Tagging transliterations from a specific script
Michael Everson
everson at evertype.com
Thu Mar 17 14:10:47 CET 2011
On 17 Mar 2011, at 12:17, Kent Karlsson wrote:
> Den 2011-03-17 12:02, skrev "Michael Everson" <everson at evertype.com>:
>
>> Then why not tt-alalc97-Arab, where Latin is implied in the hierarchy first?
>
> Because the ABNF does not allow for that; once you start with 5-8 alphanums variant subtags, you cannot have a script subtag after that, except in a private use part or in an extention part (in the latter two cases, what looks like a script subtag might not be referring to a script, or it might).
I expected this answer, but how about IGNORING that stricture for the moment and dealing with the conceptual implications of this level of hierarchy?
There are a zillion ways to transliterate (source script, target script), and I want to avoid adding a zillion new special subtags for handling this if it should be handled in terms of hierarchy.
> You could argue that that was due to lack of foresight, but that is how it is now without revising BCP 47.
Perhaps it is a good idea to consider revising it.
Michael Everson * http://www.evertype.com/
More information about the Ietf-languages
mailing list