Speaking JavaScript

If you’d take the time to read the free online copy (which OP links to) you’d see that there are plenty of “pure examples” in the book. In fact, there’s a code example for every two paragraphs of text, or so.

Give Underhanded JavaScript a try then. It teaches you how to be an arsehole JS developer :P - published today too (http://leanpub.com/underhandedjavascript/c/NackerHews)

The book is laid out by logical topics (on how to purposely write bad/underhanded code, followed by explanation on why/how they work) instead of keywords (such as IIFEs, automatic semicolon insertion, function calls etc)

disclaimer: I wrote it

EDIT: I skimmed through Speaking JavaScript. It's a bloody good book, much better than mine. You probably should read that

One more EDIT:

Since there was another thread which I gave a coupon for HNers, here's the link:

http://leanpub.com/underhandedjavascript/c/NackerHews

Agreed. Books that focus on teaching syntax and common use cases really fall flat for me. The best programming books teach you how to think in the syntax of the language. Eloquent Javascript [1] does a good job of that and is also free.

The best programming book at teaching a practical contemporary language in depth while teaching you to think with it is Metaprogramming Ruby. That's a great read for anyone considering writing a book to teach a programming language. And the best book I've ever read on thinking about programming comprehensively is SICP. Which again, I would recommend reading at least half of if you are going to write a book teaching a language.

1. http://eloquentjavascript.net/

Don’t judge a book by its TOC! I’d love to get your opinion once you have read or at least skimmed, say, chapter 1: http://speakingjs.com/es5/ch01.html

john resig- secrets of javascript ninja

absolutely the best

I hate books with contents like that too - I've read about those things many times already and unless the language I'm currently learning really does something interesting in one of those areas reading whole paragraphs of text on them is just a waste of time. I'm recently learning Forth, and I really do appreciate more detailed explanations, because holy cow, that's some really crazy shit going on :) However, for languages and technologies closer to mainstream - or just similar to what I already know - I use these web sites for an intro:

http://rosettacode.org/wiki/Rosetta_Code

http://learnxinyminutes.com/

http://hyperpolyglot.org/

And then go straight to the most advanced books I can find. It saves me a lot of time and it's more fun than following yet another book structured in exactly the same way, with pages and pages of text about minor syntactical differences and rare interesting things buried deep inside the book.

I humbly disagree. Especially in the case of JavaScript it's a good idea to review JavaScript's specific perspective on functions and object but also variable scope. I personally found "JavaScript Enlightment" quite interesting in this respect.

I agree. I enjoy weird quirky books that actually have a narrative and you build projects and learn the syntax while you are doing more interesting stuff.

However, I've read many books with tables of contents that look like that, yet they were actually pleasant to read and much more interesting than their contents implied.

Personally, I love learning from books like this one and consider tutorial-style books a complete waste of my time. It's a matter of learning style, and the differences are fairly well understood. Look up deductive vs inductive teaching.

Teachers try to use both methods, but book authors don't have to. Readers are free to learn from any book that suits them.

I second Eloquent Javascript. It's the book I recommend to JS beginners. Secrets of the Javascript Ninja by John Resig & Bear Bibealt is another good one once you get past the terrible title.

I will checkout Learning Javascript Design Patterns on your recommendation. I bookmarked that book a while ago but forgot to return and give it a read.

Sweet, looks like something fun to work with this weekend, thanks for sharing!

Also see Higher-Order Javascript by Sean M. Burke which is a "JavaScriptish companion" to Mark-Jason Dominus's Higher-Order Perl book - http://interglacial.com/hoj/

I wonder if it's a good idea to read SICP for the first time in a... let's say a language that people have mixed feelings about rather than the original version of the book. I guess nothing wrong with learning Scheme but as a JavaScripter this looks too good.

It’s a matter of deliberate focus! If you are interested in browser stuff, you can either combine Speaking JS with the content on MDN [1] or buy a different book. I like the books by Nicholas Zakas and “JavaScript: The Definitive Guide”.

[1] https://developer.mozilla.org/en-US/

There's no shortage of books about JavaScript in a browser context. There's fewer books on JavaScript that don't fairly exclusively focus on that context.

I would be surprised if it were. All languages I know says NaN == NaN is false.

That's the canonical way to check for NaN, in fact.

It makes sense. Just because both things are not something does not mean they are the same.

The reason is simple:

"hello" is Not a Number. "world" is Not a Number.

(Number)"hello" == (Number)"world" returns false.

EDIT: Another reason why this is the case is because IEEE mandates it to be. All numbers in JS are 64 bit floating point numbers. This allows for a crazy amount of NaNs (and oh the fun stuff you can do with that). To be sane, NaNs are not the same when comparing them (there is a lot more subtlety in that - comparing quiet and signalling NaNs, and so on and so forth)

if a !== a then a is NaN

This makes it easy to test for NaN! It's the only value for which this is true. (And the built-in isNaN() function has misleading behavior.)

It's actually part of some cryptic and ancient standard and is common for nearly all languages nowadays. I never cared about it very much, but here's the post on NaN semantics from someone who apparently designed the standard:

http://stackoverflow.com/questions/1565164/what-is-the-ratio...

This reminds me of a hilarious Destroy All Software screencast about oddities in certain programming languages: https://www.destroyallsoftware.com/talks/wat

It’s an experiment of mine (but I am aware of at least one more O’Reilly book that has been published this way). I’m hoping that it’s win-win: The public version lets people share and link to what they have read (click on a title to create a link to it). And it allows them to figure out whether the book is for them. But I believe that it will also help sell books.

I strongly considered the approach of releasing the contents of my recent book online with Mining the Social Web 2E [1] but instead decided to pursue what I felt was a standard OSS model: release a really high quality version of the source code that's optimized for easy learning in IPython Notebook format (optionally packaged as a turn-key VM) on GitHub [2] with the book being a form of "premium support" for the codebase if people want to learn more or dig deeper. I touch on all of this somewhat in the book's blog's "book as a startup posts" [3], and it seems to be working well so far.

I'm increasingly becoming interested in the prospect of releasing the entire contents of the book (both prose and source code) in IPython Notebook format so that you could read and work in the book seamlessly as "executable paper" like this full-text sampler of Chapter 1 [4] if it were hosted on Wakarii or a similar platform that offers a free tier. It really seems to me that this is the future of tech books: learning platforms with prose and example code integrated seamlessly.

[1] http://amzn.to/GPd59m

[2] https://github.com/ptwobrussell/Mining-the-Social-Web-2nd-Ed...

[3] http://miningthesocialweb.com/category/book-as-a-startup/

[4] http://bit.ly/IW3cbc

I saw some tweets yesterday from which I inferred there have been some DNS propagation issues. I notice the IP is an S3 one so I tried.. http://s3.amazonaws.com/speakingjs.com/es5/index.html and it seems to work, so might be a good temporary workaround.