It's a common situation that hash tables and other algorithms use only some bits of a hash so all hash bits must equally represent the object (currently not documented, filled as Issue 15380). It may be useful to require an algorithm working with hashes cut only highest bits (as current Druntime's associative array implementation does). This allows e.g. unique indexed elements use its index as a hash (if I'm not missing something regarding probabilities of usage patterns): size_t toHash() { return myIndex; } // Will be allowed
THIS ISSUE HAS BEEN MOVED TO GITHUB https://github.com/dlang/dmd/issues/17317 DO NOT COMMENT HERE ANYMORE, NOBODY WILL SEE IT, THIS ISSUE HAS BEEN MOVED TO GITHUB