Editorial: Replace Number-specific term "finite" with "finite Number" by gibson042 · Pull Request #3911 · tc39/ecma262

gibson042 · 2026-07-13T22:59:48Z

...fixing inappropriate auto-linking for uses of "finite" as a colloquial term

This does not affect Web IDL, HTML, or WebAssembly. ECMA-402 has a corresponding PR at tc39/ecma402#1091 , which can land independently of this one.

...fixing inappropriate auto-linking for uses of "finite" as a colloquial term Fixes tc39#3840

Ref tc39/ecma262#3911

github-actions · 2026-07-13T23:07:08Z

The rendered spec preview for this PR is available as a single page at https://tc39.es/ecma262/pr/3911 and as multiple pages at https://tc39.es/ecma262/pr/3911/multipage .

michaelficarra · 2026-07-14T00:25:07Z

          1. If _value_ is a Number, then
-            1. If _value_ is finite, return ! ToString(_value_).
+            1. If _value_ is a finite Number, return ! ToString(_value_).


Alternatively:

Suggested change

1. If _value_ is a Number, then

1. If _value_ is finite, return ! ToString(_value_).

1. If _value_ is a finite Number, return ! ToString(_value_).

1. If _value_ is a finite Number, return ! ToString(_value_).

1. If _value_ is a Number, return *"null"*.

But either one is fine.

michaelficarra

Changes LGTM otherwise, but I haven't checked that all the remaining uses of the term "finite" are still appropriate yet.

…Number"

nicolo-ribaudo · 2026-07-16T15:41:22Z

@@ -33068,7 +33068,7 @@ <h1>Number.prototype.toFixed ( _fractionDigits_ )</h1>
          1. Assert: If _fractionDigits_ is *undefined*, then _fractionCount_ is 0.
          1. If _fractionCount_ is not finite, throw a *RangeError* exception.


This is the result of ToIntegerOrInfinity, it's be clearer to say "is not an integer". Something similar happens in Function.prototype.bind for targetLengthAsInt.

Updated both sites.

michaelficarra · 2026-07-16T15:42:20Z

I see some additional uses of finite time value. I think we're going to need to bite the bullet and just dfn it.

michaelficarra · 2026-07-16T15:44:28Z

In MakeDate, there's a missed finite -> finite Number.

michaelficarra · 2026-07-16T15:45:22Z

All other uses of finite LGTM.

…Number"

Ref tc39#3911 (comment)

…ntial,toFixed,toPrecision}

Ref tc39#3911 (comment)

…ting AO parameter types)

Ref tc39#3911 (comment)

…ainder with other AOs

gibson042 · 2026-07-16T18:50:49Z

I see some additional uses of finite time value. I think we're going to need to bite the bullet and just dfn it.

Agreed 😭 . Done.

In MakeDate, there's a missed finite -> finite Number.

Fixed.

gibson042 · 2026-07-16T18:50:59Z

+          1. If _number_ is not a finite Number, return Number::toString(_number_, 10).
+          1. If _fractionCount_ is not an integer, throw a *RangeError* exception.


These steps are being aligned across Number.protototype.{toExponential,toFixed,toPrecision}. The integer check is optional because relational operators support extended mathematical values, but these are cases where we tend to be explicit for minimizing cognitive burden. Further refactoring could combine both checks into a single step and/or use inclusive interval in the range check, and I'm deferring both of those options to the future as out of scope for a PR whose focus is introducing the term "finite Number".

gibson042 · 2026-07-16T18:53:44Z

            1. If _numerator_ is either *+0*<sub>𝔽</sub> or *-0*<sub>𝔽</sub>, return _numerator_.
-            1. Assert: _numerator_ and _denominator_ are finite and non-zero.
+            1. Assert: _numerator_ is a finite Number and is neither *+0*<sub>𝔽</sub> nor *-0*<sub>𝔽</sub>.
+            1. Assert: _denominator_ is a finite Number and is neither *+0*<sub>𝔽</sub> nor *-0*<sub>𝔽</sub>.


"is a finite Number and is neither +0_𝔽 nor -0_𝔽" is a very common check in Number-related algorithms; we should consider a shorter way to express it.

I'd be okay defining nonzero finite Number.

In this PR, or later?

Either would be fine.

I guess it would be non-zero finite Number; we always include the hyphen.

michaelficarra · 2026-07-16T19:24:45Z

        1. If _x_ is an Object and _y_ is either a String, a Number, a BigInt, or a Symbol, return ! IsLooselyEqual(? ToPrimitive(_x_), _y_).
        1. If _x_ is a BigInt and _y_ is a Number, or if _x_ is a Number and _y_ is a BigInt, then
-          1. If _x_ is not finite or _y_ is not finite, return *false*.
+          1. If _x_ is a Number and _x_ is not a finite Number, return *false*.


So this is just NaN, +∞, and -∞, right? Would it be better to just say that?

michaelficarra · 2026-07-16T19:26:30Z

We should give esmeta a heads up about this and see if they could add support for the new term before landing.

gibson042 · 2026-07-16T19:41:36Z

We should give esmeta a heads up about this and see if they could add support for the new term before landing.

Indeed: es-meta/esmeta#346

gibson042 · 2026-07-16T19:49:30Z

        1. If _x_ is a BigInt and _y_ is a Number, or if _x_ is a Number and _y_ is a BigInt, then
-          1. If _x_ is not finite or _y_ is not finite, return *false*.
+          1. If _x_ is a Number and _x_ is not a finite Number, return *false*.
+          1. If _y_ is a Number and _y_ is not a finite Number, return *false*.
          1. If ℝ(_x_) = ℝ(_y_), return *true*.
          1. Return *false*.
        1. Return *false*.


So this is just NaN, +∞, and -∞, right? Would it be better to just say that?

I'd rather not enumerate three exceptions. If we don't like this phrasing, my preference would instead be

Suggested change

1. If _x_ is a BigInt and _y_ is a Number, or if _x_ is a Number and _y_ is a BigInt, then

1. If _x_ is not finite or _y_ is not finite, return *false*.

1. If _x_ is a Number and _x_ is not a finite Number, return *false*.

1. If _y_ is a Number and _y_ is not a finite Number, return *false*.

1. If ℝ(_x_) = ℝ(_y_), return *true*.

1. Return *false*.

1. Return *false*.

1. If _x_ is a BigInt and _y_ is a finite Number, or if _x_ is a finite Number and _y_ is a BigInt, then

1. If ℝ(_x_) = ℝ(_y_), return *true*.

1. Return *false*.

...taking advantage of type mismatch after the first step Ref tc39#3911 (comment)

…Number"

Editorial: Replace Number-specific term "finite" with "finite Number"

f9cecf1

...fixing inappropriate auto-linking for uses of "finite" as a colloquial term Fixes tc39#3840

gibson042 added editorial change editor call to be discussed in the next editor call labels Jul 13, 2026

gibson042 added a commit to gibson042/ecma402 that referenced this pull request Jul 13, 2026

Editorial: Replace Number-specific term "finite" with "finite Number"

08de55e

Ref tc39/ecma262#3911

gibson042 mentioned this pull request Jul 13, 2026

Editorial: Replace Number-specific term "finite" with "finite Number" tc39/ecma402#1091

Open

gibson042 added the request preview ask the bot to trigger a PR preview label Jul 13, 2026

github-actions Bot removed the request preview ask the bot to trigger a PR preview label Jul 13, 2026

michaelficarra reviewed Jul 13, 2026

View reviewed changes

Comment thread spec.html Outdated

michaelficarra reviewed Jul 14, 2026

View reviewed changes

Comment thread spec.html Outdated

michaelficarra reviewed Jul 14, 2026

View reviewed changes

Comment thread spec.html Outdated

michaelficarra reviewed Jul 14, 2026

View reviewed changes

fixup! Editorial: Replace Number-specific term "finite" with "finite …

669db13

…Number"

nicolo-ribaudo reviewed Jul 16, 2026

View reviewed changes

fixup! Editorial: Replace Number-specific term "finite" with "finite …

07de1be

…Number"

gibson042 added a commit to gibson042/ecma262 that referenced this pull request Jul 16, 2026

Editorial: Refine an "is finite" assertion to "is an integer"

bc07c90

Ref tc39#3911 (comment)

gibson042 added a commit to gibson042/ecma262 that referenced this pull request Jul 16, 2026

Editorial: Refine an "is finite" assertion to "is an integer"

eda5ab8

Ref tc39#3911 (comment)

gibson042 force-pushed the gh-3840-finite-number branch from bc07c90 to eda5ab8 Compare July 16, 2026 18:10

Editorial: Align finite-argument checks in Number.prototype.{toExpone…

4bd33ed

…ntial,toFixed,toPrecision}

gibson042 added a commit to gibson042/ecma262 that referenced this pull request Jul 16, 2026

Editorial: Refine an "is finite" assertion to "is an integer"

00bfbea

Ref tc39#3911 (comment)

gibson042 force-pushed the gh-3840-finite-number branch from eda5ab8 to 00bfbea Compare July 16, 2026 18:11

gibson042 added 2 commits July 16, 2026 14:26

Editorial: Define "finite time value" (as an autolink target for exis…

1d6601d

…ting AO parameter types)

Editorial: Refine an "is finite" assertion to "is an integer"

5dbf2af

Ref tc39#3911 (comment)

gibson042 force-pushed the gh-3840-finite-number branch from 00bfbea to 5dbf2af Compare July 16, 2026 18:26

Editorial: Align the non-zero finite Number assertions in Number::rem…

a749ea1

…ainder with other AOs

gibson042 commented Jul 16, 2026

View reviewed changes

michaelficarra reviewed Jul 16, 2026

View reviewed changes

gibson042 mentioned this pull request Jul 16, 2026

ESMeta should support new terms being added by ecma262 PR 3911 es-meta/esmeta#346

Open

gibson042 commented Jul 16, 2026

View reviewed changes

This was referenced Jul 16, 2026

Editorial: Introduce term "non-zero finite Number" #3917

Open

Editorial: Refactor IsLooselyEqual for better symmetry #3918

Open

gibson042 added a commit to gibson042/ecma262 that referenced this pull request Jul 16, 2026

Editorial: Refactor IsLooselyEqual for better symmetry

7b3b843

...taking advantage of type mismatch after the first step Ref tc39#3911 (comment)

gibson042 added a commit to gibson042/ecma262 that referenced this pull request Jul 16, 2026

Editorial: Refactor IsLooselyEqual for better symmetry

995df79

...taking advantage of type mismatch after the first step Ref tc39#3911 (comment)

nicolo-ribaudo approved these changes Jul 17, 2026

View reviewed changes

nicolo-ribaudo reviewed Jul 20, 2026

View reviewed changes

Comment thread spec.html

fixup! Editorial: Replace Number-specific term "finite" with "finite …

7993e1b

…Number"

		@@ -33068,7 +33068,7 @@ <h1>Number.prototype.toFixed ( _fractionDigits_ )</h1>
		1. Assert: If _fractionDigits_ is undefined, then _fractionCount_ is 0.
		1. If _fractionCount_ is not finite, throw a RangeError exception.

		1. If _number_ is not a finite Number, return Number::toString(_number_, 10).
		1. If _fractionCount_ is not an integer, throw a RangeError exception.

Uh oh!

Conversation

gibson042 commented Jul 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Jul 13, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

michaelficarra left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

michaelficarra commented Jul 16, 2026

Uh oh!

michaelficarra commented Jul 16, 2026

Uh oh!

michaelficarra commented Jul 16, 2026

Uh oh!

gibson042 commented Jul 16, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

michaelficarra commented Jul 16, 2026

Uh oh!

gibson042 commented Jul 16, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

gibson042 commented Jul 13, 2026 •

edited

Loading