found drama

Why Web Tools Like AngularJS Need To Keep Breaking Themselves: Matt Asay, writing for ReadWrite.

tl;dr: AngularJS 2.0 is targeting the future because the future is where we’re going and anyway let’s not forget that the existing architecture/design is at least four years old anyway, and who’s on a 10-year software development lifecycle anymore?

In large part, I think the points that he’s making are good, but easy to make. What he doesn’t seem to address, and is the thing that keeps coming up in the conversations I’ve been having, has been the real-world developer concerns around the upgrade path. Or rather: the fact that none has been communicated. In other words, “the point” of picking a framework like AngularJS has a lot to do with not wanting to solve the problems that are not central to your application’s domain. You’re building a travel planning app or a photo sharing site or a CMS or… fill in that blank with just about anything. You’re not building dependency injection and data-binding and a whole host of other abstract things because… well, that’s not what you’re customers are paying for. And I think that by and large no one minds the occasional refactor of their application, so long as the things that you’re refactoring are your features.

Not having an upgrade path is a source of anxiety because it goes from feeling like a “big refactor” to “complete rewrite”. This is the cringe that I hear in everyone’s voice.

That being said… (1) I’m still not convinced that we won’t see any upgrade path – even if it’s a painful one. And (2) I agree that we want the frameworks we’re using (when we’re using them) to be powerful and tap into the powerful underlying modern APIs.

About !undefined