I agree with some things. But on the other hand people started using npm (or bower) because they were tired of manually downloading and linking dependencies. Of

I agree with some things. But on the other hand people started using npm (or bower) because they were tired of manually downloading and linking dependencies. Of course you had to do like 5, but now that everything is on npm it turns out that those 5 have 10, and those than have 20, so... CSS preprocessors I don't like that much either, but having an autoprefixer is actually nice. The alternative could be in the future to just not use it. Maybe we're already there? It used to be difficult to get anything done. It was kind of, at first figure out how you will even get this complicated design to work. And every time you thought you had it, it turned out no, something broke in IE. This part got progressively better, to the point that I don't really test in multiple browsers anymore. But as things got easier, we found ways to make it complicated in other areas. And we can do nice things, but sometimes they don't add that much, yet cost a lot in terms of effort. There is some pain and friction with colleagues that always want to do something new because others do it. When you get older you stop caring about that, and want to keep things simple and replace things that are an actual problem. As others have pointed out, this is the same in any area. We had a discussion what to do a couple years ago, and one colleague was dead bent on using GraphQL for an API. He's good at his job, but the answer to why was basically "because others do it". We didn't and we're fine -- we didn't need it. So you're going to get that pressure to learn new things everywhere. I don't have a good answer to this other than focusing on learning things that will last, even beyond this one area of work. But figuring out what that is can also be difficult.