| I thought it was a good thread. For those unaware of the missing context, what he's talking about is how in the 2010s, progressive enhancement — long seen as a webdev best practice — was largely replaced by a JS-first approach by JS frameworks in practice. It does seem like the JS-first bubble is deflating a bit, but I'm not sure it's on its way to fully popping like the Flash bubble did. I do agree it was a lost decade though. More than a decade. I miss when everyone agreed progressive enhancement was the way to go. It's entirely possible to build a SPA without abandoning progressive enhancement, but the frameworks do not encourage those best practices. As such, it's rare to see a web app built with a framework that doesn't create a hard dependency on JavaScript or isn't an accessibility disaster. When I reflect on the last decade of frontend development, the lesson I draw from it is everyone is susceptible to ill-conceived fads, including people who think of themselves as evidence-driven. We're good at convincing ourselves we're objective, especially at times when we're not. What's really depressing is this tendency applies to all the applied sciences. Tons of people who think of themselves as motivated solely by evidence fall for terrible fads in their field all the time, including, terrifyingly, in areas like medical practices. I think we'd all do better to spend less time emotionally attaching to our tools and more time looking at evidence and metrics in a dispassionate way. As Paul Graham once said, keep your identities small. A great book that's vital to grokking this stuff is The Scout Mindset by Julia Galef. She argues we should always just go where the evidence leads. But our monkey brains are bad at this, so it takes a lot of conscious effort or we'll do it poorly. I think JS framework mania is yet more evidence of her thesis. |