ISO 639 "cnr" should be mapped to 'SRB '

MicrosoftDocs / typography-issues

Creative Commons Attribution 4.0 International

47 stars 21 forks source link

ISO 639 "cnr" should be mapped to 'SRB ' #590

Closed dscorbett closed 4 years ago

dscorbett commented 4 years ago

Montenegrin (ISO 639 code "cnr") is a variety of Serbian. Since it doesn’t have its own language system tag, "cnr" should be mapped to 'SRB ' (Serbian).

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

ID: 662c6f54-f90c-52ed-dca2-0f0a0a49fee5
Version Independent ID: 6aa33924-d45e-60c3-a9fd-e317ff662993
Content: Language system tags - Typography
Content Source: typographydocs/opentype/spec/languagetags.md
Product: typography
GitHub Login: @PeterCon
Microsoft Alias: PeterCon

PeterCon commented 4 years ago

Hmmm... People that insisted that ISO 639 needed to code "Montenegrin" as distinct from "Serbian" might disagree. Should 'SRB ' be mapped to cnr, or should a separate language system tag be created?

dscorbett commented 4 years ago

I don’t think it’s necessary, unless there is some evidence that the two varieties have different glyph preferences, though it wouldn’t be wrong to create a new tag. A benefit of mapping it to 'SRB ' is that it is easier to update shaping engines to map "cnr" to 'SRB ' than it is to update all fonts that use 'SRB ' to also use 'CNR '. I leave it to you; my main concern is that it be mapped to something, because otherwise text tagged as Montenegrin will end up using the default (probably Russian) glyphs.

PeterCon commented 4 years ago

I don't have any strong opinion on this issue beyond that reasonable/relevant alternatives have been considered. Your rationale makes sense to me.

Changing for next version.