Normative: Add new numbering system "tnsa"

[anba]

Add "tnsa" to the table of numbering systems with simple digit mappings. The
Tangsa script ("Tnsa") is a new addition in Unicode 14.

Also see: unicode-org/cldr#1326

[Constellation]

I have one comment. This "tnsa" is just added at 08/18 and not shipped yet in ICU. Requiring this explicitly in the spec means that only engines shipping trunk ICU can fully implement ECMA 402. Isn't it too restrictive as a spec?

[srl295]

ICU4C 70RC, for example, is available <https://github.com/unicode-org/icu/releases/tag/release-70-rc> last week including the numbering system. Tangsa is already a part of Unicode 14 which shipped last month. As with other locale data, I think it is typical that an implementation may only implement a subset of the options. Specifically §4.4 “Subsets of Unicode”. So if your implementation does not implement Unicode 14 then my reading is that there might be a difference of behavior if Tnsa characters were requested. https://tc39.es/ecma402/#sec-implementation-dependencies

…

El oct. 7, 2021, a las 12:30 p. m., Yusuke Suzuki ***@***.***> escribió: I have one comment. This "tnsa" is just added at 08/18 and not shipped yet in ICU. Requiring this explicitly in the spec means that only engines shipping trunk ICU can fully implement ECMA 402. Isn't it too restrictive as a spec?

[anba]

ECMA-262 also requires to always use the latest Unicode standard (emphasis mine):

A conforming implementation of ECMAScript must interpret source text input in conformance with the latest version of the Unicode Standard and ISO/IEC 10646.

The PR for updating related properties to Unicode 14 is tc39/ecma262#2515.

And ECMA-402 requires to use the latest UTS 35 version (emphasis mine):

The following referenced documents are required for the application of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.

[...]

• Unicode Technical Standard 35
  • Unicode Locale Data Markup Language
  • Unicode BCP 47 Locale Identifiers

ECMA-402 hasn't yet fully formalised the update process for UTS 35 changes → #584.

---

Names and Keywords has this text (emphasis mine):

The Unicode code points in the specified categories in the latest version of the Unicode standard must be treated as in those categories by all conforming ECMAScript implementations. ECMAScript implementations may recognize identifier code points defined in later editions of the Unicode Standard.

Is that really "later than the latest edition" or am I misreading this? So, it's okay to use future, unreleased Unicode versions?! 😄

(This paragraph was referring to specific Unicode versions in older ECMAScript versions, cf. tc39/ecma262#620.)

[srl295]

Probably "later than the one supported by this edition of 402"? But probably need a separate issue for this.

[FrankYFTang]

I have one comment. This "tnsa" is just added at 08/18 and not shipped yet in ICU. Requiring this explicitly in the spec means that only engines shipping trunk ICU can fully implement ECMA 402. Isn't it too restrictive as a spec?

ICU 70.1 which scheduled to release in Oct 27, 2021 will have such support as well as ICU 71.1 which scheduled to be ship in April 2022.
IF we merge this PR in Oct 8 2021, the first ever edition of ECMA402 will be "shipped" would be in Sumer 2022, which then will be 8+ months after the first version ICU released and ~10 months after Unicode 14 published, isn't that giving the engine implementation enough time to act?

[FrankYFTang]

I am supporting this PR but prefer we hold it unmerged until Nov 2021- after ICU 70.1 public released. The deadline for the 2022 version of ECMA402 is after Feb 2022 anyway.

[srl295]

I am supporting this PR but prefer we hold it unmerged until Nov 2021- after ICU 70.1 public released. The deadline for the 2022 version of ECMA402 is after Feb 2022 anyway.

Should you set the "ES 2022" milestone on this then?

I think that's probably what the "later than latest" wordage means above, @anba - say ES2022 has U14, then an ES2022 compliant imp could support U15 (which will be out in 2022 :)

[srl295]

maybe we even want a branch for es2022. dunno

[sffc]

@ryzokuken Can you follow up on the feedback from last month's meeting?

[srl295]

ICU 70.1 has been released

[ryzokuken]

@litherum said they'd post a comment on the thread after talking to their release engineering team. Any updates Myles?

[litherum]

Yes I'll comment later today.

[ryzokuken]

@litherum are you still on this?

[ryzokuken]

@sffc can we put this on the agenda for next week? I'd love to hear the JSC team's thoughts on this before we move ahead.

[sffc]

2022-01-13 approval in TC39-TG2: https://github.com/tc39/ecma402/blob/master/meetings/notes-2022-01-13.md#normative-add-new-numbering-system-tnsa

[litherum]

@litherum said they'd post a comment on the thread after talking to their release engineering team. Any updates Myles?

We're comfortable adding tnsa to the spec.

[ryzokuken]

I'll present it to plenary during this meeting, after which it can be merged.

[ryzokuken]

2022-01-24: Approved by TC39-TG1.