> developers do not know that they mustn't call
> datetime.now()
Plenty of languages where it's easy to mock the equivalent to that call.
Re:Testable IO in Haskell
by crdoconnor on Sunday June 21, 2015 @09:48PM
(#9755752)
I don't have a problem with it. I send a mocked time to the application under test using libfaketime using this: http://hitchtest.com
Et voila, no more time issues.
Wouldn't it be nice if you could just obstruct such patterns within the first millisecond of their first interaction with the compiler?