So I'm going to chime in here, because I both agree with and disagree with Jason, and because I actually _have_ written a framework (for Flash, not JS), but it was internal to a company and got truly productive around the time Steve Jobs convinced the world that all the web's performance problems were Flash's fault, not the fact that developers who just want to build something without understanding the fundamentals are never going to produce anything performant. Jason probably has, too, in that he has probably produced an opinionated architecture that handles a lot of the heavy lifting, and other people need to fit their code into it in order for everything to work.

I agree with Jason that frameworks tend to focus people's attention on the wrong things, and that you really need to be good enough to ignore all the noise (i.e. truly understand the fundamentals) if you're going to produce anything that's maintainable and useful to users.

But let's say for the sake of argument that Jason wrote a framework that demonstrated all the virtues he rants about, for free. Then, in order to get it adopted, he'd have to provide some sort of training and documentation, probably, again, for free (at least at first). Look at all the people who go on his posts and tell him he just doesn't understand "modern" development because he's living in the past. He would have to somehow convince all those people that the ideas they've uncritically accepted are wrong and spend time learning this framework, when at this point there are no companies with jobs using it. Or, he would have to convince companies to use it when there's no talent pool. Do you start to see now why popular frameworks tend to originate with tech giants? Once you understand that, you can understand why the frameworks themselves focus more on esoteric things that are more about the coder demonstrating how facile he is than the "boring" fundamental things about how browsers actually work. Just because you can write a knapsack algorithm does not mean that you can style a responsive site.

We all know that Betamax was better than VHS, but VHS won because of market forces. There may well have been hundreds of frameworks written that were excellent, but died in obscurity because their authors couldn't get enough people's attention.

At the same time, developers develop what they can get people to pay them to develop, and I can't tell you when the last time a recruiter contacted me with an opportunity to code a vanilla JS project.

I do think Henrik has a point in that most developers aren't very good and a lot don't even care to become very good, and if you turn such developers loose in vanilla JS, no telling what you will get. Something is very wrong with how we train new developers, and no one person is likely to be able to fix that. And having frameworks at least somewhat constrains what type of mess you wind up with.