John Roth wrote: > Another thing that stands out: the explicit versus dynamic typing debate > has moved on from program correctness (which is a wash) to > other areas that explicit (or derived) type information can be used > for. I see this in PyFit: the languages where explicit type information > is available by reflection have cleaner implementations.
What makes myself a fat and lazy Pythonista is simply motivated by not wanting Guido throwing any language feature that someone found accidentally in another language into Python. My personal attitude is that I have enough self-confidence that I can live with just the second most popular programming language in the world. And maybe I'm also fat because I like to feed myself ( my topics of interest ) and not people with consumerist attitudes. So you have recognized that dynamic languages don't support type reflections well. Surprise, surpise! Other wellinformed authors like Benjamin Pierce wouldn't even agree about Python as being typed but dynamically *tagged*. Now what do you recommend? Sacrificing __delattr__, __setattr__, decorators and metaclasses for having colorcoding editors and a Java-like language with indentation syntax? What do you believe to ensure with type-declarations extracted by tools in presence of magic methods that enable to rewrite your class on the fly? Of course some things become harder in Python because they are impossible to guarantee in general. The PyPy team spends a lot of effort in exploring the boundary between what can be inferred from Python code and translated into lower level languages before program execution and what has to be passed to some code specializer at runtime. That's where new insights about dynamic languages may come from. Kay PS. I don't have anything against Ruby. May it have many cheerleaders! It's just that Haskell is on top of my list of interesting PLs I didn't devote much time yet and Ruby is far away. -- http://mail.python.org/mailman/listinfo/python-list
