Other errors besides Error should probably be able to have causes?
domenic opened this issue · comments
Is there a reason the proposal is restricted to Error? I think TypeErrors and RangeErrors and the like should all be able to have causes.
I’d indeed expect them all to have it, since they all inherit from Error (and an idiomatic implementation of each would call super).
The proposal is started with minimal changes. If adding cause to *Error constructors is generally acceptable, I believe it is a good idea to have them all.