Almost all applications (which use gettext, at least) are written with their C (untranslated) locale in en_US. This means that they never have an explicit en_US translation, but are always 100% translated to en_US.
Should the <languages> element list en_US explicitly, in which case it should probably be injected by the appstream XML generator?
Or should the spec say that the application’s C locale is never listed as a <lang> element, and should be assumed to be en_US?
If we go with the second option, perhaps the assumption about the C locale being en_US could be overrideable (in the few cases where an application is written in something other than US English by default) by setting an attribute like <languages default-language="es_ES"> and defining that language as 100% translated.
I can put together a merge request to update the specification either way, but need to know what approach would be preferable 🙂
The text was updated successfully, but these errors were encountered:
Almost all applications (which use gettext, at least) are written with their C (untranslated) locale in
en_US. This means that they never have an expliciten_UStranslation, but are always 100% translated toen_US.Should the
<languages>element listen_USexplicitly, in which case it should probably be injected by the appstream XML generator?Or should the spec say that the application’s C locale is never listed as a
<lang>element, and should be assumed to been_US?If we go with the second option, perhaps the assumption about the C locale being
en_UScould be overrideable (in the few cases where an application is written in something other than US English by default) by setting an attribute like<languages default-language="es_ES">and defining that language as 100% translated.I can put together a merge request to update the specification either way, but need to know what approach would be preferable🙂
The text was updated successfully, but these errors were encountered: