Comment by greenspot | Hacker News Reader

Comment by greenspot | Hacker News Reader

Books are tempting, they have glossy covers, they are about mastering a subject. All things which lure us into buying them with 1 click on Amazon.

But I learn way more if I find a problem I want to solve. Then, I find my way through countless online resources and it's so much more effective than reading a book from beginning to end (which I can't do anyway).

Just an example: when I got into JS Promises/Generators/Iterators/Async/Await I read more than 100 online sources. I read multiple articles about the same topic again and again just because I didn't like the example of one author or there was an interesting detail in the other version. Getting just one book, means just one author and yes I love it to dive into one book on the plane. But nowadays with the Internet books are overrated. I remember going through the book store when I was a kid and buying an Assembler book for my Commodore Amiga for incredible 60 bucks—this was my only resource to knowledge next to print publications, they were expensive and maybe these memories make books still valuable to us.

Another thing I recently realized: reading source codes of popular libs (my last one was Koa). Once you get used to it it's not just very helpful but it teaches you so much and often there is an underlying goal (you want to use a lib you want to fully understand).

Books are not bad and for getting into something entirely new there are maybe the best start because in the moment you buy them you commit yourself to learn something from beginning to end. That's good and sometimes really necessary to fully comprehend new languages for example.

> Books are tempting, they have glossy covers, they are about mastering a subject.

A lot of the books recommended on that site are about expanding horizons rather than mastering a subject. You google for an answer when you know what you don't know. Books can be great for those times when you don't know what you don't know. :)

> reading a book from beginning to end (which I can't do anyway)

This doesn't concern you, in the slightest?

The benefit to books is that the best books are significantly better curated than the best pages on the internet, especially as the topic gets increasingly specialized or esoteric. Sure, you shouldn't buy the JavaScript API if you want to look up documentation or skip stackoverflow if you want to quickly grasp a concept, and I'm not saying you should buy a textbook every time you want to learn a new library, but mastering most technical subjects are only accessible from a thorough explanation. Good books have the most thorough explanation.

One point in defense of books is that by having you focus your attention for an extended period of time, they can unfold an interesting and coherent perspective on their topic. Of course this only holds for good books and most books---like most anything---aren't that good.

Also, not all topics are necessarily well suited to books. I agree it's probably better to browse varied sources online to learn how to use async/await. But I can't imagine even approximating the experience of reading Structure and Interpretation of Computer Programs through random-ish web browsing. Though both involve reading, these examples are really different sorts of activities, each well supported by different features of the product design of the book and the web.

> But I learn way more if I find a problem I want to solve. Me too, that's why I only keep recommendations in a list, that I don't get/buy right away. However, many people have lists of 'totally essentials' that I try to read first (e.g. Jeff Atwoods recommending Code Complete)

> the moment you buy them you commit yourself to learn something from beginning to end Well, not always, I remember studies (can't find the source) where the majority of the books bought are never even started. That's why I like subscriptions better, where you can just open a book to the parts that interest you. Writing a book involves much more work than a blog post, and it's also motivated by other factors. Specially in technical books, I've heard the rewards are little, so authors usually write a book to gain reputations, which is fine, but definitely requires much more thought and effort, and some times you can tell. Relevant articles by Atwood: http://blog.codinghorror.com/programmers-dont-read-books-but... http://blog.codinghorror.com/do-not-buy-this-book/

different people have different styles of learning. for some, building an overall idea first is more important than hacking something together. for other hacking something together is more important than getting the full overall picture. stick with whatever works for you + be open to do things differently once in a while :)

Reply on news.ycombinator.com