in reply to What is simply not possible without static typing?

"Simply not possible" is too strong, but in a dynamically-typed (or "run-time typed") language, you need to do these things in a different way. With a statically-typed language, you can analyze the program text without running it to figure out all the identifiers' types. With a dynamically-typed language, you have to look at a running instance of the program. You *can* just look at the program text, but at that point you're just guessing.
  • Comment on Re: What is simply not possible without static typing?

Replies are listed 'Best First'.
Re^2: What is simply not possible without static typing?
by locked_user sundialsvc4 (Abbot) on Apr 30, 2011 at 13:10 UTC

    When you don’t have static typing, what you really miss the most is the ability to make data-type related assertions and to do so at compile time.   But there are many ways to implement that, which do not unduly restrict your flexibility in handling data at run time.

    At run time, the notion that “a piece of data is a self-describing object” really is convenient, and it really is efficient.   Sometimes much more convenient than the static-typing of the days of yore, when computers were much slower and smaller and “you did what you had to do in the very-tiny shoebox that you had.”