and yet every single project to reimplement vim "cleanly" has failed. the only successful other implementation out there essentially ported the vim codebase directly to java. i can only conclude that vim has built up a tremendous amount of sheer value along with the cruft that is not easy to reproduce.
He's not trying to reimplement vim. He's doing a new thing. Part of vim's cruft is in its design. An important part of a clean break is a new interface. Obviously, getting traction with other devs will be difficult, exactly like any other tool.
