Live data from Hacker News

Why is Ember fading away?

medium.com

11–20 of 81 posts

Re: Why is Ember fading away?

#11
I am a junior frontend dev working on my first dev job and both the main web apps (customer facing web app and internal tools web app) are in Ember.

I like it, I could do real, valuable (if small) contributions within a month. I am 9 months in it now and I can handle significantly more complex features.

As it was explained to me, this was part of the decision of choosing Ember.

I will try to list the reasons for us:

- Timing: it was a time when the Angular broken upgrade from 1 to 2 was happening and React was starting to get traction.

- Ember was a solid and mature framework, stable enough to use on a startup with product-market fit entering the growth phase.

- There were clear signs that the framework was not going to be abandoned by its developers. Not only the maturity, but its future seemed solid. In retrospect, this was right. Ember keeps staying as a modern framework (even if not hyped that way).

- Its "conventions over configuration" approach was very important as we hadn't enough experienced software developers in the team. It made it welcoming to junior developers (like myself) that could not properly anticipate all the possible outcomes of early configuration and architectural decisions. React is very tricky in this aspect. I believe a lot of React applications being created these days by not so experience engineers will cause a lot of problems in the future.

The downside is mostly human capital. Hard to find Ember developers these days. But not such a big deal, as our hiring policy focus on more fundamental frontend skills rather than framework knowledge, so we hired good frontend developers that first got in touch with Ember here. And it is fairly easy to pick up Ember for junior and senior developers both.

That said, talking to some senior developers here, it appears that, were they doing that decision today, they would probably go with React. Being easier to attract new hires is a big deal of that decision, as we are growing a lot our development team and if we could offer React options we would certainly attract an order of magnitude more candidates.

Also, React is more mature and we have more developers that are experienced enough to make the right calls on its architecture and configuration.

But, my personal opinion, is that Ember would still be the right choice for our situation today. Something solid, with a large enough range of problems already solved, welcoming to new developers, easy enough to be productive from start.

Re: Why is Ember fading away?

#12
I think it's the natural evolution of frameworks. About 5 years ago, we had the Cambrian explosion of front-end frameworks -- Angular, React, Ember, Meteor, Backbone, even jQuery was still hot back then, and certainly dozens more. Technology had advanced enough that we finally were seeing new innovative ways of building web apps, and everyone had their own opinion on how to do it the Right Way.

This was certainly good for technological progress, but if anyone recalls the arduous process of "which frontend stack should I choose?" back then, it was an absolute nightmare. I had no idea which framework was production-ready, which framework would be around in 2 years, and heck, which framework even worked as advertised (and I remember distinctly wasting two weeks of my life trying to get a-then-very-beta ember-data to even fetch correctly, and eventually gave up and apologized to the client that their project wouldn't be done on time).

Fast forward to 2018, as happens with basically every technological advancement ever, we are starting to standardize. Most of the evolutionary branches are dying out in favor of one or two of the strongest candidates (in this case, React, as far as I can see). This has resulted in huge maturity gains, performance improvements, library standardization, lower training costs, generally lower software maintenance costs, and overall stability in the ecosystem. And as someone who is generally a fan of the "more wood behind fewer arrows" philosophy, having suffered through the world war that was JavaScript in 2013, I am really enjoying the calm before the next storm.

Because, as always, the next wave of disruption will rise again. This current consolidation is at the expense of change -- there will be less paradigm shifts, there will be less new ways of doing things, and that both provides for stability and stagnation. But like clockwork, come about 2020, there will be a new catalyst for change (maybe our next Cambrian explosion will be thanks to WebAssembly?).

So in short order, don't worry, we'll get to play the "which of these libraries even works" game in no time.

Re: Why is Ember fading away?

#13
It came and went, plain and simple. It made assumptions and design choices which at the time seemed sound but by today’s standards look silly. Mustache templates for instance. Separating logic from presentation should be a choice not a hard/fast rule with a severely limited templating engine.

Re: Why is Ember fading away?

#14

I think it's the natural evolution of frameworks. About 5 years ago, we had the Cambrian explosion of front-end frameworks -- Angular, React, Ember, Meteor, Backbone, even jQuery was still hot back then, and certainly dozens more. Technology had advanced enough that we finally were seeing new innovative ways of building web apps, and everyone had their own opinion on how to do it the Right Way. This was certainly goo…

Angular 1 and Backbone belong to an older generation of frameworks imo and shouldn't be lumped in with React and Meteor.

Re: Why is Ember fading away?

#15
As someone who has moved from using Ember daily for 2+ years to now using React/Redux, I miss Ember on a daily basis. React suffers from lack of standardization like no other. I hate starting a new project or working on an existing project and having to think about where to put things. Ember grew from developers who existed in the Rails space and understood that convention over configuration is the way to go. React devs likely came from Express land where everything is composed of tiny modules that you'll spend days carefully putting them all together.

The React world is the wild west. Libraries come and go so quickly that you're better off building it yourself if you see your app as long living. Ember, on the other hand, does a great job of keeping everything in check. Because the community is so small, things just work because there's less things to maintain.

Admittedly, when I first got into Ember, it was pre 1.0 days. Even prior to v2.0, Ember struggled to keep things working as intended. Now that the framework has matured, most pain points have gone away. The way they do their versioning is the smartest in the industry. React has been around for years and still the community seems fragmented.

What it boils down to is:

Is your app fairly small and intended to only last about a year? Use React

Is your app core to your business used by many and intended to last for the foreseeable future? Use Ember

Re: Why is Ember fading away?

#16
I think there is a lot of tribalism in the js community and liking one framework (often the one you use for work) is synonymous with hating the others. Take this survey done where the people who have "heard of Ember and would NOT use it" is completely outsized with the rest of the data. (https://stateofjs.com/2017/front-end/results/)

I've been developing professionally for 15 years and even in that short a time I've had to frequently start fresh in new frameworks or languages as the old ones lost their edge. The concept that someone would all-caps "NOT" use something is hilariously junior to me because the sands _will_ shift and you _will_ one day face the cliff of: my existing skills have no job demand, do I learn new skills or change profession? Harboring tribalism toward frameworks will not help you make the right decision in times like these.

Re: Why is Ember fading away?

#18
I invested a lot of my time over the last few years into Ember and though it was a difficult decision I eventually had to accept that I would be better off moving to Vue.js. I have written a bit about my decision here: https://hackernoon.com/why-i-am-migrating-propertywebbuilder...

Its a shame but the reality is that ember will not stand the test of time.

Re: Why is Ember fading away?

#19
Ember felt very Rails-y to me. The ember community felt entrenched in rails, and rails is reaching end-of-life. It is old-school thought. Ember made existing ideas better, a 2x gain maybe. I want new ideas, the 100x leap.

Re: Why is Ember fading away?

#20
Using a framework per se is a drag. If you want to do something quickly or have a bunch of choices made for you, it can be great. But as soon as those choices don’t make sense for you, it’s too late and you’ve painted yourself in a corner. You need to use their tooling and generally shape your workflow into what they’ve envisioned. Instead, it’s all about smashing a bunch of libraries together these days.

I know Ember has been composed as a set of libraries, but it seems like a lot of the advantage it offers comes from buying into things like their server-side rendering, templating approach and so on. There are plenty of engines out there already that will render a virtual DOM or whatever, though I’m sure theirs is really nice.

Plus, maybe people are just wary of something that doesn’t have the clear backing of a big company. Not that I’m saying people should feel that way.

Post reply on HN