Encoding scripts in tags: evil or just unpleasant?

[Michael Everson]

Serbian, Croatian, and Bosnian are, linguistically speaking, closely 
related dialects of the same language. We've got three codes for them 
now, and Bosnian was totally political, and now how many more do you 
want? Let's say they can ALL be written in two scripts. Shall we add 
six more codes to the three we already have?

I thought we did not wish to encode duplicates and redundancies.

More and more I go back to the original thought: 15924 codes should 
be applied in some kind of script tag rather than trying to conflate 
all this stuff in a language tag, and there should be an RFC to 
specify script tagging. Then languages would be identified as 
languages and orthographies would be orthogonal to that.

Wrong? Possibly. But I want to know that everybody think this RFC is 
for precisely this thing because I am not sure of that.
-- 
Michael Everson * * Everson Typography *  * http://www.evertype.com