Hi Peter, > ...even more Stream-y (with JDK 9 changes to Optional): > > otherMods > .stream() > .flatMap(mod -> finder.find(mod).stream()) > .forEach(mref -> ...); > > > Yes, it takes some time for people to get used to new tools before they use > them optimally. This is a normal learning process. So I wonder whether it is > only the method name that is to blame for the observed percentage of wrong or > sub-optimal usages of Optional or would it be more-or-less the same with a > more descriptive name at this time. > > But the argument that getWhenPresent() is easier to spot than get() when > reading code is on spot so at least some bugs would get spotted more quickly > if the method stood out.
Even when you call this new method Optional.dudeCallIsPresentBeforeCallingMe() it won't help. Against the stupidity, the inability to read JavaDocs and the laziness helps only a compiler error. Therefore when you really want to help developers to avoid bugs due to wrong usage of Optional.get(), then you must change the compiler to recognize the usage of Optional.get() without the isPresent()-check and to raise an error in such cases. I mean if IntelliJ is able to give you a hint that you use Optional.get() without the isPresent()-check, then the Java compiler should be able to do it too. Just my two cents. Best regards, Andrej Golovnin
