Thu, 24 May 2001 12:17:24 -0700
I agree. I think we should define languages using standard locales. So an
English text would be "en", and a US-specific English text (e.g. the NIV)
would be "en_US". Language codes are defined by ISO 3166, and country codes
by ISO 639. Use the 2-letter language code (ISO 639-1) when available, or
the 3-letter terminology language code (ISO 639-2/T) when that is not
available. See http://www.din.de/gremien/nas/nabd/iso3166ma/ for language
codes, and http://lcweb.loc.gov/standards/iso639-2/englangn.html for country
codes. In the case where neither ISO 639-1 nor ISO 639-2/T includes a code
for the language, someone from SIL suggested marking the country as "xx" and
using their Ethnologue codes in the country position. So, for example, the
Ketchi Bible would have "lang=xx_KEK". See
http://www.sil.org/ethnologue/countries/ for the SIL Ethnologue.
And when/if OSIS comes up with another standard, we can probably switch over
to that rather painlessly as this seemed to be the direction that was
proposed for a future standard.
This system should handle any still living language and any ancient language
that we are likely to use.
And while we're at it, I propose we also add some sort of denomination= tag.
I'm willing to take suggestions on this issue since there aren't any ISO
standards on denominations. :) We may want to define this as a comma
delimited list so that a text could be categorized as "calvinist, academic,
protestant" for example. It would help show our reason for esteeming some
texts over others and allow users, eventually, to screen out texts by
denomination when using Sword or downloading new texts. I'm not sure if
denomination is the best tag name, since it shouldn't necessarily be limited
to denomination, but also academic vs. inspirational, dynamic vs. literal,