![](https://secure.gravatar.com/avatar/a0f6b595b0e3b7f3abba19eeb8261fc2.jpg?s=120&d=mm&r=g)
21 Dec
2010
21 Dec
'10
5:17 a.m.
But context uses some non-standard codes also: “deo” instead of “de-1901” for sane German orthography, and “agr” for ancient Attic instead of “grc” or “el-polyton”. (Testfile appended.)
Yes, that's what I said: ConTeXt has used non-standard codes in the past, now we try to follow BCP 47, while attempting to stay backward-compatible, and providing easy-to-remember names for the user as well. But of course, it's always easier to write a three-line sentence that contains one troll and one severe inaccuracy.
Note that Unicode Language Identifiers are not the same as BCP 47 tags, even if they're very close in their goals and shape. Arthur