AngularJS versus Ember
eviltrout.com
eviltrout.com
One nit from the essay: the author comments on the missing 'model' from Angular and provides an example of a computed property to illustrate this. In many (not all) Angular applications, the server is responsible for the model (think a Parse or other REST-style backend), and the Angular app is simply responsible for rendering it. Angular (err Restangular to be specific) has a way to do this with less code and an aesthetic (data binding in HTML) that I personally like. If, however, you were writing code that needed the client to have a lot more business logic (i.e. computed properties, browser local storage, etc) then something like Ember would be a better choice because of its 'Model.' A lot of shit crud apps, however, don't need that much business logic in the client (at least for an MVP), and that would be my narrative for why Angular appears more popular. That said, I use computed properties in a couple places with Angular, and watches have performed perfectly fine for my use (gallery of 1000s thumbnails). He should disclaim that YMMV.
Neither is inherently better, and both teams working on them are full of smart, quality engineers that understand these problems in ways I never will.
I'd like to call it, "the appeal to the bold new future fallacy, with sensationalist titles"
<template>
Is X trying to change the paradigm ? No. Use Y.
Is Y trying to change the paradigm ? Yes. Use Y
</template>
Some dumb geeks just can't handle "worse is better" !
My question is then: why are people purposely choosing not to use the advantages of client side MVC code? All of a sudden we have access to all these awesome features of long lived applications and people are choosing patterns that don't embrace it.
> That said, I use computed properties in a couple places with Angular, and watches have performed perfectly fine for my use
I do say in the end that there are viable workarounds. It's an issue of maintainability and consistency. Ember embraces patterns for ambitious applications.
If it works for you I'm glad, I just think we should open ourselves up more to the advantages of rich client side apps and not hold back!
You can't trust the client. Therefore a good part of the logic MUST be built on the server side. When you've had to build that logic once, building it again on the client side is a maintainability problem of exactly the same kind that you decry with Angular.js. Except worse, because frequently the client and server are different languages and so are harder to keep in sync.
Secondly, client and server interfaces are not as symmetric as you indicate. Discourse has much more validation logic on the server than the client.
We generally "assume success", because the vast majority of requests do go through correctly. So if you submit a spam post for example, we'll show it in the client side view right away, but once the server asynchronously replies a failure, we'll display an error message and remove it. This can be jarring in the case of a failure as you'll see something appear then go away, but the vast majority of the time it doesn't happen.
The client side interface should do what it does best - display stuff quickly and avoid server round trips as much as possible.
So, the app functions on its own in the client, and only checks-in periodically with the server to see if anything is not valid?
(As opposed to sending a request to the server for every change.)
So for example: when I submit a comment, it doesn't go to the server right away. It immediately displays successfully in my browser, and could be submitted to the server 2 seconds from now, or something?
Ajax by definition is asynchronous. When you submit a post, the Ajax request is fired right away. But we don't wait for a response before showing the post in the stream (success case). We show it right away.
Then WHEN we receive a reply, we check the error status and if it was success, do nothing. If it failed, we revert the UI to the previous state. This is really easy with a client side MVC framework.
The downside is the failure case can be disorienting, as something will appear and then disappear. But the vast majority of requests, say, 99% never fail so the user so we optimize for that case.
I will never forget my first interactions with a team who took this line of argument seriously. They were an outsourced team who had been given a spec for a website. They said they were on track. Our product team went down to talk to them. The most memorable quote was their lead developer saying, "We will just have to train our users to not hit the back button." Almost as bad was, "We have no idea how to support IE." (Over half of our business was IE, and many of our users were casual browsers.)
There is a legitimate debate between rich client-side applications and keeping work on the server-side. Anyone who tells you - either way - that one approach is obviously the right thing to do simply lacks full perspective.
But personally I've been burned enough by people who want to create complex client-side interactions with serious UI mistakes that I have a certain reflexive caution about arguments for pushing everything to the client.
YMMV (and apparently does).
> My question is then: why are people purposely choosing not to use the advantages of client side MVC code?
I didn't understand this question (it seemed too obvious to me and felt like a bait to get into a broader discussion). In both Ember and Angular apps you still have a 'model' that is stored outside of the DOM, the only difference is how the model is works, which seems orthogonal to 'advantages of client-side MVC.'
I wonder, though this is speculation, if this question and post are some kind of elaborate troll. EvilTrout hails from Forumwarz, which is a game that may use these techniques every day:
Forumwarz is a parody role-playing game that takes place on the Internet. In fact, Forumwarz is the Internet...in game form.
Magical dragon-faeries? Flaxen-hair'd elflords? Dank scary dungeons, reminiscent of Grandpa's basement? Kids' stuff. In Forumwarz, you can pwn trolls in ridiculous web forums...buy hacked warez from shady Russian websites...or upgrade your skills with breast implants, malt liquor and antidepressants.
To spell out the question in a different way:
I have an object that represents some kind of object, like a User, for example. On every page, I display the user's name.
With the 'long lived object' strategy, I fetch the User on the first page load, and display it. The JS object then sticks around, and as I navigate, the same User object in memory is used on each 'page' to put in the username. No difference.
With the 'mostly server side' strategy, I fetch the User on the first page load and display it. When I navigate around, I then re-fetch the User on every single 'page'.
I am not familiar enough with the details of Angular (and, frankly, Ember, which I'm still new to) to tell you if this maps 1-1 with the way that they work, but it's the two different approaches in a general sense.
(I will look into exactly what Services are, thank you.)
FYI according to Stack Overflow, Steve is in his late 20s and has been a professional developer for several years. I'm in my mid-40s and have been a developer since the last millennium. Both of us have been on this site longer than you, and have much higher reputation here. Neither of us are kids. Assume that we can understand your actual substantive points, make them, and move on.
You're questioning the decision making behavior of developers. Developer stereotypically like to think in black and white and therefore as developers we assume that all developers follow one uniform pattern.
To answer your question - I don't know, but this is just like asking why people like chocolate ice cream over vanilla. Here's my theory - Angular.js is "by Google" and therefore has a larger community.
Having the server contain the "single true source" is important for many reasons, most of them being the main benefits of the web in the first place - automatic backups of everything, automatic sync of everything, etc. And if you're backing up everything to the server anyway (e.g. every action requires a REST call to the server), then the server is in any case acting as the source of truth, which means it can also do some business logic processing.
That's not to say the client is "dumb" by any means, but you'll be talking to the server anyway.
The point is that due to the difference in architecture between an SPA and a more traditional all-server-side app, there's no reason to hit the server every single time. Isn't this the advantage of writing so much JavaScript? If you're going to make a bunch of requests on every page, why not just ditch the JavaScript and write it all server side?
The only atomic source is probably your database, and from the millisecond that you query it, it could be stale.
It is not the same thing as the class of bug I am explaining where you have multiple copies of the same object in memory leading to confusing state errors.
There are some objects that you just don't need to refresh constantly. In ember there's nothing stopping you from calling refresh when you enter a route, it gives you the object and leaves it up to you. If you think it's important to refresh it as the route changes, by all means do so!
For example, I have an Ember application running on both mobile and desktop that gives the user complete read/write access to their data even when the network goes down. When they get a connection again, everything synchronizes automatically.
This is an important requirement for us, and once we implemented it we realized that is has very useful side benefits as well: the entire application feels dramatically faster, because you're very rarely waiting for the server. Changes take effect instantly.
It also makes our backend infrastructure simpler. First, because we can take it down for a bit and nobody will even notice. Second, because once you have a true distributed synchronization algorithm running, adding another redundant server is no more complicated than adding another client. They're actually quite symmetric in many respects, running the same codebase.
http://www.youtube.com/watch?v=W13qDdJDHp8&feature=c4-overvi...
For instance, scope will be optional in the future, not only that but the roadmap for Angular Next is very impressive (see around the 44:40 mark).
---
Angular is accelerating; it has more stars/watchers/staff than Ember. Embers' core devs barely have time for pull requests. To the best of my knowledge Ember has no full-time staffers. Angular is fully staffed by Google. Both take contribution from the community.
I suspect Angular's popularity has more to do with a logo than features.
- The unnecessary new module system is a massive waste of time and obviously doesn't play well with others
- It's the first MVC framework I've used that can't properly loop over an array or primitives (the partials for the primitives be displayed,but changes to them won't flow back to the parent as Angular can't make a $parent (IIRC) key.
Also: 'NG' just fucking doesn't stand for Angular.
> Q: Why is this project called "AngularJS"? Why is the namespace called "ng"?
> A: Because HTML has Angular brackets and "ng" sounds like "Angular".
http://docs.angularjs.org/misc/faq
sigh
The fact there's an FAQ question on an abbreviation should tell you something.
Thanks for the sigh though, and the substantive response to the points raised in the body of my post, it really adds a lot.
Promoting it as 'by Google' is misleading.
Have you written about them? I haven't come across anyone who has enough experience with more than one of them to comment broadly.
(To the people downvoting my previous comment: wtf?! Politely questioning someone's stated experience when it borders on absurdity is perfectly valid. And the GP provided a perfectly reasonable explanation: s/10k+ apps/10k+ loc apps/.)
Backbone was tiring for me and my partner with all the files and code to set up each view. I also never liked putting templates in script tags -- it just felt dirty. We see Ember and Angular as similar (how do you pick which language you like?), and we only picked one for atheistic reasons (data-binding in HTML) in our current project. Since I haven't written the same app in each framework, it would be unfair to comment further.
What are you working on? I did a quick trawl of your history but didn't see any mention.
If you're like me, then you'd want to see some examples of craftsmanship to weigh the value of different opinions. Unfortunately, I can't provide links to these example works due to a very strict NDA (we did the work on behalf of a couple brand names, and ironically they wanted privacy). The projects definitely fell into the category of 'shit crud app' :)
<div {{bindAttr class="highlighted:paid"}}>{{name}}</div>
I am pretty sure you prefer Angular but are just too pussy to say it.
Maybe you think people will think less of you for preferring the framework that is less complex.
I assumed 'not even close' would mean this guy would be siding with AngularJS.
Its sad but many people have the mistaken belief that programming languages and frameworks that are harder to use are more sophisticated and therefore better.
What I think is that deep down people are afraid that others will think they weren't smart enough to figure out the more complex and difficult to use thing.
What I am sure of is that a system that is less complex and easier to use is better engineering.
I agree. I think this is what made the success of very popular libraries (such as Jquery), but also protocols like email and the web.
I've been trying very hard to use Ember on a couple small projects, and I found it very hard not to get frustrated by how difficult it is to execute some trivial tasks.
Part of the problem, I believe, is the way I learn. I usually read a quick tutorial, hack a little bit, read a bit more, hack more, read the full documentation, hack a real project (and getting into the wire of the API at this point). It's a bit oversimplified, but I believe you get the picture.
However, with Ember, it seems to me like it's hard to get started and running without having a full understanding of the full system and how everything fits together.
So yeah, in the end, maybe if you know every part of Ember and how it intrinsically work, then that's an awesome framework to work with. But for the others, like me, who learn in a more spiral-ish way, I think Ember makes it very hard to get shit done and appreciate the value of the framework.
And please, don't get me wrong, I admire programmers who can read the full documentation first.. A good analogy might be the manual (man) on Linux. I know people say RTFM all the time. But for me, it's really more about googling non-stop to get a feeling on the quickest solution for my problem.
I guess it's just a different way to learn and approaches solutions to problem. I know a couple excellent programmers who would just go reading the man and figure everything by themselves.
That being said, not sure why you're being so aggressive in your statements.. I feel you had very valid arguments but these are totally buried under this arrogance.
Funny, I felt the same way about AngularJS. I had been working for a few months with Ember. I was building a new internal tool and wanted to learn Angular but I found a lot of the concepts poorly explained in the documentation.
And I forgot that language like that simply was not accepted on this website unfortunately. However, I believe that the aggressiveness was warranted because there are quite a large number of organizations losing a lot of money on the basis of a poorly informed engineering decision.
Next time I will hopefully remember to be more political and use acceptable language when I feel so strongly about something.
Thank you for your honest comment.
I would ask that you consider, for a moment, that it is possible for someone to hold simultaneous opposing views. I also would ask for you to go back and re-read Brooks' essay about accidental and essential complexity. I believe that the choice of framework and the difficulties associated with a specific framework fall into the accidental complexity category. If after reading the essay you believe the same, then you would conclude that both choices in this thread have similar inherent merits. Such an opinion wouldn't be wishy-washy or a cop-out, but, perhaps, closer to a deeper understanding of our craft.
Every MVC framework I've used in production (Catalyst, ASP MVC, Cake, Spring, Rails, Rails with Spine, this batch of JavaScript frameworks) has had warts and frustrating moments just as all of the web languages have frustrations (Perl with maintainability, PHP with amateur hour inconsistencies, Python with tab/spaces that different IDEs will autocorrect, Rails with too much 'magic' that leads to serious bugs, Java with too much verbosity, Spring with too much XML, C# with too expensive support, Django with too many required add-on libraries, Rust with too few libraries, Go with too new, etc, etc).
Ember was not any harder than Angular to get going (ever try to untangle the service/factory/module/controller mess?). And I have no shame in admitting that I took at least three full workdays to learn each. At the end of the day the two frameworks were just different, in the same way that writing an app using JavaScript MVC is just different than how I used to write ASP MVC apps. The reason I use the newer stuff is because the newer stuff is more fun, I can be more responsive, and (in some cases) I can host an entire webapp on s3 (managing servers can be a headache). Despite some clear advantages, none of these different frameworks is inherently better. They all have tradeoffs. If Microsoft were to come to me and (with a large check) demand that I use ASP MVC, I would do so.
My first response reading your comment is that you have less experience than I do with software engineering. That may be untrue and unfair, but rash responses are correlated with inexperience from my career. I would ask you to consider your strong, aggressive reaction as unhealthy for merit-based engineering decisions. Learning and respecting that most decisions in software are about tradeoffs and not black-and-write-strictly-better-or-worse I believe will help you in your career. Furthermore, I hope you eventually come to point where you realize many of these different languages and frameworks all look the same after a while :) It's not until we can tell a computer what we what and it figures out how that our productivity will massively improve.
This is exactly what I was trying to figure out how to type.
Right now ember is especially hard to learn since it is constantly changing and ember-data isn't done yet. As a result there are 5 different ways of doing something and when you get stuck you finally solve it by finding someone else who had the same problem and that is when you learn that the tutorial was for rc3, not rc5. But the huge problem is that it took you forever to get there.
I personally think Ember will win, and it will win hard, for the professional developer market. I think it will come down to Rails : PHP :: Ember : Angular, but in the mean time angular will get a huge head start.
Once people get used to those common single bugs that keep them stuck for hours, they will be far more productive for non-trivial applications than angular devs will be.
1.0.0.rc1 was released on February 15, and (to my knowledge) there has been _one_ backwards-incompatible change since then.
> ember-data isn't done yet
You don't need to use ember-data: http://eviltrout.com/2013/03/23/ember-without-data.html
Want to become a truly great developer? A real one-percenter?
Get in the habit of diving into other people's big, scary codebases when you hit a bug. Learn how things work, learn why they're breaking, and fix it.
Who has time for that? That's part of my point. Many people won't invest in their own capabilities in this way. But once you do, you get dramatically faster at it. Soon it's no big deal to jump into the guts of Rails or Ember and bend them to your will.
This is also the best way to truly be able to judge the quality and usefulness of competing projects. Instead of following the internet popularity contest, you can go to the source and see pretty quick how good the ideas really are.
Rejecting 'magic' and just reading the code definitely helped me to get to another level of coding.
The best success I've had for my modest web projects has been when I've used Flask and accompanying tools for server-side (I'm guessing Sinatra would work just as well), and Backbone for client-side apps. It takes a while to set up, but I understand everything that is going on, so I can easily fix problems as they occur.
Some time ago, I started off with AngularJS, I liked what I saw (except the dirty checking part) and started building my product / startup with it. I don't have the time to stop all dev and experiment with EmberJS, but would like to know if it's really better or if it's to-mah-toh vs tom-ae-toe.
I mean, how hard could it be to have an AngularJS port where the dirty-checking part is removed and Ember-style getters & setters are introduced (thus removing the only efficiency-related issue that I ever came across).
On the other hand, what's the equivalent for "directives" in EmberJS? It's singularly, the most powerful concept that I've come across in AngularJS.
Btw, AngularJS has an accepted way to define models -- inside "factories" (I know they just sound like Java-esque hell, but they're really simple, once you get to understand them).
I've written non-trivial apps for the same backend, first in Ember, then in Angular. The Ember app is the consumer-driven frontend, while Angular runs the administrative interface. It wasn't my intention to use two different frameworks, but we started the Ember one when Angular was really just starting to gain traction. A few months later on the advice of a friend, I thought I'd give it a shot.
The biggest problem between the two isn't really anything about the frameworks. They're very different beasts, and they both have their strengths and weaknesses. The noticable difference out the gate is the documentation. Ember is a mess. It's older, and things keep changing, so documentation gets out of date. Also most of the tutorials and cookbooks online are all for older versions, and that can be very confusing when you're starting out.
I personally find Ember's patterns to be confusing at times, but I also have this hunch that there should be a point I get it and it will all make sense. For example, there's controller, which should manage representations of your models for UIs, and there's views, which manage events in your templates. Right now that's where I always get lost. I have no idea where to put anything. When you click on a button, is it the controller or the view that handles it? The answer appears to be "yes, depending...".
I think this confusion comes from Ember's philosophy: it's trying to be a desktop-like UI kit for web development. That works 80% of the time, and the other 20% introduces "fudge factors" like this.get and this.set. There's also too much magic happening in Ember, which makes it difficult to learn. Things happen based on the names of objects and templates, and then you hook into this magic to override certain steps. Since I don't spend all my time in Ember, when I go back to look at code I've written previously, I'm often lost at where something was updated or triggered.
Angular on the other hand is very straightforward as to where things go. For example, people mention directives. You start off by hooking events and values into your templates in the controllers. After a while, you'll start migrating that code out of controllers into directives, so you can reuse it in many different templates. As your application grows, your templates become filled with reusable directives and your controllers get simpler and simpler. It's a very nice philosophy.
Also, Angular doesn't buck the web workflow as much as Ember. Like I said, Ember is trying to be a desktop-like UI for the web, while Angular feels more like you're just transferring your older, do-it-on-the-server templates to JavaScript. This gives you a lot more flexibility with UIs and animation, but it doesn't completely break the browser, which is why I think Angular doesn't have as many edge cases to deal with as Ember.
On the downside, Angular has this initial easy learning curve, which gets steep later. The basic stuff is easy, and you'll have a web site up in a day or so, vs. a couple weeks to figure out the basics of Ember and get something meaningful online with that. However, Angular quickly gets complicated when you want to dive into complex directives and some of the more esoteric aspects of the framework.
For support, I haven't been hugely happy with Angular. There are more people using it, and there are more blog articles and tutorials out there on it, as well as lots of plugins to add features (like websockets and local storage support). That's great. But the community itself doesn't feel as helpful. Most of my posts to the mailing list went unanswered. I haven't tried Angular's IRC, so maybe that's where I should be.
The Ember folks on IRC are hugely helpful and very friendly. There is a bit of "our way is the best way", but I think that comes from understanding Ember and everything you can do with it and knowing it's probably a more powerful framework in the end. However, the discussion forums are horrible. They're using Discourse, which was written in Ember, and it's nearly impossible to find answers to anything previously posted. Because it's in Ember it's all JavaScript, so nothing is indexed on Google, and the built-in search is awful.
I would recommend if you're starting out and you don't have super, super strong JavaScript skills, you should go with Angular. You'll feel more productive quickly. Ember will be frustrating. If you're an advanced JS developer and you've built desktop UIs, maybe Ember is more your cup fo tea.
Also take a look at http://yeoman.io. It's by far the easiest way to get an Angular "build" system going. Yeoman supports Ember too, but like everything else with Ember, it's a bit tricky (for example, updates to the templates will crash grunt server if there's an error).
Not true. All Discourse content has been indexed by Google since day one. Even though it's a Javascript app it's search engine friendly.
Secondly, Stack Overflow is the official place for Ember questions. The forums are for general discussion about features and other Ember related topics. Go to SO if you have a question and tag it as ember!
> Our site is indexable by Google and it doesn't do much fancy. On certain URLs, we generate a small HTML view of the content in the <noscript> tag. You can see this by viewing source or disabling JS in your browser. It's just a simple ERB template in Rails, and uses the same object graph that we serialize via Active Model Serializers.
> Google can see it and index it, we've confirmed by searching post launch.
> As time goes on we'll probably work more on it to make the SEO even better. As you can imagine it was tough to do when we were in stealth mode ;)
Maybe we could get an update from eviltrout on that.
I was leaving it to him to explain how it works :)
However, the process of checking whether JS objects have changed, is not very pretty. AngularJS maintains a copy of every object in the "scope". After every "digest cycle", Angular compares the copy of each object with the previous known state/value and figures out what changed. (I'm sure internally they're using really tight loops & optimized code to make this fast).
A "digest cycle" is triggered after every event that may possibley change the state/value of the tracked JS objects. Eg. a keypress, an HTTP event, a click, etc.
When you have several-thousand bindings, each digest cycle gets computationally very intensive and the app starts to feel sluggish. Typing a key in a form-field and waiting for it to appear on the screen, can also lag. Please note: this happens if you have several-thousand bindings on one page -- something that is actively discouraged in the Angular world.
In my case, I was building a dashboard app, where the nature of the app required me to show several thousand data-points with the option of "zooming-in" or "zooming-out" from data points (not in the visual sense).
Later, I realized that this limitation is easily overcome if you stop using the implicit two-way data bindings provided by Angular for such cases. If you know that those JS objects aren't really going to change with every digest cycle, it's best to not use {{ var }} to display them in your views and come-up with a custom directive that does NOT set-up a watcher on the variable. The downside is that you have to manually update your views whenever the underlying data does actually change -- not very different from how EmberJS does it.
http://brandid.github.io/parse-angular-demo/
"In AngularJS, every time you visit a route, it passes an idand has to resolve it in your controllers."
- not true, just use nested controllers and prototypical inheritance to access the already resolved object
"AngularJS doesn’t follow this philosophy; it encourages you to throw away what you already had and find it again (probably from the server!)."
- not true, thats why $rootScope exists
"None of the examples or blogs I read through demonstrated how to reuse object instances, for example. Is that just not done in AngularJS?"
- not true, all scopes inherit from each other, javascript prototypical inheritance applies
"The problem is that the $watch function belongs to the $scope and NOT the Room."
- not true, you can put a $watch on the model if use Backbone for your models
"If you had an array of Room instances for example, you can’t watch them all without looping through them all constantly."
- not true, you can watch a collection of models, and you can do things like watch the length of the array for less processor intensive use of $watch
If you're arguing that with a bunch of extra code on top of AngularJS it does what Ember does out of the box, that was kind of my point.
What's your take on points #1, #2, #3 and #5?
What you are suggesting is good practice but something that I suspect very few Angular developers do as it's not encouraged by the framework itself.
If you follow the guides on the official Angular site, they're done none of the things you suggest.
If you're argument is about what Ember 'encourages' vs what AngularJS 'encourages', that is your opinion and not a weakness in the framework.
I think it's great you've taken some time to learn AngularJS, and I think if you dig a little deeper you will be pleasantly surprised with what you find.
Yes, thats why I choose Angular over Ember every time. EvilTrout obviously wants to see Ember succeed, I would to if I spent so much time writing a big Ember app and see it loosing out in the Javascript framework wars.
The one thing I will say is that his point might be true about performance now, but once Object.Observe lands we won't need dirty checking anymore. It will be easy for Angular to take advantage of it but as far as I understand, Ember will require some major changes to take advantage of it.
The way I see it Angular is a forward facing framework showing you how new technologies coming like Object.Observe and Web Components could work together in a framework today.
I find the concept of directives one of the most powerful and maintainable features of any framework I have ever used. It's probably the "killer feature" of Angular.
You didn't really rebut anything in his post, just asserted that Angular is 'forward-looking.' EvilTrout makes very specific claims, as someone quite familiar with Angular, you're well-poised to give a real answer, not just chest-puffing.
In my opinion this sucks. I love being able to use plain old javascript objects in Angular. Not having to extend every object with ember.observable is a big win for me. When O.o lands Angular will be just as fast (if not faster) as Ember but without the need to instantiate new Ember objects, and always use get() and set() on them.
Of course Ember could change its API in the future, but Angular won't need an API change because its already using plain old javascript objects.
While I'm with you and love being able to write my own simple classes and POJOs, ultimately I'll value being able to support users who won't or can't upgrade their browsers.
I also question your assumption that Object.observe will be as fast as Ember's observation system. I wouldn't be surprised if it was several times slower.
Object.observe also doesn't support array observation, which is something we support in Ember.
If I chose my languages or frameworks based on what was the most popular, I'd be using Java or PHP, not Ruby and Javascript.
Additionally, Object.Observe benefits Ember just as much as it does Angular.
I tried Ember first. Got excited replicating the sample blog app, but got stuck trying a simple task: show the most recent post by default (see this bug: https://github.com/emberjs/ember.js/issues/1642). Later I started to trip into namespace clashes - you have to jump a few hoops if your model has a field named "content", "is_new" or "is_deleted" (http://stackoverflow.com/questions/16515900/reserved-atribut...). My general impression is: there is nice ideas here but Ember is too opinionated and has to many magic going on (hard to debug).
In contrast, Angular shows better design everywhere. No name clashes, framework methods and attributes use special prefix $ (like python uses double underscores). Everything I got stuck with ember, I was able to achieve elegantly with Angular.
I love the eviltrout but this post is lacking a bit of diligence. Angular learning curve is a little steeper but IMHO it is way more finished and well designed than Ember (Ember data has good ideas, they have a cute mascot, and it ends here).
Framework simplicity is good if your application is meant to be simple. But you should be cautious when choosing simplicity if your application is ambitious and you care about supporting it for a long time.
This is what I like about Clojure. It is so composable, that simplicity doesn't come at the expense of features. The language makes composing disparate libraries so easy, that it's quite trivial to mix and match different libraries and have them all work in tandem. You don't only use what you need, but for every aspect of your application you can choose the best tool for your particular needs.
(ns my-app.my-namespace
(:require [some-library.some-namespace :as foo])
Composition is tedious to reason about in Ruby when your only clue is a menagerie of `require`s at the top of the file. Or, more commonly, when your only clue is a Gemfile because all the gems are auto-required and omnipresent.
Today I was seeing this -- https://en.wikipedia.org/wiki/Knuth-Morris-Pratt_algorithm
How can I begin to implement it in clojure ?
Specifically, how do I _ingeneral_ deal with assignments and while loops.
In common-lisp, for example I would use the `loop` macro and defvar-sefq combo, (beginning lisper).
In ruby, = and while itself.
http://clojuredocs.org/clojure_core/clojure.core/let
(let [username "prollyignored"
karma 7]
(println username "has" karma "karma points."))
; prollyignored has 7 karma points.
(loop [n 0]
(when (< n 10)
(print n)
(recur (inc n))))
;=> 0123456789
(for [x [0 1 2 3 4 5]
:let [y (* x 3)]
:when (even? y)]
y)
;=> (0 6 12)
(filter even? (map #(* % 3) [0 1 2 3 4 5]))
;=> (0 6 12)
Now in,
while m+i is less than the length of S, do:
if W[i] = S[m + i],
if i equals the (length of W)-1,
return m
let i ← i + 1
....
When you write:
(defn square [x]
(* x x))
(def square
(fn [x]
(* x x)))
(loop [n 0]
(if (= n 10)
n ; true case: return n
(recur (inc n)))) ; false case: loop with n++
;=> 10The two issues that this author are singling out aren't exactly addressed fairly so I sort of feel inclined to defend Angular here.
Author states that one downside of AngularJS is that it doesn't conform to the Uniform Access Principle. JavaScript the language itself doesn't conform to this either. From what I can tell from my brief review of AngularJS, it tries to be as close to the implementation of JavaScript as possible. I don't see this as a bad thing but rather as a good thing. Using methods like #get and #set drove me crazy when I was using Batman.js. I understand why they exist, but why are #get and #set acceptable ways of accessing object properties but #area isn't? I think they both have the same performance implications, don't quote me on that as I'll never claim to be a JavaScript method cache expert.
His first point is that in AngularJS you don't have a true base object that your models inherit from and instead you use plain JavaScript objects. This sounds like another bonus to AngularJS in my eyes! And if it doesn't seem like a bonus to you, you can quickly create a light abstraction to serve as a Model.
All in all, having not used either of these frameworks this blog post actually makes me want to use AngularJS more than it makes me want to use Ember.
Because Ember only has to recalculate the area after set is called on length or width. If area is just a function, it gets recalculated every time it's called/get'ed.
That said, I have a pretty large Ember app, I adopted Ember for my app when it was known as SproutCore 2 (before AngularJS was known). Both my app and Ember have gone through a very long, very painful maturing process. But it's finally in a really good place. Ember's patterns are the result of months (years really) of spinning, cycling, iterating and collaborating with developers using the framework. Seeing the process of evolution and open collaboration unfold in the community is just as exciting and thrilling as the state of the framework it's self.
My advice is to try them both and decide for yourself. The only thing I'd ask is please don't assume one is easier or better than the other with only a few hours of use. If you're planning to build a sizable app, please take the time to evaluate them both thoroughly and decide what's best for you.
I'm happy to accept the problem is hard, but I doubt it's so complex it's impossible to decompose.
This isn't unique to js frameworks. There's plenty of systems like this e.g. Hibernate. At least backbone was small enough that I could seriously evaluate it in a month long project. (For the record, I didn't like it, but at least I actually can have a considered opinion on it without being well past the point of no return.)
- Derby.JS is one such a full-stack framework.
- There's also Meteor but this has a little less community support.
- Eyebrow.JS is coming (there's a number of blog posts and talk videos) but I'm not sure if it's released yet.
Discourse 'private topics' leaks (1) are good examples of typical security Rails apps bugs (the result of overusing ActiveRecord). I'm not blaming Discourse, the same kind of bugs appear in other popular Rails codebases (Diaspora, Spree, Redmine).
Ember.js is very Rails-like. I didn't use Ember much, but the patterns look so similar, that it may lead to similar kind of problems as in Rails apps.
I'm not saying that we shouldn't use Rails or Ember. In fact, I'm a big fan of Rails. It's just worth being careful of not falling into the trap of overusing the framework patterns.
The frameworks are best for the infrastructure parts of the app - http, persistence. At the end, it's better that you can control your models. That's the part of Angular that I prefer over Ember - you own your Model part.
(1) http://meta.discourse.org/t/digest-mail-ignores-secure-group... http://meta.discourse.org/t/non-authenticated-users-see-priv...
Had we used raw SQL instead of an ORM we would have had the same issues. All projects are open to this style of bug. The correct thing to do is report, close them quickly and add tests to prevent them from happening again (which we do.)
I've seen so many of such bugs in other apps, that I treat them as part of "ActiveRecord price".
ActiveRecord over-usage makes it very easy to miss the places, where queries need to respect new rules.
You're doing an awesome job with Discourse and the team approach to such bugs is very good - no doubts here. Thanks for your work!
We did try to write it in Ember first, but found it much harder, both because of a much bigger lack of documentation, and because the versions weren't very stable and broke from RC to RC.
The title is argumentative and inflammatory. I think lots of people will read the title on Twitter and roll their eyes at the Ember attack-marketing machine when the article itself contains some valuable points.
I think it's a very good point and I doubt I'll do it again.
;-)
1) The DOM is the model. JS reads elements and data attributes and builds a model from them. Then updates the model when the DOM changes. (Angular)
2) The model is loaded from JSON. Then the App builds the DOM based on the model. (Ember)
So is the question: Does it start with the DOM, or start with javascript objects?
Am I understanding this debate correctly?
If nested views are in the basic Angular docs, it's even more strong. Don't just say someone is ignorant, show them how to not be ignorant any longer. It helps them, and makes you seem like less of a jerk.
>> If you were using jQuery to do this, you’d likely bind a function to the click event on each row. When fired, your function would change the CSS class of the row to highlight it.
>> You can’t stop there, though. Your function also has to remove the CSS class from any other rows that were previously selected. That’s a little crazy if you think about it! Changing which item is selected means we have to know everywhere that it’s been represented in our UI.
An easy way is just to do:
$('<all-row-selector..>').each(<unselect>)
.find('<selected-row>').each(<select>)I have the feeling that, by choosing the simpler framework, I can better maneuver around future complexities as I better understand what sits under the hood. Ember, which seems more elaborate than Angular, tries to anticipate solutions to problems I have yet to encounter. Maybe I'll regret it later. I don't know.
The reason I don't like either, or any framework in particular, is that none seems to really make a sane toolchain out of html, css and js.
Especially since the opening argument of the author is that AngularJS is gaining momentum because "it is simpler".
My personal experience is that when I ported my application from Ember to Angular, my code base (~5,000 lines) get simplified considerably. Based on what I've been reading, I'm not the only one making these observations.
My main beef with Ember was not about complexity or simplicity, just that it's not as powerful as AngularJS and it also covers only a small part of what AngularJS gives me out of the box (especially server side includes).
And even if AngularJS really is better, it will be left behind. Just because...
However, today, I feel, Python is more popular than Ruby (at least in my part of the world) because of various other reasons. Even though a lot of web frameworks are doing things the Rails way.
this application also has to have a pluggable(200+ different company business/validation rules) rule engine, and being able to configure angularjs services through its providers interface has been very helpful with that problem.
we're not done yet so i can't give a final verdict but we've been overall pleased with the framework so far. i also agree with some of the sentiments in here where angularjs does not currently seem to have a culture of what is "idiomatic" angular which can be very detrimental in the short and long run.
one of the reasons we didn't go with ember is that the decision was made at the time where in they were making huge changes to ember.
For example, there's a wide common case of simple apps hacked together by 1 or 2 developers, traditionally composed of jQuery spaghetti. Maybe Angular is a perfect tool to better organize these typical, not-particularly-ambitious apps, where Ember would be a pre-mature optimization.
And perhaps Ember is the perfect tool for sophisticated apps written by teams of developers.
As the article says at the end:
> Ultimately, I think you should examine your application’s goals. Do you want to build something that pushes the boundaries of what people expect from the web? Is your application going to be super simple, or do you want to add powerful features and maintain it well over time?
> If you’re serious about front end development, I advise you to take the time to learn Ember properly.
The article isn't fair to Angular. It's like saying, If you're a professional web designer, you should always use semantic markup and a CSS pre-processor, and never Twitter Bootstrap. In fact, Twitter Bootstrap and Angular both have their places alongside Sass and Ember.
DFA drove a lot of early refinements in how AngularJS works. Since then, it's largely community driven.
Really? google frame ? google reader ? (...) they dont give a f..
While Angular is quite complete feature wise , the UI responsiveness of a AngularJS application is not that great. At least that's my personal experience.
I disagree. I've built many applications using both frameworks and Rails is overkill 80% of the time, especially in front-end heavy apps where most of your routes are GET/POSTing JSON data.
I like to write my own SQL-Queries, so Sinatra is the better choice. But if you'd want to use an ORM, why not directly use Rails?
An ORM (the only example listed by you or the author) does not a Rails app make. There are dozens of ORMs of all types across every language; DHH did not invent the Active Record pattern and Rails' implementation of it is not the only viable option, even in Ruby (consider Sequel's ORM for example).
It seems like you must've had to implement a fairly big subset of something like Angular in the process, no?
Did you encounter similar issues or do you rely mostly on unit tests for debugging?
If you don't need that level of functionality, then fine, don't pick up the framework with a few extra KBs. But keep in mind that with any luck, your app will start to grow. When it does you'll need to scale and perhaps you'll discover there was value in using a more functional framework.
I'm not sure why this is a problem. A model is just a representation. Could be one, could be many objects.
That's the point of using a framework. You need to stick to conventions since frameworks have an opinon on how things should be(Ember does i guess).
I never tried Ember ( I will ), but AngularJS comes with a lot of battery included , that's why it is really awesome.