Reshat Sabiq's requests for two Tatar orthographic variants

"Reshat Sabiq (Reşat)" tatar.iqtelif.i18n at
Wed Jan 10 07:14:58 CET 2007

Hash: SHA1

Frank Ellermann yazmış:
 > I'm not yet sure, what other (apparently non-Turkic) languages are you
> talking about ?  
Some that i've cursively verified are:
Abkhaz, Chechen, Tajik. But i think almost any language in that
geography other than Georgian, Armenian, and Slavic languages is a
candidate for being one of the 70. This can be looked at if it comes to it.

> Your option (2) appears to be better defined, and it's
> also more directly related to what you really want.  But otherwise it's
> IMO okay to register a variant with some prefixes, later adding more
> prefixes as needed - probably not all 70, purely theoretical constructs
> aren't relevant for the language subtag registry, it's no encyclopedia.
> JFTR, if you register "whatever" as variant for tt (e.g. prefix tt-Latn),
> you can't later reduce this to prefix tt-Latn-newturka resulting in tags
> tt-Latn-newturka-whatever.  On the other hand if you register "newturka"
> with (among others) prefix tt-Latn, you can later add "whatever" with a
> prefix tt-Latn-newturka.
> The reason for this rule in 4646 are the potential tags tt-Latn-whatever
> created before the introduction of newturka, it's okay to make the set
> of permitted prefixes larger, it's not okay to remove permitted prefixes.
I never looked at it that way, but looks like you are referring to
nesting, as in:

I see 2 disadvantages to doing that:
1. it gets harder to read
2. it doesn't give a whole lot of advantage over just having 1 variant
per language.


- --
My public GPG key (ID 0x262839AF) is at:
Version: GnuPG v1.4.2.1 (Cygwin)


More information about the Ietf-languages mailing list