How Google broke the OSS compact with Angular 2.0

#1

Post by Garbage » Mon, Oct 27, 2014, 2:21 PM UTC

How Google broke the OSS compact with Angular 2.0
codebetter.com

Re: How Google broke the OSS compact with Angular 2.0

#2

Post by untog » Mon, Oct 27, 2014, 3:58 PM UTC

TL;DR: It's being developed in secret. Blog posts takes a lot of words to say one thing.

Re: How Google broke the OSS compact with Angular 2.0

#3

Post by whitten » Mon, Oct 27, 2014, 4:06 PM UTC

I think this article has a good point re community building. When a group of people move, it takes more effort to get them to start and stop (social momentum). Transparency of the source of software allows the external community to plan and train, and do other group-oriented things that a single-developer shop can do with more agility.

Re: How Google broke the OSS compact with Angular 2.0

#4

Post by mindcrime » Mon, Oct 27, 2014, 4:14 PM UTC

untog wrote: post #2

TL;DR: It's being developed in secret. Blog posts takes a lot of words to say one thing.

This isn't exactly a surprising thing, coming from Google. Look at the way Android development is handled. It's pretty much all "develop in secret, throw some code over the wall know and then".

Re: How Google broke the OSS compact with Angular 2.0

#5

Post by ryanpardieck » Mon, Oct 27, 2014, 5:09 PM UTC

I've grown kind of tired of these sorts of posts, where people assume that OSS means handling the community exactly the way you prefer.

Does it have an OSS license? Do they adhere to it in practice? If so, then your complaint isn't one of OSS, but rather community politics.

Re: How Google broke the OSS compact with Angular 2.0

#6

Post by CmonDev » Mon, Oct 27, 2014, 5:36 PM UTC

And what is the deal with AtScript? If you are going to randomly "fork/borrow" a language you can at least consult the original authors. Typescript is inspired by C# not Java, so those are attributes - not annotations, and they are in square brackets not start with an @ character. And it's called 'reflection' not 'introspection' in C#.

Re: How Google broke the OSS compact with Angular 2.0

#7

Post by general_failure » Mon, Oct 27, 2014, 5:37 PM UTC

If Angular 2.0 is architected the way it says in the docs, then it's really scary. They obsoleted everything - something which many developer put a lot of time and effort into. It doesn't matter if the design is better; what matters is a nice clean migration.

Atleast for me, if this change goes through, I won't return to Angular. Angular 3.0 might rewrite everything again for all I care. I just wouldn't trust the angular devs anymore.

Re: How Google broke the OSS compact with Angular 2.0

#8

Post by general_failure » Mon, Oct 27, 2014, 5:39 PM UTC

ryanpardieck wrote: post #5

I've grown kind of tired of these sorts of posts, where people assume that OSS means handling the community exactly the way you prefer. Does it have an OSS license? Do they adhere to it in practice? If so, then your complaint isn't one of OSS, but rather community politics.

There is some implicit expectation when people release a project as open source. As a OSS author if you don't understand this expectation, your project is going to hit massive roadblocks sooner or later.

Re: How Google broke the OSS compact with Angular 2.0

#9

Post by deedubaya » Mon, Oct 27, 2014, 5:47 PM UTC

To make big changes and stay with the times, you have to burn some bridges. Sometimes you have to burn a bunch of them.

I'll feel the same pain as everyone else when 2.0 is released, but instead of wasting time complaining about it, I'll pick it up, embrace it, and keep my whimpering to myself.

NG Core is kind enough to share their work with me. Sure, I've invested time in learning it, but they don't owe me anything for that investment.

Feel free to apply down votes here to express your anguish.

Re: How Google broke the OSS compact with Angular 2.0

#10

Post by mckinnsb » Mon, Oct 27, 2014, 9:01 PM UTC

I was just having a conversation with a friend of mine regarding if we should spend more effort in learning Angular.

While I can understand making breaking changes, and sometimes very big ones, I feel like this timing is incredibly poor, and is coming when many people are just learning the framework. At this point in time I do not want to invest any more time in the framework or its changes: the stable version I use will soon be obsolete; I can't learn the "new" version or use the "new" version now; and I don't know when the point will come when the AngularJS core team - and Google - are happy with the decisions they have made, if the point EVER comes ( especially re: Google ).

I also feel that while they addressed many of the concerns that the community had, the decisions they made were not strictly speaking "necessary". For example, removing all the ng-* stuff "looks" nicer, but it is now less clear which code is repsonsible for handling that declarative behavior. I would have preferred a mechanism which allows you to specify other "included" directives in a directive definition in a clear and concise manner ( without hacking on compile or link functions ).

I think I'm going to try Ember or something else. I liked Angular for what it showed me - that a data binding approach could really ease the pain of radical design changes, for example - but I'm not going to be sticking with it after this project. Time to move on.