1. 10
  1.  

  2. 5

    Numerical math is very fun, very ‘unfinished’, and a good mix between computer science and math. It is mostly the bridge between theoretical math and computations, and the reason that implementations of mathematical methods keep changing.

    1. 3

      And that especially hits data science. I want JavaScript to be a contender for data science in the browser, and – amongst some other issues, like number types and a confounding lack of a commonly-used data-frames library – an inability to produce replicable results means adding more crisis to the replication crisis in the sciences.

      I think the author’s heart is in the right place, but isn’t it reasonable to require that someone who is concerned about numerical accuracy actually use an environment that’s built for that, as opposed to having the convenience of the tools in the browser?

      The examples given would scare the heck out of me if I were concerned with validation and accuracy.

      1. 3

        I am concerned about how this affects JS crypto libraries… Does anyone have any insight into this? /cc @nickpsecurity

        1. 1

          You mean apart from “tests that break for bad reasons tend to be skipped sooner or later”?

          1. 1

            If you’re interested in an answer from @dchest, check out this issue: https://github.com/dchest/tweetnacl-js/issues/190