Results of Duplicate Busters Survey #2

Doug Ewell doug at
Tue Aug 26 06:49:51 CEST 2008

And here are the results from the second survey, the one that dealt with 
(what I consider to be) duplicate Description fields within a single 
draft-4646bis Registry entry.  These results, and my recommendations 
based on them, are submitted to Michael Everson as Language Subtag 
Reviewer for his final decisions, which I will capture in draft-4645bis.

I divided the relevant Registry entries into two types:

1.  those where the Description fields differ only in diacritical marks 
or differences in hyphens or apostrophes

2.  those where the Description fields are equivalent except that one 
ends with a parenthetical comment such as "(macrolanguage)" or "(Japan)" 
to differentiate it from another.

Only four people besides me (John Cowan, Frank Ellermann, Kent Karlsson, 
C.E. Whitehead) stated preferences in this survey.  If anyone feels 
their comments are not fairly expressed here, or believes they made 
comments (on- or off-list) that are not captured here, please reply as 
soon as practical, citing the date of the message.

Regarding the first type, there was fairly strong preference for leaving 
all of the Description fields intact.  Only John (and I) supported 
deleting any of the existing fields.  Therefore, my
recommendation is to leave the existing records for 'Ethi', 'Hang', and 
'Hano' alone, and furthermore to add a new Description field to the 
existing subtag 'nqo' for consistency with this preference:

Type: language
Subtag: nqo
--> ADD: Description: N'Ko
Description: N’Ko
Added: 2006-06-05
Suppress-Script: Nkoo

The additional Description field, with the plain ASCII apostrophe, is 
from ISO 639-3.

ISO 639-3 solved the conflict between "Macedo Romanian" and 
"Macedo-Romanian" for us, by adding the hyphen they had previously 

Regarding the second type, most respondents favored keeping the 
Description field with the parenthetical comment (typically from ISO 
639-3) and deleting the one without (typically from ISO 639-1 
and/or -2).  John and Kent concurred with me on this; I didn't see an 
opinion from C.E. here.

Frank dissented on the "(macrolanguage)" entries, preferring to keep 
both "Swahili" and "Swahili (macrolanguage)" in those cases, and 
concurred on the "(country name)" entries, but only if a Comments field 
were added to the record, something like this:

Type: language
Subtag: ain
Description: Ainu (Japan)
Added: 2005-10-16
Comments: Listed as "Ainu" in ISO 639-2

Resolving the duplication between "Ainu" and "Ainu (Japan)" by 
relegating the former to a comment, but retaining the duplication 
between "Swahili" and "Swahili (macrolanguage)", continues not to make 
any sense to me.  Based on the support of other respondents, my 
recommendation to the Reviewer is to keep only the ISO 639-3 names as 
Description fields for the subtags listed below, and -- at the 
Reviewer's sole discretion -- to either capture the ISO 639-1/2 names in 
Comments fields, or omit them altogether.

The subtags in question are:

ms, sw, ain, bas, bem, chm, doi, fan, gba, kam, kok, men, war

As an example, the recommendation is to change the existing record for 
language subtag 'ain' as shown above.

Please note that I have not consulted with Michael on anything related 
to these surveys.  He may not even have remembered that I wanted him to 
make the final determination.

Doug Ewell  *  Thornton, Colorado, USA  *  RFC 4645  *  UTN #14  ˆ

More information about the Ietf-languages mailing list