I think the author's got a kernel of an idea and does an absolutely horrible job outlining it, considering few here seem to be catching on. Localhost will like

I think the author's got a kernel of an idea and does an absolutely horrible job outlining it, considering few here seem to be catching on. Localhost will likely never die for certain sectors of programming. There's a significant chunk of stuff that doesn't need to be custom localhost setups, though - and ultimately it can smooth out the differences between dev/production in teams where this is a stumbling block. You don't need localhost to be some server somewhere that kills your ability to work with an outage - stuff like StackBlitz (https://stackblitz.com https://stackblitz.com) effectively uses WebAssembly to run e.g Node directly in the browser sandbox. This model is the first one that makes me think it's possible. You no longer have that dangling issue of "which of my NPM packages is going to try and steal from me" since it's walled off in a sandboxed environment that spins up faster than you can blink. The caveat to all of this is that we should fundamentally never let localhost die, as ceding total control to services to build on is a pretty bleak future. There's a good middle ground to find here though. A final aside: an article telling me I'm wrong for wanting to code while flying is a junk take.