Errata to the first printing by MIT Press

[martin-henz]

This issue collects the errata of the print edition of SICP JS (print 2022):

• 4.1.5, page 351 (top): Extra semicolon in function strange:

function strange(f) {
    return halts(f, f)
           ? run_forever();
           : "halted";
}

should be

function strange(f) {
    return halts(f, f)
           ? run_forever()
           : "halted";
}

fixed in online edition; fix is PR #747.

[martin-henz]

• 3.2.4, page 220, exercise 3.11, function dispatch:

               : "Unknown request: make_account";

should be

               : error(m, "Unknown request: make_account");

to make more sense and to be in line with SICP.

fixed in online edition; fix is PR #798.

[martin-henz]

• 2.2.1, page 90, typo in second line from top:

one at at time

should be

one at a time

Fixed in online edition; fix is PR #778.

[martin-henz]

• 2.1.3, page 80, line 5:

(an efficient implementation of pairs might use JavaScript’s primitive vector data structure)

should be

(an efficient implementation of pairs might use JavaScript’s native vector data structure)

The issue is discussed in #756.

Fixed in 489637f

[martin-henz]

2.2.3, page 105, last function of Exercise 2.37:

function matrix_times_matrix(n, m) {

should be

function matrix_times_matrix(n, m) {

The issue is discussed in #831.

Fixed in #1027.

[martin-henz]

3.5.1, page 322, middle of the page:

pair(head(stream), 
     stream_filter(pred, stream_tail(stream)));

should be

pair(head(stream), 
     () => stream_filter(pred, stream_tail(stream)));

as in SICP (using cons-stream).

The issue is discussed in #839.

Fixed in #1028.

[martin-henz]

1.1.6, page 15, the first two bullet points in the middle of the page:

e1 && e2 is syntactic sugar for e1 ? e2 : false, and e1 || e2 is syntactic sugar for e1 ? true : e2. Strictly speaking, in JavaScript, e1 && e2 is syntactic sugar for e1 ? e2 : e1 and e1 || e2 is syntactic sugar for e1 ? e1 : e2.

Direct link to the relevant portion of the language specification: https://tc39.es/ecma262/multipage/ecmascript-language-expressions.html#sec-binary-logical-operators-runtime-semantics-evaluation

Solution: Weaken the phrasing to:

"We assume16 this syntactic form to be syntactic sugar for
expression1​ ? true : expression2."
and add a clarifying footnote
"16 This assumption is justified by the restriction mentioned in footnote 14. Full JavaScript needs to consider the case where the result of evaluating expression11​ is neither true nor false."

The issue is discussed in #776.

Fixed in #1033.

[martin-henz]

References, pages 566, 568, 570, page headers:

The headers say "Chapter 5 References", but these are the references for the whole book.

The issue is discussed in #889.

Fixed in #1039.

[martin-henz]

Several places in 2.4.2, 2.4.3, and 2.5.2: math_atan was used instead of math_atan2.

The issue is discussed in #1006.

Fixed in #1041.

[martin-henz]

3.3.3, page 238, function insert.
Wrong:

const record = assoc(key_2, tail(table));

Right:

const record = assoc(key_2, tail(subtable));

The issue is discussed in #1000.

Fixed in #1043.

[martin-henz]

1.2.6, page 46. Inconsistent exercise.

The question makes use of a function report_prime that returns undefined. As a result, the given function start_prime_test either returns undefined or true. This makes it awkward to use the given setup in a solution, because undefined hasn't been properly introduced at this point. Instead of undefined, report_prime should return true. Then start_prime_test should be changed so that it returns true for primes and false for non-primes. That will allow for a solution that doesn't need to deal with undefined.

The issue is discussed in #973.

Fixed in #1045.

[martin-henz]

3.5.2, page 294, Exercise 3.60, second line, currently says

is implemented by add-streams

but should say

is implemented by add_streams

The issue is raised in #1048.

Fixed in #1049.