3 ms·
You missed the point, so I'll repeat it. The parent I was responding to made the comment that python has type hinting, responding to OP's criticism of Python's
by computerex 8y ago
You missed the point, so I'll repeat it. The parent I was responding to made the comment that python has type hinting, responding to OP's criticism of Python's lack of static typing. My point was, there was no notion of type hinting in python until PEP 484 or whenever it was released, long after the initial release of python. Python was released in 1991 and type hinting was introduced in like 2014. So for the 23 years the language existed prior to type hinting, people made all sorts of arguments/excuses for why that's not needed/unnecessary for Python. Hell, even now as a consequence of this type hinting in python isn't all that common. It may have to do with the fact that for 85% of the language's lifespan, there was no type hinting and duck typing was rampant, which like I said is an approach that goes against type hinting.
So if you want people to be pythonic in py3, they have to make massive changes like replacing duck types with abstract classes.
I think it's weird that you don't question why vast majority of python code has no type hints. Do you have experience in Python?
- duckerude 8y agoIt makes sense now, thank you. I think the word "pythonic" was misused in the parent - usually it's normative, but type hints are strictly optional. I don't think I've seen anyone argue that Python 3 code without type hints is unpythonic. Except maybe the parent, but I think that's just unfortunate wording. It's explicitly using a non-standard definition of pythonic ("part of the language").