You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I see the specification of the language property is expressed in reference to ISO 639-3. ISO 639-3 is now replaced (like other parts of the former 639 multipart standard) by a single ISO 639:2023 Code for individual languages and language groups. In this standard you still have sets (in this case set 3) reminiscence of the former parts, but it is more adequate to follow the recommendation of IETF BCP 47 which articulates the ISO 639 language code with other standards (scripts, locale, etc.) thus allowing a finer-grained representation of language varieties. This is what is adopted for instance for the generic xml:lang attribute in the XML recommendation and would ensure a better interoperability across applications. You can find a good introduction under: https://www.w3.org/International/articles/language-tags/
The text was updated successfully, but these errors were encountered:
Thanks! This is simply about the note Language of the ... expressed using ISO 639-3 placed in the README.md, right?
If we state there, it is according to ISO 639:2023 Set 3 then it is not a breaking change; however, if we would just state ISO 639:2023 then it is a breaking change?
The point is whether you want to stick to the 3 letter codes (which was meant by the reference to 639-3 I gather) or adopt the larger framework of IETF BCP 47 which recommends 2 letter codes when they exist (e.g. fr, en) and allows one to use more elaborate identifiers (e.g. fr-CA for Canadian French). I think it would provide a better interoperability framework with other applications using language information.
Ahh I see... that would be better, but we introduce a breaking change in DCS then (so potentially for v2.0)? Or does BCP 47 also allow 3 letter codes (without additional)?
It does not, but hardly anyone is using the 3 letter codes alone. So, yes, you need to introduce a deprecation for the old format and organise the switch .
I see the specification of the language property is expressed in reference to ISO 639-3. ISO 639-3 is now replaced (like other parts of the former 639 multipart standard) by a single ISO 639:2023 Code for individual languages and language groups. In this standard you still have sets (in this case set 3) reminiscence of the former parts, but it is more adequate to follow the recommendation of IETF BCP 47 which articulates the ISO 639 language code with other standards (scripts, locale, etc.) thus allowing a finer-grained representation of language varieties. This is what is adopted for instance for the generic xml:lang attribute in the XML recommendation and would ensure a better interoperability across applications. You can find a good introduction under: https://www.w3.org/International/articles/language-tags/
The text was updated successfully, but these errors were encountered: