Counting Heads

John Cowan jcowan at
Wed May 28 14:30:28 CEST 2003

Doug Ewell scripsit:

> I suppose I should have been clearer.  Of course "1901" is not a script.
> What I meant to say was, since RFC 3066 specifies that two-letter second
> subtags are ISO 3166-1 country codes, the country code (if present)
> should always be the second subtag.  

That's not actually stated, merely that *if* a second subtag is two letters
it is always 3166-1.  There is no reason why (if other considerations
compel it) that 3166-1 tags can't appear in the third subtag or even later.
For example, if Hakka came in national variants, it would be unreasonable
to encode them as zh-TW-hakka and zh-CN-hakka; surely the natural forms
are zh-hakka-TW and zh-hakka-CN.

> I couldn't agree more that script differences are more significant than
> national usage differences or orthographic reforms.  I just didn't think
> there was unanimity that the subtags had to be ordered left-to-right by
> significance,

Well, left matching is specified in the RFC as an option, with the warning
that it may not always work: this of course primarily serves people who
prefer en-us but are willing to fall back to en and expect that to match
en-uk, and similar cases.

> especially since the order of significance isn't the same
> for every application (I'd rather read es-Latn or de-Latn than en-Arab,

So would I, but given that we agree that script is usually more significant
and that some choice needs to be made, we might as well put script left of

> But no matter what order is finally chosen, software will be written
> that supports these extended tags, and everything will be OK.  No
> buildings will collapse if the order of subtags isn't defined
> "correctly."

No, but fallbacks may be made unnecessarily more difficult, a thing important
during periods of transition (and our own period, as Adam said to Eve, is
inevitably a period of transition).

John Cowan   <jcowan at>
"One time I called in to the central system and started working on a big
thick 'sed' and 'awk' heavy duty data bashing script.  One of the geologists
came by, looked over my shoulder and said 'Oh, that happens to me too.
Try hanging up and phoning in again.'"  --Beverly Erlebacher

More information about the Ietf-languages mailing list