Re: Language tag education and negotiation

John Cowan wrote:
> Asmus Freytag scripsit:
>
>   
>> There are parts of the planets where it is common for people to command 
>> more than one language. 
>>     
>
> Most of it, indeed.
>
>   
>> Of course, a meta tag that (reliably :-) ) described something as 
>> 'translation', or conversely as 'official language version' would be 
>> useful, too.
>>     
>
> This would be a good use case for a BCP 47 registered extension,
> something like 't-*' to report the translation status of a document.
> Off the top of my head, the obvious candidates would be t-original,
> t-authentic (for documents which are "equally authentic" in all language
> versions), t-polished, t-rough, and t-machine.
>   

I see some overlap with the functionality ITS "Translate" offers, see
http://www.w3.org/TR/its/#trans-datacat
while ITS "Translate" is meant to express a directive ("Translate this 
part, but not this!"), your "t-*" extension is meant to express a 
status. This seems to be useful, however, I'm not sure if having it as 
part of a language tag helps. Imagine a localization project in which 
several people and tools work on the same parts of a document, and you 
need to assign several status, changing over time, like 
"t-rough-version1", "t-machine-implementation1-version1", 
"t-machine-implementation2-version2" etc. This is an open, 
project-specific list of values. Another example of objects which need 
different values, but are in the same domain (from a localization point 
of view) seem to be images which are "localized", but not translated.

Felix