You can make the calling interface look like that (that’s definitely good), but the implementations in the example were still very document centric inline html.

You can make the calling interface look like that (that’s definitely good), but the implementations in the example were still very document centric inline html. It all just seems like it’s a lot more low level work than it ought to be. I feel much the same way when using a (bad) orm: there’s just a vague feeling that the solution isn’t quite an appropriate fit to the problem. Maybe this gets better when you’re composing from a library of core components, I don’t know. Again, I’m very much an outsider here. I’m not saying that old GUI component frameworks are vastly superior either. I just imagined that we’d have progressed well past the kind of approach the article is mentioning. It’s a side rant, but I don’t think it’s entirely without merit.