frr, fy, ngo, tt

[Frank Ellermann]

Michael Everson wrote:

> There are *thousands* of languages that use the Latin script.

If Web sites wish to tag their content the Suppress-Script is
relevant for backwards compatibility, there's no difference
between say Nkoo for ngo, (missing) Latn for fy, or (missing)
Latn for frr.

> I do not see why we should add a tag for Northern Frisian.

Because I found a kind of proof that it's correct.  Copying
Cyrl for tt from CLDR on the other hand could be dubious.  I'm
not in any linguistic business, propose a better solution.

We might be able to copy this info from your book, IIRC based
on public PDFs that would do the trick for frr, frs, and fy,
and not run into any tt trap.

> This is just too weird. It makes no sense at all.

The Suppress-Script makes sense for implementors of RFC 4647,
and potentially for CLDR, otherwise it's not very interesting.

In the case of frr its 3..5 dialects are more important than
a Suppress-Script, but I think it's better if somebody actually
needing this to tag content requests these variants.

Some folks collect obscure scripts, others are fascinated by
obscure country codes and TLDs, but for "weird" and "nonsense"
it should be Talossa or worse (and yes, that's possible, see
the wikipedia micronation talk page).

BTW, I think that all requests based _alone_ on Wikipedia info
should be rejected, it's relatively simple to game Wikipedia.

Frank
-- 
<http://en.wikipedia.org/wiki/Talossan_language>