Missing subtags 003 and 172

Doug Ewell doug at ewellic.org
Fri Jul 30 19:13:40 CEST 2010

Mark Davis ☼ <mark at macchiato dot com> wrote:

> We deliberately decided in the course of the BCP47 update to include
> all of the codes except for the following.
>        B.  UN numeric codes for 'economic groupings' or 'other
>            groupings' MUST NOT be registered in the IANA registry and
>            MUST NOT be used to form language tags.

In fact, that passage was unchanged during the update from RFC 4646
(where it was Section 2.2.4, item 3.B) to RFC 5646 (where it became item

> We made 830 optional, specifically because it is somewhat deprecated
> by the addition of the country codes for the islands instead. That is
> not the case for 003 and 172; these are as or more useful than other
> region codes.

I'm particularly curious about the usefulness of 172 in language
tagging.  I'm trying to think of a language, commonly spoken in the
far-reaching supranational region currently known as the CIS, such that
there would be a benefit to tagging content in that language as "xx-172"
instead of either pinning it down to the country level, or else simply
writing "xx".  So far I'm drawing a blank.

003 is a little better.  There might arguably be a use case for
expressing "Spanish in the U.S. or Cuba or the Dominican Republic or
Costa Rica, but not Colombia or Chile or Spain."  I wonder how detailed
we want to try to get with region subtags.

> There are some other codes (like 062) that we can deprecate in CLDR,
> but not these.

So this is just for CLDR after all.  Do you have use cases for these
subtags *for language tagging*?

> At the time you made the update to add the other M.49 codes I had
> presumed that they were included. It wasn't as if you raised the issue
> and we agreed.

We didn't add any other M.49 code elements between 4646 and 5646, did
we?  We did add exceptionally reserved ISO 3166-1 code elements.

> Do you really think we should drag everyone through the pain of doing
> another RFC just to fix these two codes?

No, I don't.  On the contrary, I don't think it is obvious that these
should be "fixed."  I believe the LTRU threads from 2005 will show we
excluded them intentionally and with justification.  I can dig through
the archives to find chapters and verses, but since the archives are
available only via HTML and my saved digests (in Outlook Express format,
not as easily searched as plain text) this will take some time.

What I really think is that we should follow the rules we created.

Doug Ewell | Thornton, Colorado, USA | http://www.ewellic.org
RFC 5645, 4645, UTN #14 | ietf-languages @ is dot gd slash 2kf0s ­

More information about the Ietf-languages mailing list