Ask HN: Is React too big to fail? | Hacker News Reader

Ask HN: Is React too big to fail?

12 pointsmacando16 comments

In the sense "will you still find new job listings in a decade if you learn it now", then I would say highly likely. In the case its not, the path from React to the Next Big Thing™ will be very publicised and you'll be drowning in blog posts describing the differences.

Yeah, I wonder will we see the Next Big Front-end Thing™ in the following 5 to 10 years. I remember how fundamentally novel and different React felt when it started getting traction.

Now, all its competitors (Vue, Flutter, Svelte, Solid) look very React-like to me. A few minor improvements in some areas but nothing groundbreakingly different.

Elm is different but remained niche and obviously too exotic for mainstream. There's something about the pure-functional approach that doesn't sit well with most people.

Svelte is different from the rest of that list. In Svelte you don't use a virtual DOM to watch for changes, instead you compile the app and hard-code all changes.

That's a difference in implementation. You don't typically think about the virtual DOM when writing react.

React will fail under a few possible conditions:

* Browser tech evolves to not need it (WebASM and similar tech becomes mainstream)

* Browsers and the open web fail in favor of some other cross-platform standard emerging

* Browsers and the open web fail because an existing platform (like MacOS/iOS) reach 70% or greater marketshare with lower costs of development than the web.

Right now, none of these seems feasible, so I believe React will endure. I can't see any of the other JS frameworks having a strong and compelling reason to take it down. The only risks are more of a fundamental change.

Do you say that as assuming react is currently the biggest and widest adopted framework? Angular still seems to be more popular in corporations (at least in the UK) although I'd prefer react generally. I guess that might be because angular is more batteries included.

It's a pretty safe bet React is the biggest and widest adopted.

https://www.npmtrends.com/@angular/core-vs-angular-vs-react-...

I'm not disagreeing with you and if I had to learn more in depth I'd choose React. I'm just saying that big (non-tech) businesses favour Angular in my experience, same as these companies traditionally favoured Oracle/SQLServer over Postgres.

Thanks for the link

How could it fail?

A bank can fail and then you can't get money out of your checking account.

React might decline in popularity compared to a new alternative, or Facebook might decide to stop maintaining it -- but unless it gets deleted from npm life will go on.

Personally, I prefer Vue but why does it matter if it is too big to fail or not ? It is just another JS framework and I am sure there will be many more in the future. Also hearing good things about Svelte.

Your team has 1M LOC in React, and then they stop supporting it, you are in trouble, because Chrome in 2023 might not work with 2021 React.

In the real world a sudden loss of support from a big company like Facebook for dev tools is unlikely. Even Google still supports Angular (tongue in cheek)

And some companies take it as a known. For example companies who use Elm in production, probably at some level know they might have to tinker with the compiler if the founding team which is quite small (compared to other frameworks) gets hit by that so called bus.

I don't personally like it. But rather than see it "fail" I'd prefer there be many great options and less tribalism around frameworks and stacks.

I like React just fine, except for the hooks. (Pro tip: if it matters what order your functions get called you're not doing "functional programming")

I don't like many of things people build in React because (like Docker) it (1) doesn't subtract any complexity, but (2) lets people go further than they otherwise would without discipline.

For instance many React apps have multiple widget sets, some of which duplicate very simple functionality, but some of which do really amazing things. Some of them are customizable with CSS (but in different ways) other use toolkits like Emotion that require a lot of conceptual knowledge to master. Either way you have to know CSS and Emotion and the specifics of how they are used in all the widget sets used in your app to be in control.

What gets me is that React/MobX and Vue haven't had much market penetration whereas they come pretty close to solving the problem of "how to update the view when the model changes" systematically. That problem is a bear, however, in any GUI framework, many times so when asynchronous comm is the only kind of comm you have.

All frameworks will go out of fashion. React is probably going to outlast others but who knows. It's really hard to predict. Technology in the web front end is in constant flux.

What does that phrase mean as far as a framework goes?

To suddenly be replaced by something better just like happened to jQuery.

Or is it to invested in and spread across the industry that it can't fail. Similar to Java/Spring.

Reply on news.ycombinator.com