A charitable explanation of Sara's tweet might be that, like with jQuery, it is becoming difficult to convince new developers that React may not be necessary for their next project.

The other comparative downside of JQuery was that components started to rely on it as a shared library, which meant developers suddenly needed to do dependency management. That is (usually) a bad idea. React absolutely does have that problem as well - probably to an even worse degree than jQuery widgets ever did.

This is kind of the same concern I have with Vue to be honest. I use Vue for prototyping and will probably use it in some final apps as well. But I'm probably not ever going to use a component that someone else has written if it relies on Vue.

IMO Lodash went the right direction with this. A component or library can depend on Lodash, pull in just the functions that they use for a final build, and then nobody else in the entire dev toolchain needs to know or care. No risk of conflicting dependencies, build size stays low, etc...

I kind of wonder if the problems Vue solves for me could be solved better by a smaller, lodash-style library instead of a framework.

But jQuery wasn’t just a tool, it was also community. What that community did/perpetuated was some pretty bad and ill advised practices. Plugins for everything - the old $.sum(2,2) joke applies here. DOM for state. Even sanctioned projects like jQuery UI were guilty of some terrible, awful practices.

That said, React isn’t the new jQuery, it’s the new Backbone. Backbone lacked so much - and did so little - that it’s deficiencies needed to be made up for by “other” tooling. As a result, there was O(n) ways to “Backbone” – making backbone apps very difficult to study/build/maintain.

React is fine. The community just needs to coalesce.

Edit: typing on my phone; and am lazy

However in my experience, a couple of things were key to creating a negative view of jQuery:

1. Security and unsafe defaults, for example: evaluating remote JavaScript by default [1]; and unsafe DOM parsing by default [2]

2. Not really to blame on jQuery, but its ease of use for DOM manipulation led to programming patterns that aren't easily scalable, for both code and performance

[1] https://nvd.nist.gov/vuln/detail/CVE-2015-9251 [2] http://api.jquery.com/jQuery.parseHTML/

I cant even name a similar piece of software in web development history that is as important as jQuery.

Despite Web development as a whole, back end or front end is still no where near being good enough, it is still many times better then the old days.

I would understand if comparing VueJS with jquey. Even if they are very different, the philosophy is the same: a small elegant easy to use versatil tool.

Yet, Sara is neither of those, she's a well respected developer, who's work with CSS and SVG is world-class.

If she's using jQuery as a comparison she's doing it from a place of experience and knowledge

React is great, mostly for bigger web apps that do need a framework and more structure, such as an SPA with multiple team members working on it.

Just like jQuery is great to do some quick and simple dom manipulation.

I've tried both opposites; I've been on teams working on large apps in jQuery, and it's hell. And I've also used React to implement simple little things, and it's also hell (part of it for reasons Brad mentions in this article).

I don't have much experience with Vuejs but I've been using it instead of React recently, and it's refreshing indeed. Just like jQuery you can just import it with a script tag and start doing some dom manipulation and it makes the code much cleaner than jQuery.

And as others have said, jQuery is not bad - it's awesome. I guess it didn't age very well with the JS env evolving, but it's great.

I can usually jump into someone's react code and get what's going on pretty quickly unless they did something really gnarly with incredibly nested props. Even then, at least I still know where to drill down. With jQuery soup, you didn't get that luxury.

Being able to have an entire team know how to build components and provide them with functionality is an incredible advantage. I don't love everything about react either, but the ecosystem, somewhat enforced convention, tooling, etc - it's worth it (so far).

I love some things about Vue, but it isn't a silver bullet either. Brad seems to be pretty butt hurt about react. If you've worked on a large team without something like react, then worked on one with react, it really shines. It's a major productivity booster.

It's not perfect, but it's very good.

Well, of course he will struggle to learn React. Web design and software engineering are two different fields… Not saying he needs a master in CS to learn React, but he will surely need to learn a bit of programming and JavaScript before jumping on it.

Wikipedia seems to agree with me [1], but maybe I'm misinterpreting it:

[1] https://en.wikipedia.org/wiki/Proprietary_software

I was more into those two but jQuery made everything dead simple.

Originally, jQuery's big selling point wasn't that it handled animations, it was that it gave you a query selector, and one that could be used in any browser. jQuery predates the release of chrome, and came out a few years before document.querySelector was added to Firefox and Chrome was released.

jQuery was monumentally more useful for common actions compared to its contemporaries, such as Dojo, MooTools and Prototype.

The people for whom ripping out React or Vue leads to a blank page are those who wanted an app from the ground-up.

If anything, React is losing because it doesn't include enough out of the box. If you wanted to build a news widget, fine. But if you wanted to build an app, you have to start putting pieces together from 3rd parties (or write your own custom pieces). That's not what a framework sounds like to me, that sounds like figure-out-your-own-framework, and I think that's why Vue is winning.

Create React App was a very late response, and I still don't think it's as good as Vue's up-and-going tooling experience.

I'm sorry but this is not a problem. If you commit to React then you commit to it. You are not suppoused to change it every other year.

Imagine a big C++ Qt application. Will you ever think of changing Qt on it, and expect no huge cost? No. That's unreasonable.

So what he wants is not only a cross-browser, cross-platform framework (and much more) but also a framework that is somehow "cross-framework"!!

This is really getting me. People, such flexibility and ubiquity is NOT FREE. People this days complain about the web being to complicated, and that it is a lot of complexity to make something "simple as a webapp". I agree, it is. But that simple multimedia webapp you get today works over a distributed network, in almost every platform, from mobile to desktop to almost everywhere. This kind of ubiquity will be unthinkable in the early 90s/2000s without investing on a huge number of experienced dev teams and a huge especialized codebase(s). Today some javascript framework and markup will give you almost the same product. Yes, I think the web ecosystem can be improved. A lot. And the end result is not always 1:1 comparable at many levels with what you would get back in the day with native apps and a load of work and experience. But still.

People complain but they don't know what they are getting.

I totally agree, and in fact I would take it a little farther and say it's making functional programming mainstream on the backend as well. Many people I've worked with got their taste via React/Redux and then started adopting things like Elixir/Phoenix for the backend.

Maybe it's an overstatement, but I credit React with driving at least some of that.

You hit on the two points that makes React what it is. A pure state -> ui abstraction and the ability for large projects or teams to use it efficiently.

You can of course prove me wrong and point us to a complex vanilly JS application you wrote.

It's not something you can easily implement yourself either.

That's the main part that makes React code maintainable.

I don’t think the same could be said of your random JQuery or. Backbon based project

Is it worth the effort for a fraction of a percent of technical users whining about the need for JS? probably not.

For SSR you are sending the HTML text instead of JSON. Time ago you needed nodejs for SSR I do not know the alternatives today.

    $(".formname").html(n_input);

- `this` might not be document

- find might return multiple elements, not just one

- if nothing matches, jQuery does not throw unlike the vanilla version

I do understand it's just an example, but if you'd consider all the implicit cases the jQuery example covers, you'd end up with a lot more than a single line. I see your point, but using off examples like this don't really sell it too well.

Modern libraries such as React and Vue however play a totally different role. They introduce proper architecture to web development such as separating the view from your business logic with data binding, they can not be compared with jquery which was just a nicer way of doing the same old manual DOM manipulation.

same! Jquery is still quite useful and powerful. This is also what rails does for me; makes web development fun. Or highcharts making charts fun, or elasticsearch making search fun. To me that's a testament to elegantly solving a problem. I just don't get those fuzzies with react.

Any thoughts on https://news.ycombinator.com/item?id=17213264?

I don't disagree but will push back slightly on the appeal to authority. I think in JS land we're getting a little too carried away with celebrity developers (I often joke about "what would Dan Abramov say?" and it's only half-joke). I've been trying hard to focus on the ideas, without regard from who said it.

What does that even mean?

It's frightening, the degree to which many professional web app developers do not know the fundamentals of Javascript. Our technical interview starts out with ridiculously simple questions and gets progressively more difficult and esoteric. And a dismayingly high number of candidates don't get past the first couple of questions. And these are people with years of experience and at least some of the hot JS frameworks on their resume.

The problems with jQuery is that it's mostly a set of functions to simplify Ajax and DOM manipulation so it's not really helpful for the structure of an application.

React (and other frameworks surch as Angular, Ember and Vue), on the other hand, helps you to build reusable component which helps a lot to scale.

Also jQuery's golden age was when browser compatibility was a big issue, nowaday, most browsers behave properly.

There's no harm in learning jQuery, its interface is way more intuitive than native DOM apis. It could be a good idea to try and build a simple app with jQuery and see what are the pain points and then compare with a modern framework.

React is for building applications, especially with teams. Combined with redux and a handful of other pieces, it's very nice.

Angular, I don't get why anyone would choose it today (I've said it time and again).

jQuery and the plain DOM are retained mode APIs. You manipulate stateful objects to achieve your goals. It is ok for simple interactions, but it quickly becomes unmanageable for more complex apps.