Blog Post

YOW! 2012: Kevlin Henney - Cool Code, Modern Programming, Big Problems in Computing

Sign in to queue

The Discussion

  • User profile image

    80% of videos start with already.

  • User profile image


  • User profile image

    @dupa: Yep. Signifies that we're rolling...



  • User profile image

    In long run the future belongs to higher order and more abstract languages.

    Let us consider the problem of solving the equation ax^2+bx+c=0.
    We can write that program in many different form (depending on the programming language).
    However, the mathematical content is exactly the same.
    Maybe in the future the code optimization will allow to find the optimal representation of given problem, which is not dependent on the programming language (in theory we can do this at the machine code level).
    At present such tools don't exist. Because of that in order to get more optimal code it is necessary to use lower level programming languages. However, if the code optimization process will be good enough, then the choice of programming language will be irrelevant.

    End users do not care about the programming language.
    The software is distributed usually as a machine code.
    The real question is ... who is better in code optimization ... developers or code optimization tools.
    So far developers are better (because of that we still need lower level programming languages). However, I am not so sure about that in the future.

    Development process is much faster in more abstract programing languages.

    Well ... but today ... we still need to say that the most important are developers, developers, developers, ... and lower level programing languages.

  • User profile image

    There's some people at MIT that appear to think that the abstractions will break down and we need to program in something like RALA for optimal performance on future hardware - but let's hope not.

Add Your 2 Cents