types-grammar, ch4: rearrangement for coherency

getify · getify · commit 2d3493436979 · 2022-08-22T11:36:39.000-05:00
diff --git a/types-grammar/ch4.md b/types-grammar/ch4.md
@@ -9,9 +9,39 @@ We've thouroughly covered all of the different *types* of values in JS. And alon
 
 In this chapter, we'll dive deep into coercion and uncover all its mysteries.
 
+## Coercion: Explicit vs Implicit
+
+Some developers assert that when you explicitly indicate a type change in an operation, this doesn't qualify as a *coercion* but just a type-cast or type-conversion. In other words, the claim is that coercion is only implicit.
+
+I disagree with this characterization. I use *coercion* to label any type conversion in a dynamically-typed language, whether it's plainly obvious in the code or not. Here's why: the line between *explicit* and *implicit* is not clear and objective, it's fairly subjective. If you think a type conversion is implicit (and thus *coercion*), but I think it's explicit (and thus not a *coercion*), the distinction becomes irrelevant.
+
+Keep that subjectivity in mind as we explore various *explicit* and *implicit* forms of coercion. In fact, here's a spoiler: most of the coercions could be argued as either, so we'll be looking at them with such balanced perspective.
+
+### Implicit: Bad or ...?
+
+An extremely common opinion among JS developers is that *coercion is bad*, specifically, that *implicit coercion is bad*; the rise in popularity of type-aware tooling like TypeScript speaks loudly to this sentiment.
+
+But that feeling is not new. 14+ years ago, Douglas Crockford's book "The Good Parts" also famously decried *implicit coercion* as one of the *bad parts*. Even Brendan Eich, creator of JS, regularly claims that *implicit coercion* was a mistake[^EichCoercion] in the early design of the language that he now regrets.
+
+If you've been around JS for more than a few months, you've almost certainly heard these opinions voiced strongly and predominantly. And if you've been around JS for years or more, you probably have your mind already made up.
+
+In fact, I think you'd be hard pressed to name hardly any other well-known source of JS teaching that strongly endorses coercion (in virtually all its forms); I do -- and this book definitely does! -- but I feel mostly like a lone voice shouting futilely in the wilderness.
+
+However, here's an observation I've made over the years: most of the folks who publicly condemn *implicit coercion*, actually use *implicit coercion* in their own code. Hmmmm...
+
+Douglas Crockford says to avoid the mistake of *implicit coercion*[^CrockfordCoercion], but his code uses `if (..)` statements with non-boolean values evaluated. [^CrockfordIfs] Many have dismissed my pointing that out in the past, with the claim that `ToBoolean()` isn't *really* coercion. Ummm... ok?
+
+Brendan Eich says he regrets *implicit coercion*, but yet he openly endorses[^BrendanToString] idioms like `x + ""` (and others!) to coerce the value in `x` to a string (we'll cover this later); and that's most definitely an *implicit coercion*.
+
+So what do we make of this dissonance? Is it merely a, "do as I say, not as I do" minor self-contradiction? Or is there more to it?
+
+I am not going to pass a final judgement here yet, but I want you the reader to deeply ponder that question, as you continue throughout this chapter and book.
+
 ## Abstracts
 
-The JS specification details a number of *abstract operations*[^AbstractOperations] that dictate internal conversion from one value-type to another. It's important to be aware of these operations, as coercive mechanics in the language mix and match them in various ways.
+Now that I've challenged you to examine coercion in more depth than you may have ever previously indulged, let's first look at the foundations of how coercion occurs, according to the JS specification.
+
+The specification details a number of *abstract operations*[^AbstractOperations] that dictate internal conversion from one value-type to another. It's important to be aware of these operations, as coercive mechanics in the language mix and match them in various ways.
 
 These operations *look* as if they're real functions that could be called, such as `ToString(..)` or `ToNumber(..)`. But by *abstract*, we mean they only exist conceptually by these names; they aren't functions we can *directly* invoke in our programs. Instead, we activate them implicitly/indirectly depending on the statements/expressions in our programs.
 
@@ -380,33 +410,9 @@ IsLessThan(NaN,1 /*LeftFirst=*/ true );          // false
 IsLessThan(41n,42n, /*LeftFirst=*/ true );       // true
 ```
 
-## Explicit vs. Implicit Coercions
-
-Some developers assert that when you explicitly indicate a type change in an operation, this doesn't qualify as a *coercion* but just a type-cast or type-conversion. In other words, the claim is that coercion is only implicit.
-
-I disagree with this characterization. I use *coercion* to label any type conversion in a dynamically-typed language, whether it's plainly obvious in the code or not. Here's why: the line between *explicit* and *implicit* is not clear and objective, it's fairly subjective. If you think a type conversion is implicit (and thus *coercion*), but I think it's explicit (and thus not a *coercion*), the distinction becomes irrelevant.
-
-Keep that subjectivity in mind as we explore various *explicit* and *implicit* forms of coercion. In fact, here's a spoiler: most of the coercions could be argued as either, so we'll be looking at them with such balanced perspective.
-
-#### Implicit: Bad or ...?
-
-An extremely common opinion among JS developers is that *coercion is bad*, specifically, that *implicit coercion is bad*; the rise in popularity of type-aware tooling like TypeScript speaks loudly to this sentiment.
-
-But that feeling is not new. 14+ years ago, Douglas Crockford's book "The Good Parts" also famously decried *implicit coercion* as one of the *bad parts*. Even Brendan Eich, creator of JS, regularly claims that *implicit coercion* was a mistake[^EichCoercion] in the early design of the language that he now regrets.
-
-If you've been around JS for more than a few months, you've almost certainly heard these opinions voiced strongly and predominantly. And if you've been around JS for years or more, you probably have your mind already made up.
-
-In fact, I think you'd be hard pressed to name hardly any other well-known source of JS teaching that strongly endorses coercion (in virtually all its forms); I do -- and this book definitely does! -- but I feel mostly like a lone voice shouting futilely in the wilderness.
+## Concrete Coercions
 
-However, here's an observation I've made over the years: most of the folks who publicly condemn *implicit coercion*, actually use *implicit coercion* in their own code. Hmmmm...
-
-Douglas Crockford says to avoid the mistake of *implicit coercion*[^CrockfordCoercion], but his code uses `if (..)` statements with non-boolean values evaluated. [^CrockfordIfs] Many have dismissed my pointing that out in the past, with the claim that `ToBoolean()` isn't *really* coercion. Ummm... ok?
-
-Brendan Eich says he regrets *implicit coercion*, but yet he openly endorses[^BrendanToString] idioms like `x + ""` (and others!) to coerce the value in `x` to a string (we'll cover this later); and that's most definitely an *implicit coercion*.
-
-So what do we make of this dissonance? Is it merely a, "do as I say, not as I do" minor self-contradiction? Or is there more to it?
-
-I am not going to pass a final judgement here yet, but I want you the reader to deeply ponder that question, as you continue throughout this chapter and book.
+Now that we've covered all the abstract operations JS defines for handling various coercions, it's time to turn our attention to the concrete statements/expressions we can use in our programs that activate these operations.
 
 ### To Boolean
 
@@ -468,6 +474,12 @@ In fact, as mentioned earlier in the `ToBoolean()` discussion, some developers d
 
 My take: `Boolean(..)` is the most preferable *explicit* coercion form. Further, I think `if` / `for` / `while` statements are *implicitly* coercing non-`boolean`s, but I'm OK with that.
 
+Since most developers, including famous names like Doug Crockford, also in practice use implicit (`boolean`) coercions in their `if`[^CrockfordIfs] and loop statements, I think we can say that at least *some forms* of *implicit* coercion are widely acceptable, regardless of the rhetoric to the contrary.
+
+### To Primitive
+
+// TODO
+
 ### To String
 
 As with `ToBoolean()`, there are a number of ways to activate the `ToString()` coercion (as discussed earlier in the chapter). The decision of which approach is similarly subjective.
@@ -555,12 +567,24 @@ Symbol("ok") + "";              // TypeError exception thrown
 
 Why the exception? JS treats `+ ""` as an *implicit* coercion, which is why when activated with a symbol, an exception is thrown. I think that's a pretty ironclad proof.
 
-Nevertheless, as I mentioned earlier, Brendan Eich endorses `+ ""` as the *best* way to coerce values to strings. I think that carries a lot of weight, in terms of him supporting at least a subset of *implicit* coercion practices. His views on *implicit* coercion must be a bit more nuanced than, "it's all bad."
+Nevertheless, as I mentioned at the start of this chapter, Brendan Eich endorses `+ ""`[^BrendanToString] as the *best* way to coerce values to strings. I think that carries a lot of weight, in terms of him supporting at least a subset of *implicit* coercion practices. His views on *implicit* coercion must be a bit more nuanced than, "it's all bad."
 
 ### To Number
 
 // TODO
 
+### Nullish
+
+// TODO
+
+[^EichCoercion]: "The State of JavaScript - Brendan Eich", comment thread, Hacker News; Oct 9 2012; https://news.ycombinator.com/item?id=4632704 ; Accessed August 2022
+
+[^CrockfordCoercion]: "JavaScript: The World's Most Misunderstood Programming Language"; 2001; https://www.crockford.com/javascript/javascript.html ; Accessed August 2022
+
+[^CrockfordIfs]: "json2.js", Github; Apr 21 2018; https://github.com/douglascrockford/JSON-js/blob/8e8b0407e475e35942f7e9461dab81929fcc7321/json2.js#L336 ; Accessed August 2022
+
+[^BrendanToString]: ESDiscuss mailing list; Aug 26 2014; https://esdiscuss.org/topic/string-symbol#content-15 ; Accessed August 2022
+
 [^AbstractOperations]: "7.1 Type Conversion", ECMAScript 2022 Language Specification; https://262.ecma-international.org/13.0/#sec-type-conversion ; Accessed August 2022
 
 [^ToBoolean]: "7.1.2 ToBoolean(argument)", ECMAScript 2022 Language Specification; https://262.ecma-international.org/13.0/#sec-toboolean ; Accessed August 2022
@@ -597,12 +621,4 @@ Nevertheless, as I mentioned earlier, Brendan Eich endorses `+ ""` as the *best*
 
 [^StringPrefix]: "7.2.9 IsStringPrefix(p,q)", ECMAScript 2022 Language Specification; https://262.ecma-international.org/13.0/#sec-isstringprefix ; Accessed August 2022
 
-[^EichCoercion]: "The State of JavaScript - Brendan Eich", comment thread, Hacker News; Oct 9 2012; https://news.ycombinator.com/item?id=4632704 ; Accessed August 2022
-
-[^CrockfordCoercion]: "JavaScript: The World's Most Misunderstood Programming Language"; 2001; https://www.crockford.com/javascript/javascript.html ; Accessed August 2022
-
-[^CrockfordIfs]: "json2.js", Github; Apr 21 2018; https://github.com/douglascrockford/JSON-js/blob/8e8b0407e475e35942f7e9461dab81929fcc7321/json2.js#L336 ; Accessed August 2022
-
-[^BrendanToString]: ESDiscuss mailing list; Aug 26 2014; https://esdiscuss.org/topic/string-symbol#content-15 ; Accessed August 2022
-
 [^SymbolString]: "String(symbol)", ESDiscuss mailing list; Aug 12 2014; https://esdiscuss.org/topic/string-symbol ; Accessed August 2022
diff --git a/types-grammar/toc.md b/types-grammar/toc.md
@@ -28,6 +28,7 @@
     * Fundamental Objects
     * TODO
 * Chapter 4: Coercing Values
+    * Coercion: Explicit vs Implicit
     * Abstracts
-    * Explicit vs. Implicit Coercions
+    * Concrete Coercions
     * TODO
