Re:Ask HN: How do you read programming books?
by scoggs on Sunday November 19, 2017 @05:50PM
(#15736392)
Right. I think learning this way forces yourself to solve the problem by teaching the discipline to yourself. It's always said that you don't truly understand something until you are able to teach it to another so in that sense you a sort of "tricking" yourself into teaching it to yourself as you learn it. It's a round about way of thinking / doing things but those mental gymnastics feel necessary to me to absorb such unnatural and abstract (to me) topics.
Re:Ask HN: How do you read programming books?
by justbaker on Monday November 20, 2017 @09:59AM
(#15740278)
>The only way to learn something (and certainly programming) at a high level is to force yourself to write down a well- defined, closed-end problem, and then sit down and solve the motherf_____r. You learn what you need to learn, build, and invent along the way, and that survives longer than the original problem.
This is how I've been successful as a self-taught developer.
Re:Ask HN: How do you read programming books?
by pgathogo on Wednesday November 22, 2017 @04:06AM
(#15755426)
I tried this technique with C++ but soon I realized it was a dead end. You try to build anything with C++ without some inner understanding of concepts like const, templates and STL, and I can bet things will blow up pretty fast.
Re:Ask HN: How do you read programming books?
by nurettin on Sunday November 19, 2017 @11:07PM
(#15737709)
This is how I learned most of my mathematics before highschool. I just made up a problem such as calculating the angle which minimizes the time to intercept a slowing
rocket which taught me a lot about differentiation.
However, I didn't become amazingly successful in life.
The only way to learn something (and certainly programming) at a high level is to force yourself to write down a well- defined, closed-end problem, and then sit down and solve the motherf_____r. You learn what you need to learn, build, and invent along the way, and that survives longer than the original problem.