Generic variants and Armenian dialects

[Doug Ewell]

Misha Wolf <Misha dot Wolf at reuters dot com> wrote:

> The only bit of your message that worries me is the discussion of 
> "en-1901".  Yes, I know you were commenting on a mail of John's.
>
> It's one thing to argue that we should use the Armenian word for East 
> or West to ensure the variants are not used in an arbitrary manner 
> elsewhere.  It's quite another to assign a year number to one language 
> only.  What if two languages underwent spelling reform in 2007?  Would 
> one of them not be allowed to use "2007"?  This seems wrong to me.

Given that we have implemented "spelling reform" variant subtags for 
only one language, and the practical need for those has been called into 
question, this scenario seems quite remote.  Still, anything is 
possible.

There is a special production in the ABNF to allow a 4-character variant 
that begins with a digit.  The obvious intent of this is to indicate the 
year in which some change was made in the language that is significant 
enough to justify a variant (such as a spelling reform).  Perhaps LTRU 
should consider defining a distinction between this type of "date" 
variant, which could in theory apply to more than one language (even 
though the reforms themselves would be different), and the "usual" type 
such as Resianic or Boontling or Scouse or Eastern [Armenian], which in 
theory could not ("Eastern Quinqui" would not have the same relative 
meaning as "Eastern Armenian," except geographically).

--
Doug Ewell
Fullerton, California, USA
http://users.adelphia.net/~dewell/
Editor, draft-ietf-ltru-initial