First, I don't judge myself, or any other web developer, for getting pulled into the single-page application way of building to begin with. We all said it resulted in better developer experience, and it did, for a while; we said it resulted in faster apps, and sometimes it did, but really we just weren't checking. It felt fast to us on our fancy, up-to-date dev machines, but I don't think I was alone in thinking it just felt cooler, and that that was enough.

The ol' hipness pendulum seems to be swinging away from SPAs now, and back to building pages on the server, augmenting them with a touch of client-side code. A library called htmx is becoming popular enough for that augmentation that it's become emblematic of the approach; other brands for this new model of the old ways include "hypermedia" (a sure way to get my attention), HOWL, and the truly awful acronym HATEOAS. (Whenever I see this acronym I want to pronounce it "Hatey-O's." Have you eaten yours today?)

My personal reasons for getting on this train are pretty big: I crashed out of the React scene for a few reasons, but the biggest was complexity. The jungle of files you had to hack through by the end there, with GraphQL and Apollo and Redux - holy gods, Redux - and server-side shells for everything (I'm not talking about good code here) and JSX through build steps and all the configs. I would feel dizzy whenever I started to think about changing anything. The call of YouTube from the other tab got louder and louder. So no, when people get radical about the present state of things and say to chuck it all in history's fire, I don't judge.

But I don't have to judge - someone will do it for me. People are rolling their eyes saying "oh boy, guess it was time for a new hype cycle, guess there was too much money to be made on contrarian takes," and like, thanks for inching closer to a critique of capitalism, comrade. It's not as if there's any money to be made with complexity!

But would we have made everything much too complicated anyway, before the train managed to turn around? I think so. I think it is okay, natural, and maybe inevitable that the field of computing lurches back and forth like a giant, busted strandbeest in its search for the best way to build information systems. How could our thinking ever not have been socially influenced, to the point of herd behavior? How could we not have let the implications of machines that you write into being carry us off toward endless complexity?

This post isn't really about technology, but I will end by saying that like Ursula Le Guin did, I think hard times are coming. So efficiency in computing is going to become a lot more important than it is now, but at the same time, the severe C++ priests who demand it at every turn are also not the keepers of the one true path forward. When it comes to machines that are written into being, we will always have to think about tools, they'll never be a settled question. But we will have to learn what it really means to think about people first.