Some years ago, in 2022, I was at the office of my employer of the time. I was seated close to the Tech Lead of the team, he turned to me and say:
— Did you see the new React killer? I’m reading a blog post saying this will replace it.
— which tool?
— qwik. — he answered, laughing.
We are at the end of 2025, React still has not been replaced by any of the “React Killers”, Qwik is still not a relevant tool for the industry in terms of market usage.
So, what happened?
The frontend myth
We had this joke of “a new framework JS every day” and the bad part is that people took the joke seriously. Yeah, years ago we had a vivid ecosystem, with people experimenting and creating projects all the time. But there is a big difference between number and relevance.
When you look at the big picture, you see that little things changed over time. Eight years ago when I started in the Frontend area, the big three tools for Frontend were React, vue and angular. It’s 2025 and it keeps the same big three.
For sure, we had new tools such as Svelte, Solid and Qwik, but the truth is: any of the new ones are relevant in terms of jobs and market usage.
Even influencing the community thought leaders, with ideas such as signals, resumability, shipping less javascript to the client, etc, in terms of numbers, there are very few reasons to really study and invest time in these tools.
The good ideas will be gotten by the big 3 ones, especially Vue, that is keen to implement good solutions from alternative tools. They implemented hooks years ago, now they are investing in integrating signals to their system, a community that does not have the fear of admitting the good ideas from competing projects.
But this is the thing that is wrong with fatigue. The discourse is: it’s too many things to learn all the time, but they don’t say that if we filter the really important things, the number goes down.
Signals in solid was created almost a decade ago, React server components are available in beta since 2020, even the new generation with tools such as Svelte, has more than 8 years of existence.
Learning
But even with these tools that are important, you don’t need to learn in-depth. Learn can be an abstract word, but people divide the learning process in levels, for example, I divide in 3 levels:
- 1: Understand the tool, what problems it solves, why it was created, the basic API.
- 2: Learn how to use the tool, actively and be ready to deliver production level code with it.
- 3: Deep dive in how it works internally, the evolution during time, how to implement libs or hacks with the tool, and solve complex problems.
The problem is that some think they need to reach level 2 for all tools, when it’s not the truth, level 1 for most part is more than necessary. For example, why would you learn a tool that has small expressions in the market if your goal is to land a job?
Why study Qwik if you can learn React, Vue, Angular, which really has job openings and are really used by the companies? I’m not arguing against study. I study things that are not popular in the market, but because I like to do it, not because they will help me in the market.
But, if the current number of things to learn is a problem for you, why study things that will not help you to reach your goals? It’s a question of priority, focus on the important things, skip the ones that are not. The “fatigue” will be lower, you will have more free time to worry about other things.
If you are a decision maker or just want to be updated with the latest things, level 1 is enough to understand what a tool can do and the problem it solves. If you really will use it, then you go to level 2 or 3 and really learn how to use the tool. The slow real evolution Sometimes, when you are too deep in the tech bubble in social networks, things seem to pass faster and faster, even faster than real life. That makes sense: we are following the library and frameworks creators, and influencers that make money talking about tech.
These discussions will take years to make real changes in the industry, but we see that even before it was created, in twitter/X/Bluesky, etc. And, of course, for the content creators, to use the urgency trigger is very profitable. They need you to think everything is changing everyday, this makes sure you will click on the last video, buy the last course, keep informed with the latest releases and dramas.
Just to give some examples:
- React router v7/Tanstack router with loaders: then you go to real projects still using React Router v4, v5, without any plan to update.
- CSS-in-JS is dead: then you see multiple projects relying on these strategies, even new ones using Material-UI, Chakra-UI, etc, and legacy ones with styled-components.
- Typescript is mandatory: the you see multiple projects with vanilla javascript (I worked in 2 projects last year in pure javascript, production level web apps in a global big corporation)
- React 19??: people are still using React 16, delaying upgrades due other prioritary tasks.
- Vitest is the hype: Jest been gigantic and has a usage rate until today.
- New state management libs: people still use old redux, or use useReducer with Context that has similar experience.
This is just a bunch of common cases, but we can extend to other frameworks and ecosystems. So this is the thing: javascript fatigue is not real anymore, never was, they just want you to believe it exists.