In the past I've been working on enterprise application suites for the desktop. The kind of monstrous applications that run on electronics manufacturing floors.
Such a typical application has a shell, modules, inter-module communication, dynamic module loading, plugins, and much much more.
One can think of it as Eclipse with its ecosystem.
To build such a system on the Enterprise with its typical NIH (not invented here) paranoid attitude was VERY hard.
I had to reinvent a lot of wheels.
On the Web, I've used backbone for a long time now, since almost the time it was out. I saw the same problems building complex applications with it (that backbone-marionette amends to a great deal). For a larger project, I evaluated Angular. Had very strong negative feelings about it.
Then I tried Ember. It took me a long time to "get" it. The only thing that kept me motivated is knowing that Yehuda, Tom and Trek and other capable people are contributing on it. I kept fighting through the outdated tutorials, the outdated videos, and even the peepcode video was embarrassingly confusing to me at some point.
But then it hit me. Ember and its infrastructure, the way its MVC is rigged, was very similar to what I was building from scratch on the desktop many years ago. It truly IS the one framework that "gets" desktop, or client-side, applications.
The causer of my confusion was that I didn't completely let go of the "Web-think" for building application. I was stuck at either server-side MVC (MVP), or bare-bones frameworks such as Backbone.
It's been just too long out of the real complex desktop game for me, to realize what that I'm looking at is a proper MVC framework.
So for me, Ember ended up as being great - it still takes me back to the way I was building desktop applications, and I'm sure it will become even better and better.
To understand it, you need to cold-boot your thoughts into that classic desktop MVC place; and if you were never in it, I think Ember is an excellent way to get into it as opposed to other frameworks.
One can think of it as Eclipse with its ecosystem.
To build such a system on the Enterprise with its typical NIH (not invented here) paranoid attitude was VERY hard.
I had to reinvent a lot of wheels.
On the Web, I've used backbone for a long time now, since almost the time it was out. I saw the same problems building complex applications with it (that backbone-marionette amends to a great deal). For a larger project, I evaluated Angular. Had very strong negative feelings about it.
Then I tried Ember. It took me a long time to "get" it. The only thing that kept me motivated is knowing that Yehuda, Tom and Trek and other capable people are contributing on it. I kept fighting through the outdated tutorials, the outdated videos, and even the peepcode video was embarrassingly confusing to me at some point.
But then it hit me. Ember and its infrastructure, the way its MVC is rigged, was very similar to what I was building from scratch on the desktop many years ago. It truly IS the one framework that "gets" desktop, or client-side, applications.
The causer of my confusion was that I didn't completely let go of the "Web-think" for building application. I was stuck at either server-side MVC (MVP), or bare-bones frameworks such as Backbone.
It's been just too long out of the real complex desktop game for me, to realize what that I'm looking at is a proper MVC framework.
So for me, Ember ended up as being great - it still takes me back to the way I was building desktop applications, and I'm sure it will become even better and better. To understand it, you need to cold-boot your thoughts into that classic desktop MVC place; and if you were never in it, I think Ember is an excellent way to get into it as opposed to other frameworks.