brad Frost described the experience as “extraordinarily liberating,” noting that while there are some ergonomic inefficiencies, they come out in the wash — especially compared with debugging a broken build or dependency hell.
Max Böck, responding to Brad, sketched a measured outlook: production-grade development is not quite ready to ditch build tools entirely given performance tradeoffs and other complexity issues, but for smaller sites and side projects, the buildless route is worth testing. Jeremy Keith picked up the idea with the indie web community’s “Manual ’til it hurts” approach, which reduces to a two-step process:
- Start doing what you need to do by hand.
- When that becomes unworkable, introduce some kind of automation.
Keith’s observation was that people often never reach the second step — and that premature optimization is not the root of all evil, but it is premature.
The converged message from this batch of pieces is consistent: fundamentals are transferable in ways that framework churn is not, and many of the conveniences developers reach for automatically can obscure the power of the platform underneath.
