September 29th, 2018

(no subject)

I watched the hearing of Brett Kavanaugh, and as much as I liked the beginning of his opening statement, I hated the way he evaded direct questions (filibustering they call it?). To me, that would have immediately been a red flag. Not because it necessarily means he is guilty, but because it shows lack of good will. A violation of the principle of cooperation in conversation. Compared to him, many senators were much more frank, direct, and precise in their language.

Also, minor linguistic notables:
- when asked "Do you wish she never came forward?", he couldn't bring himself to simply say "yes" (although what could possibly be wrong with wishing that, if his position is that she mistook him for someone else?) (link)
- when asked whether he thinks his accuser is a political pawn, he again couldn't bring himself to say "yes" (there was in that exchange an implied — incorrectly I think — contrast between having a real traumatic experience and being a political pawn, as if, even if she had been assaulted, she couldn't now be used, like a pawn, as an instrument against his nomination)
- when asked "are doctor Ford’s allegations true" he cannot bring himself to plainly say "no, they are not", the best he can manage is "they are not accurate as to me" (which is the same as saying that they are not true, but in a more roundabout way). I was a bit surprised to hear them repeatedly say "her truth" (and, perhaps, also "your truth"), as if declaring that there are multiple truths, each in its way valid.

(no subject)

In his mini-course Build Your Own RxJS Pipeable Operators, John Lindquist starts off by overriding an undocumented method in the Subscriber class:


class DoubleSubscriber extends Subscriber {
  _next(value) {
    console.log(value)
    this.destination.next(value * 2)
  }
}



(link)

No-one, without looking into the source code of RxJS, would even know that the Subscriber class has a _next method or a destination object. It is not in the public contract of the Subscriber class.

I wonder if it's a good practice to even touch the internals of a library, let alone build teaching materials that rely on this.