What this article essentially says is one kind of math is better than the other.
No.
The Mathematical Hacker
81–90 of 135 posts
Re: The Mathematical Hacker
#82I'm in the Yegge school, so far as I think that the mathematical ignorance and, more generally, anti-intellectualism of our industry is its downfall. We see it in the lack of design sense and the awful code that is produced. VisitorFactory nonsense is something that was invented by people who hated math and wanted to tear programming away from its mathematical/problem-solving roots with a bunch of junk complexity tha…
There aren't enough humans that can do math, and there are more openings for jobs than intellectuals. Programming is not necessarily a job for intellectuals, any more than painting or automotive repair is reserved for intellectuals. Have you heard of the balmer peak? This stuff we call code ain't that hard. I don't like your complexity argument. If you are a gear in a watch you do not deal with complexity. You are we…
The Ballmer Peak was a fictional concept in an XKCD cartoon[0]. Hardly something to use as evidence in an argument.
But even if it were scientifically accurate, alcohol doesn't make you dumber (at least not at first) - it makes you less inhibited and slower-thinking, two traits that may actually be desirable in problem-solving.
Re: The Mathematical Hacker
#83This rings True to me. It also represents potentially really good news, because it means that we are probably at the beginning of a huge wave of wealth creation as we figure out how to apply mathematical knowledge to the massive growing mountains of data generated by everyone and everything everywhere. Think about the historical impact that statistical-quality-control software has had in manufacturing, or the impact…
Yes! This is what had me thinking that this article is the best one I've read all week. It's not about language wars, it's about finding the new applications of math. With the data and computational power at our disposal, big questions are going to fall and big industries will be disrupted.
Re: The Mathematical Hacker
#84I find it disturbing that this article has been written in 2012. While I was reading it I really thought that it was at least 10 years old. Computer science researchers are doing actual mathematics, and they are clearly more in what the article calls "Lisp school" than the "Fortran school". Research in functional programming is mostly mathematics. Lambda calculi (which was originally not developed to be a programming…
Ironically, though, FP is perhaps the worst programming paradigm for actually performing numerical computations.
Math is a lot more than numerical computations, however. If you're modeling abstractions like vector spaces (as real mathematicians often will want to do), then functional languages are an excellent choice.
Re: The Mathematical Hacker
#85A fine essay, except that it ignores that the "workaday" programmer actually hates high degrees of abstraction and ignores it.
No, seriously.
We're taught in mainstream programming languages to temper our desire for abstraction with mechanical sympathy; being careful not to build too high a tower of abstraction lest our CPU god grows angry and knocks us down to earth with miserable runspeeds. Even if your algorithm is right, failing to fit in cache or having too many memory fetches can increase your code runtime by multiple orders of magnitude from where an optimal, machine-aware piece of code should be.
And the languages that do try to bring more abstraction to the table are ironically dismissed as "too academic" or "too strange" or "not practical" by most people in the industry.
Re: The Mathematical Hacker
#86Earlier quoted context omitted.
>There aren't enough humans that can do math On what basis do you make that assertion? "Lots of people program who didn't bother to learn maths" doesn't mean they aren't able to learn maths. >This stuff we call code ain't that hard. Then why do we so consistently fuck it up horribly?
> Then why do we so consistently fuck it up horribly? This is the crux of it. Coding these days "ain't that hard" because the current generation of programmers sneer at formal methods. They believe providing even semi-rigorous proofs that their code works is not worth their effort. The result? Testing code with arbitrary input is the best practice; even semi-rigorous proofs are a "luxury". Would your high school geom…
I'd say blaming programmers is a bit wrong, blame the business; most businesses don't pay programmers to write good code, they pay for fast code that generally works. They don't want to pay too much because how hard can it be to put database data onto the screen. Most programming simply doesn't require math beyond elementary arithmetic. You think most programmers have time for formal proofs?
Re: The Mathematical Hacker
#87I'm in the Yegge school, so far as I think that the mathematical ignorance and, more generally, anti-intellectualism of our industry is its downfall. We see it in the lack of design sense and the awful code that is produced. VisitorFactory nonsense is something that was invented by people who hated math and wanted to tear programming away from its mathematical/problem-solving roots with a bunch of junk complexity tha…
>anti-intellectualism of our industry is its downfall I just curious to find out, what downfall are you talking about ? Isn't software eating the world and Silicon Valley making billions ?
But there's an astounding lack of mathematical rigor in software. It may all call crumbling down like a house of cards. And it does. People don't know theory, nor do they want to learn it. They just want to get work done. Although all theory is dangerous too, I think a healthy mix is a good order.
Re: The Mathematical Hacker
#88The problem with the article is that it dreams up a false conflict and then buries its thesis, a very valid and solid one, by trying to force a contentious narrative onto that just isn't there. Very few people these days are ignorant of the points he makes and neither Yegge nor Graham are posts of some imagined counter applied math camp. I would tend to think they are pro - Yegge's interested in bio, Graham did spam…
To add to your claim that math is fundamental to computer science, I think there's a big factor that isn't being mentioned: knowing math future-proofs your career better than knowledge of any technology or language. Mathematical depth is the irreducible core of a computer scientist's value. Fundamentally, what computers provide is the mechanization of knowledge. In the same way that robotics came along and mechanized…
I'm not sure I completely agree with you, but these are some valid points. What happens once we run out of unskilled jobs due to the machines?
Re: The Mathematical Hacker
#89Can we at least restrict the word "hack" to using something in a way it was not intended? It's being thrown around for everything and it's downright confusing.
Re: The Mathematical Hacker
#90"
long int fac(unsigned long int n) {
return lround(exp(lgamma(n+1)));
}
"
Sounds great! What's the factorial of 40? Let's use some python:
>>> int(math.exp(math.lgamma(41)))
815915283247882431423526575245034982027017846784L
Huh, weird. I would've thought the factorial of 40 would be divisible by ten. Let's try it another way:
Prelude> let fac n = foldr1 (*) [1..n]
Prelude> fac 40
815915283247897734345611269596115894272000000000
The author's disdain for the lambda calculus, as if it is somehow not real math (because it doesn't involve numbers in the familiar sense?), is bizarre.