No out of bounds exceptions, and you can prove shit like the idempotence of a swap operation repeated twice on the same pair of array elements. Good stuff.
Name:
Anonymous2014-12-02 15:51
Why would you do a swap operation three times?
Name:
Anonymous2014-12-02 15:53
C has mutable arrays too. Haskal has mutable arrays too. Even Ada/SPARK has mutable arrays.
>>8 It's not about better or worse now, it's about better or worse in the future. And if dependent types can tackle such a seemingly imperative thing as mutable arrays, they have a chance to become the better option all-around.
Name:
Anonymous2014-12-02 18:07
>>9 Every time I see dependent typing enthusiasts prattling about their languages, they're talking about the types themselves. I don't see anyone talking about building software in these languages. It's suspiciously exactly the same as the OOP hype back when that was a thing.
Name:
Anonymous2014-12-02 18:24
>>10 Because the whole idea is that types come first. Development is interactive, you write down your types and the IDE writes the implementation for you, with some holes that you fill. I.e. types define not only the specification of code, but also in a large part the implementation itself.
Name:
Anonymous2014-12-02 18:34
>>11 But that's ridiculous. You write the types, and then you still have to actually write the program? Types are supposed to be there to speed up performance. No other reason.
Name:
Anonymous2014-12-02 18:54
>>12 How are things back in the 80s? Still dancing to that disco?
Name:
Anonymous2014-12-02 18:56
>>13 You can do better than that. Lisp has been around a little more than 30 years.
Also remember that Haskell could lead to typetheoritis and heavy dependent type disorders. Check your child for signs of Haskell today, and take prompt action at first sign of monad disease to prevent zygohistoparamorphitis.