4 ms·

I have not found in practice that programmers need to be mathematically inclined to become great software developers. Quite the opposite, in fact. To anyone u

I have not found in practice that programmers need to be mathematically inclined to become great software developers. Quite the opposite, in fact. To anyone unsure about the above quote. Jeff is full of shit. Often he dances on the edge of wrong advice, today he has leaped over it. You don't need to be a math whiz, but you damn sure need to be math inclined. At the very least you need logic. Any kind of graphics programming - math. But if like Jeff your career consists of simplistic apps in VB, well then I still wouldn't call you a great software developer without math. Don't aim to be at best average in your career, it makes for a shitty professional life, which is a big chunk of your whole life.

potatolicious 18y ago
Graphics programming is a very small subset of programming in general. You're basically saying "all programmers need math because some programming needs math"... There's plenty of cutting-edge stuff to be done that doesn't require an intimate knowledge of mathematics. I would have to say that being good at arithmetic seems to be pretty important in programming, though.
- biohacker42 18y ago
  Graphics, physics, all my work in biotech, all my work in video edition, all work I've ever done. Vast sectors of programming are math intensive. There are few which aren't. Also arithmetic and even more so logic pretty much = applied math.
  - catz 18y ago
    Other programming areas that require math: Pattern recognition, control systems, encryption, image and video compression (discrete cosine transforms in JPEG, discrete wavelet transforms in JPEG2000) any type of simulation (e.g. SPICE for electronic component simulation, RF simulation for electromagnetism, finite element models for mechanical systems) ... Any type of planning system requires math (e.g. convex optimisation, network routing, etc...) - these types of things have huge applications in areas such as industrial engineering and network routing. Basically all the interesting parts of programming require math.
ganley 18y ago
Thank you, HN. I started to comment on Jeff's post itself, but on the way down I saw so many people agreeing with him that I gave up in dismay. I'm happy to see the HN crowd has better sense. All I'll say is, if you lack math background, then you may well end up doing things like making mincemeat of trying to explain NP-completeness. Just to pick a random example (heh).
flatline 18y ago
I think that Jeff just takes math for granted. Without algebra, you would have a really hard time even understanding your run-of-the-mill business apps: order of operator precedence, the general notion of a variable, functions, etc. Let alone compound interest, expontential growth, etc. We're not talking combinatorial logic or multi-variable calculus here. If you complete your basic education in a first world country, you probably have some grounding in basic mathematics and logic, whether you realize it or not. Learning to program makes real a lot of these concepts that were probably sitting in the back of your brain unused for some time.

potatolicious 18y ago

Graphics programming is a very small subset of programming in general. You're basically saying "all programmers need math because some programming needs math"... There's plenty of cutting-edge stuff to be done that doesn't require an intimate knowledge of mathematics. I would have to say that being good at arithmetic seems to be pretty important in programming, though.

biohacker42 18y ago
Graphics, physics, all my work in biotech, all my work in video edition, all work I've ever done. Vast sectors of programming are math intensive. There are few which aren't. Also arithmetic and even more so logic pretty much = applied math.
- catz 18y ago
  Other programming areas that require math: Pattern recognition, control systems, encryption, image and video compression (discrete cosine transforms in JPEG, discrete wavelet transforms in JPEG2000) any type of simulation (e.g. SPICE for electronic component simulation, RF simulation for electromagnetism, finite element models for mechanical systems) ... Any type of planning system requires math (e.g. convex optimisation, network routing, etc...) - these types of things have huge applications in areas such as industrial engineering and network routing. Basically all the interesting parts of programming require math.

ganley 18y ago

Thank you, HN. I started to comment on Jeff's post itself, but on the way down I saw so many people agreeing with him that I gave up in dismay. I'm happy to see the HN crowd has better sense. All I'll say is, if you lack math background, then you may well end up doing things like making mincemeat of trying to explain NP-completeness. Just to pick a random example (heh).

flatline 18y ago

I think that Jeff just takes math for granted. Without algebra, you would have a really hard time even understanding your run-of-the-mill business apps: order of operator precedence, the general notion of a variable, functions, etc. Let alone compound interest, expontential growth, etc. We're not talking combinatorial logic or multi-variable calculus here. If you complete your basic education in a first world country, you probably have some grounding in basic mathematics and logic, whether you realize it or not. Learning to program makes real a lot of these concepts that were probably sitting in the back of your brain unused for some time.