Scripsit Andreas Prilop:
I'm not aware that any software - let alone search engines -
recognize, say "en-GB" and "en-US" and make any difference.
I think IBM Home Page Reader recognizes those codes and selected the
pronunciation accordingly. However the software has been discontinued. Yet,
it's still _potentially_ useful to declare the language variant.
If you open an HTML file in (a sufficiently new version of) MS Word, the
lang attributes will be recognized and applied in spelling checks. I guess
Word would then accept "honour" in en-GB and reject it in en-US.
You could also write (as an author or as a user) a style sheet that uses
language selectors and e.g. highlights any text in en-GB. Maybe someone
finds some use for such methods.
To summarize, the tangible benefits of using language markup in general and
refined markup in particular are rather small, but the idea is not
completely unrealistic. Using very refined codes means taking the risk that
software capable of recognizing simple codes like "en" doesn't grok your
"en-foobar" at all.
If someone really wants to spend time with such issues, RFC 4646 is probably
the best starting point.
--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/