belter creole, language codes “qbc” and “art-x-belter” (for now)
![Belter Creole [lang belta], language codes “qbc” (iso 639-3) and “art-x-belter” (ietf)](/posts/attachments/ef8937e6-24d9-4b49-8950-afe343a2ffab/image.png)
belter creole has two language codes, “qbc” and “art-x-belter”. but how?
“qbc” is from the “reserved for local use” area (“qaa” through “qtz”) of the three-letter iso 639-3 language codes, like “eng” for english, “cmn” for mandarin chinese, or “ckb” for sorani. conlang code registry is a project to informally coordinate “assignments” in this area for conlangs, and that’s why belter creole is “qbc”. who knows, maybe someday it will transcend the expanse (2015) and upgrade to a standard code, like toki pona did from “qtk” to “tok”.
this reminds me of cambridge g-rin, a project to informally coordinate “assignments” in the rfc 1918 address spaces, that is, private ipv4 addresses like 10.0.0.1 and 192.168.0.1. in both cases, a public registry that allows anyone who knows about it to avoid taking someone else’s spot.
“art-x-belter” starts with “art”, the iso 639-3 code for “artificial languages”. the “-x-belter” part uses an ietf extension to those codes (bcp 47) that lets you add “private-use” extra details, and that’s why belter creole is “art-x-belter”. internet standards love “x-” for private use. see also http headers like “X-Real-IP” or “X-Forwarded-For”, or mime headers like “X-Spam-Score”.
by the way, another language code with private-use extra details is in the html spec, of all places. thanks hixie :)
<!DOCTYPE html>
<html class="split index" lang=en-US-x-hixie>
<script src=/link-fixup.js defer=""></script>
<meta charset=utf-8>
[...]