I tend to agree with this. We need a way to natively address the complete set of each platform's capabilities without needing to worry about them falling out of lockstep if some centralized maintenance entity decides they're no longer sufficiently incentivized to maintain their social contract with their framework adopters. It may indeed be the case that we can already accomplish this with properly written (concrete implementation agnostic) modular code but fmpov there's no complete solution being offered to the public at present. My hope is that in the next few years we get to see Google adopt Swift. Then it's just a matter of reinventing the browser so we can build native web apps in Swift as well. :P
I'm curious if this will change now that webapps/electron-apps have become so popular. It may be that users have begun non-native apps as the new norm.
In five years the Electron craziness will be gone.
And, judging from what was going on 5 years ago, whatever replaces it will be even crazier and trendier :)
My bet, browser will be reduced to yet another general purpose VM, with the new trend being WebAssembly + Canvas/WebGL.
Or back to pure native.
what will replace it?
Any widget toolkit, on any platform ever, that has tried to provide it's own cross-platform widgets, has been completely terrible and ends up being native-nowhere instead of native-everywhere.