I wasn't disputing this in my previous comment, but I do actually disagree with it at this point. I haven't found a use-case for a dynamic language in many years. I cringe at the thought of working with them now.
> vanilla JS in a text editor is the best tradeoff
This defies my understanding of the word "tradeoff". It costs nothing to fire up a full IDE instead of a text editor, so why not use one? It comes with autocomplete at the very least, but also static analysis of JS.
If something costs nothing and has a benefit, it must be the best tradeoff, right?
> Any bug that would have been catched with typing will obviously be catched visually.
Catching bugs at runtime is slower by definition than catching them while you're writing the code in the first place. Why make things harder when great tooling is a few clicks away?
You are focusing on the IDE aspect which I do not necessarily disagree with, honestly I haven’t tried the recent offerings for Ruby or JS so maybe it’s 0 cost now.
> I haven't found a use-case for a dynamic language in many years.
That’s actually my initial point. The use case is small scope projects, where you get the benefits of dynamic languages (faster and easier write and easier to read) without the maintainability cost and where you’re unlikely to ever get a type error.
Edit: concrete example: Standalone web page with a GET request to get some data, create a chart, and some user events like mouseover. No build tool chain, plain JS. I don’t see how I could benefit from types and what kind of errors would be prevented, but maybe the overhead is now very low with recent tooling and I should re-evaluate.
> I haven't found a use-case for a dynamic language in many years
This seems like a rather extreme statement given the number of people in the mathematics and science communities that find plenty of use cases for Python/Scipy/Numpy/Pandas/Tensorflow, etc.
What language ecosystem would you suggest for those people?
I wasn't disputing this in my previous comment, but I do actually disagree with it at this point. I haven't found a use-case for a dynamic language in many years. I cringe at the thought of working with them now.
> vanilla JS in a text editor is the best tradeoff
This defies my understanding of the word "tradeoff". It costs nothing to fire up a full IDE instead of a text editor, so why not use one? It comes with autocomplete at the very least, but also static analysis of JS.
If something costs nothing and has a benefit, it must be the best tradeoff, right?
> Any bug that would have been catched with typing will obviously be catched visually.
Catching bugs at runtime is slower by definition than catching them while you're writing the code in the first place. Why make things harder when great tooling is a few clicks away?