On: Type inference has usability problems

[azhenley]

Running experiments would give us data on whether it is personal or not, and to what degree.

I know a lot of teams that conduct code reviews outside of the IDE, so your suggestion wouldn’t help those cases.