Re: [Ltru] Consensus call: extlang

["Phillips, Addison" <[e-mail]>]

>
> It is also permissible, within one's own application,
> to tag UK English as 'en' and require US English to be
> tagged 'en-US'.  But "permissible" doesn't mean "advisable".
> I would argue that if an application knows the difference
> and cares about the difference, that it would be much better
> to use both 'en-GB' and 'en-US', and reserve naked 'en' for
> the cases where there's no value in making the distinction
> (e.g., it can't tell or doesn't care).
>

I don't disagree. However...

The classic software fallback problem (sorry Karen) is that when someone requests "en-AQ" and you don't have any Antarctic English resources, one falls back on some content, which is typically labeled (in software resources) 'en'. This 'en' content can be one variety or another, but not both. In English this isn't always a problem. The foregoing sentence, for example, is good as both "en-US" and "en-GB".

Things are less good when there are disparities and no generic choice exists. Some years ago I work on a project that including localization of a product into Simplified Chinese--but only Simplified, not Traditional. That product also sold in Traditional Chinese speaking markets. Installing 'zh' labeled resources mean that ALL Chinese-locale configured systems might have displayed Simplified Chinese, so I had to implement certain checks to ensure that Traditional Chinese markets didn't see the Simplified (they saw the default resources instead).

Yes, the resources could have be labeled more specifically than 'zh'. But then I would run the risk of having a user with an unfamiliar locale configuration not seeing the Chinese resources they paid for. So the "permissible" and "advisable" bar may be set at different heights under different circumstances.

Addison
_______________________________________________
Ltru mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/ltru