Use of "CS" for 'Serbia and Montenegro'

Peter Constable petercon at
Mon Jan 10 06:53:03 CET 2005

> From: ietf-languages-bounces at [mailto:ietf-languages-
> bounces at] On Behalf Of Bruce Lilly

> > Of course, this is an exceptional case; "sr-CS" is of a form that
> > doesn't require registration.

> As you have noted, registration is not necessary.

I did not quite say it was not necessary; at least, there is a subtle
distinction to be made in relation to what I was meaning.

Per RFC 3066, it is not required that "sr-CS" be registered before that
language tag can be used.

Because of ambiguity of the semantics, it may be necessary to ensure
inter-operability to register "sr-CS" to eliminate that clarify the
intended semantics of that tag.

> As I have noted
> in previous discussion, registration would achieve nothing, since
> interpretation per RFC 3066 would remain as a 2-letter language
> code followed by a two-letter country code, the latter being the
> definition provided by ISO 3166 (i.e. the alleged ambiguity would
> remain).

On the contrary, registration could serve to eliminate the ambiguity. 

And just to make sure we don't end up speaking past one another due to
unstated assumptions, my assumption is that ambiguity in the semantics
of this metadata element is a bad thing.

Peter Constable

More information about the Ietf-languages mailing list