This MR optimizes views compilation by compiling them together, and then extracting one by one at mini_c stage (after this, another round of Self_mini_c.all_expression is applied to simplify remaining code, but less rounds are needed, as they were "optimized" together previously).
-
has a changelog entry