Created attachment 669502 [details] [diff] [review]
Comment on attachment 669502 [details] [diff] [review]
This changes the !AllocatedData() case to returning null instead of throwing. Why is that desirable?
Alternately, can we nix the whole AllocatedData() thing and require infallible construction for funcstringlists?
r=me modulo that.
Should this have a test?
Maybe, but it's only testable via C++; JS-observed behavior shouldn't have changed here. And I'm not sure how much we care about the C++ behavior of these methods. ;)