4 ms·

> But it's always depressed me that nearly forty years later, our high-tech and ultrafast dynamic languages still cannot hold a candle to Smalltalk for clarity,

> But it's always depressed me that nearly forty years later, our high-tech and ultrafast dynamic languages still cannot hold a candle to Smalltalk for clarity, simplicity and usability. You're aware that these are extremely subjective attributes, right? Most people not using Smalltalk (or who tried it and decided to not use it) will argue that Smalltalk is anything but clear, simple or even usable (seriously, even Pharos or Squeak have the usability of late 90s IDE's). And of course, there's the fact that Smalltalk is dynamically typed, a class of languages that's fast headed for extinction.

AceJohnny2 11y ago
> Smalltalk is dynamically typed, a class of languages that's fast headed for extinction. I'm going to go out on a limb here, and chalk that one up in the "famous last words" column. :)
- machinshin_ 11y ago
  I dunno, a lot of modern language academic research and push are going towards a very compile-time-typing-everywhere mode. see Haskell and Idris, for instance
  - moonshinefe 11y ago
    ... which are known to drive major trends in modern programming language usage? Sorry, not seeing your point.
  - vidarh 11y ago
    And a huge proportion of language research gets ignored by the mainstream, so it's not a very useful indicator. For static typing to "win" (and though my favourite language these days is Ruby, I hope it does in the long run), it will need to become near transparent. If type inference is sufficiently advanced to avoid almost all type annotation, and the type system just gets out of your way 99% of the time, then yes, dynamically typed languages might be heading for extinction. But you'll have to wait a very long time for that, judging by the current crop of statically typed languages.
david-given 11y ago
Note that I was talking about the language, not the environment! I totally agree that Smalltalk IDEs and the general development workflow is poor (and is most likely the reason why it's so little known).

AceJohnny2 11y ago

> Smalltalk is dynamically typed, a class of languages that's fast headed for extinction. I'm going to go out on a limb here, and chalk that one up in the "famous last words" column. :)

machinshin_ 11y ago
I dunno, a lot of modern language academic research and push are going towards a very compile-time-typing-everywhere mode. see Haskell and Idris, for instance
- moonshinefe 11y ago
  ... which are known to drive major trends in modern programming language usage? Sorry, not seeing your point.
- vidarh 11y ago
  And a huge proportion of language research gets ignored by the mainstream, so it's not a very useful indicator. For static typing to "win" (and though my favourite language these days is Ruby, I hope it does in the long run), it will need to become near transparent. If type inference is sufficiently advanced to avoid almost all type annotation, and the type system just gets out of your way 99% of the time, then yes, dynamically typed languages might be heading for extinction. But you'll have to wait a very long time for that, judging by the current crop of statically typed languages.

david-given 11y ago

Note that I was talking about the language, not the environment! I totally agree that Smalltalk IDEs and the general development workflow is poor (and is most likely the reason why it's so little known).