4 ms·

> I never really used math again after that, even after 20 years of programming. Maybe some arithmetic here and there for bills and to split a Venmo dinner. Tha

by Teever 3y ago

> I never really used math again after that, even after 20 years of programming. Maybe some arithmetic here and there for bills and to split a Venmo dinner. That's about it. I don't believe this at all.

kr0bat 3y ago
Really, this sounds believable. How often does the average programmer use anything beyond arithmetic? I'm only a few years out of school, but I'm re-learning calculus since I haven't touched it once since college.
solardev 3y ago
I'm not saying that no programmers use math, but I don't -- as a run-of-the-mill frontend web dev, it's just not something I regularly encounter. The overwhelming majority of my work is composing JSX components out of other people's JSX components, spending some time in Figma or Illustrator, and then going back to the JSX to tweak some pixels. In an average month (hell, average year), the most math I do is increment some versions in package.json. Maybe the occasional CSS calc(). I'm sure if you're doing games or algorithms or data science, it'd be different, but there are many programmers who work in boring business logic, APIs, databases, etc. Relationships and concurrency and graphs, yes, but not at the lowest levels where the math is actually implemented, unless you're the person writing those libraries/frameworks (in which case, thank you!). And most people don't become any kind of programmer. I think the most advanced math I ever did was figuring out how to rotate a polygon in place, but even then we used a library and I just copied and pasted a Stack answer (https://longviewcoder.com/2020/12/15/konva-rotate-a-shape-around-any-point/ https://longviewcoder.com/2020/12/15/konva-rotate-a-shape-ar...). That was before GPT. I don't remember how to do that now, or how a sin is different from a tan, except one you confess for and the other gives you cancer. Another example was a line graph where we had to comb through several tens of thousands of data points in Javascript. Our initial implementation was terribly slow and I was tasked with improving it. I started reading about binary trees and b-trees and O(n) notation, but I had no idea what any of that meant -- I couldn't even understand the first paragraph on Wikipedia (https://en.wikipedia.org/wiki/Big_O_notation https://en.wikipedia.org/wiki/Big_O_notation). I ended just copying some example from Stack, the search became much faster, and that was that. I never understood the math (nor do I really care to, if I'm being honest). And get this: Among the frontend devs I've worked with (again, just in regular small/med companies, not FAANG or any glorious science work), I actually had the strongest math background among them. That's only because I was an environmental science undergrad, and some of the introductory stats stuff transferred over into analytics (which, again, is something I just copy and paste from Stack or Lynda or whatever without ever actually understanding how a Pearson's correlation is different from any other correlation). To me and the people I've worked with, math was just a problem to be outsourced to someone else's implementation, same as any other NPM package or Stackoverflow answer. We worked in abstractions on top of abstractions in service to mid-managers and business needs, not in pursuit of algorithmic elegance. In my mind, that's another reason why web dev is often considered the low-hanging fruit/bastard child of "programming", still more akin to lightweight scripting and Dreamweaver/Frontpage than any real comp sci. There's a reason why 6 months of bootcamp used to be enough to do the job. ------ Just to support your point though, I once interviewed for a Canvas-based web app (as in the HTML element that lets you draw pixels), a mapping system similar to Google Maps. Their interview coding test was to make an analog clock with animated hands. I completely failed the test and was not hired. Maybe if I knew more math and algorithms :) Alas. But in the rest of my career, math was not important at all.

by Teever 3y ago

kr0bat 3y ago

Really, this sounds believable. How often does the average programmer use anything beyond arithmetic? I'm only a few years out of school, but I'm re-learning calculus since I haven't touched it once since college.

solardev 3y ago

I'm not saying that no programmers use math, but I don't -- as a run-of-the-mill frontend web dev, it's just not something I regularly encounter. The overwhelming majority of my work is composing JSX components out of other people's JSX components, spending some time in Figma or Illustrator, and then going back to the JSX to tweak some pixels. In an average month (hell, average year), the most math I do is increment some versions in package.json. Maybe the occasional CSS calc(). I'm sure if you're doing games or algorithms or data science, it'd be different, but there are many programmers who work in boring business logic, APIs, databases, etc. Relationships and concurrency and graphs, yes, but not at the lowest levels where the math is actually implemented, unless you're the person writing those libraries/frameworks (in which case, thank you!). And most people don't become any kind of programmer. I think the most advanced math I ever did was figuring out how to rotate a polygon in place, but even then we used a library and I just copied and pasted a Stack answer (https://longviewcoder.com/2020/12/15/konva-rotate-a-shape-around-any-point/ https://longviewcoder.com/2020/12/15/konva-rotate-a-shape-ar...). That was before GPT. I don't remember how to do that now, or how a sin is different from a tan, except one you confess for and the other gives you cancer. Another example was a line graph where we had to comb through several tens of thousands of data points in Javascript. Our initial implementation was terribly slow and I was tasked with improving it. I started reading about binary trees and b-trees and O(n) notation, but I had no idea what any of that meant -- I couldn't even understand the first paragraph on Wikipedia (https://en.wikipedia.org/wiki/Big_O_notation https://en.wikipedia.org/wiki/Big_O_notation). I ended just copying some example from Stack, the search became much faster, and that was that. I never understood the math (nor do I really care to, if I'm being honest). And get this: Among the frontend devs I've worked with (again, just in regular small/med companies, not FAANG or any glorious science work), I actually had the strongest math background among them. That's only because I was an environmental science undergrad, and some of the introductory stats stuff transferred over into analytics (which, again, is something I just copy and paste from Stack or Lynda or whatever without ever actually understanding how a Pearson's correlation is different from any other correlation). To me and the people I've worked with, math was just a problem to be outsourced to someone else's implementation, same as any other NPM package or Stackoverflow answer. We worked in abstractions on top of abstractions in service to mid-managers and business needs, not in pursuit of algorithmic elegance. In my mind, that's another reason why web dev is often considered the low-hanging fruit/bastard child of "programming", still more akin to lightweight scripting and Dreamweaver/Frontpage than any real comp sci. There's a reason why 6 months of bootcamp used to be enough to do the job. ------ Just to support your point though, I once interviewed for a Canvas-based web app (as in the HTML element that lets you draw pixels), a mapping system similar to Google Maps. Their interview coding test was to make an analog clock with animated hands. I completely failed the test and was not hired. Maybe if I knew more math and algorithms :) Alas. But in the rest of my career, math was not important at all.