Re: [Ltru] Re: Registry deltas

John Cowan <cowan at ccil dot org> wrote:

> I'd be more in favor of merging descriptions that are a prefix of 
> other descriptions:  if we have both "Foo (Bar)" and "Foo", to 
> eliminate the latter in favor of the former.  I don't feel strongly 
> about this, however.

There's some fairly strong aopposition to that, from Debbie and (I 
believe) others who feel at least one Description field needs to match 
the name in each of the corresponding standards.  That's why "bas" is 
listed as both "Basa" (from 639-2) and "Basa (Cameroon)" (from 639-3), 
and "bzw" is listed only as "Basa (Nigeria)" since it appears only in 
639-3 and not in 639-2.

I'm not at all happy with having one of the Basas be listed as just 
"Basa", but I suppose I can't argue with the logic.  We do say that the 
purpose of the Description is "to indicate the meaning of that 
identifier [subtag] as defined in the source standard."  If we point 
back to the source standard for meaning, the names do have to match.

>> Grandfathered tag sgn-CH-de
>> Add: Preferred-Value: sgn-sgg
>>      Deprecated: 2007-01-01
>>      Comments: replaced by ISO code sgg
>
> These "replaced by" comments existed in the RFC 3066 registry as a 
> kind of crude Preferred-Value, and we now have a better mechanism.  I 
> can easily see someone looking at this entry and not knowing whether 
> to use "sgn-sgg" or "sgg".

I did some spelunking and found where the "requirement" came from to add 
these bogus comments.  Early copies of draft-ietf-ltru-registry, the 
document that became RFC 4646, included a section 3.7 called 
"Initialization of the Registry" (note, singular -- there was no 
extensions registry yet) that contains a good deal of text that was 
later moved into draft-ietf-ltru-initial (RFC 4645).  That section 
included this passage:

"All other tags in the RFC 3066 registry that are deprecated will be 
maintained as grandfathered entries... The 'Comments' field will contain 
the reason for the deprecation."

By that time we already had invented Preferred-Value, a much better 
mechanism (as John points out), but the passage was there.  I never 
liked it -- for example, in 
http://www1.ietf.org/mail-archive/web/ltru/current/msg02481.html from 
June 2005 (search for the word "mindless") -- but decided to live with 
it.  In the meantime, the "will" clause became "SHOULD" and then "may 
optionally" but the comments remained.

I am 100% in favor of deleting, as part of the RFC 4645bis Registry 
reload, every Comments field of the form "replaced by ISO code xxx" 
where at best the same information can be gleaned from the 
Preferred-Value, and at worst the comment may add confusion.  That is to 
say, all of them.  RFC 4646 allows Comments fields to be added or 
deleted with impunity; if we can do this as part of the update instead 
of taking it to ietf-languages I am all for it.  Comments that add no 
real information do not need to clutter up the Registry.

--
Doug Ewell  *  Fullerton, California, USA  *  RFC 4645  *  UTN #14
http://users.adelphia.net/~dewell/
http://www1.ietf.org/html.charters/ltru-charter.html
http://www.alvestrand.no/mailman/listinfo/ietf-languages

_______________________________________________
Ltru mailing list
Ltru@ietf.org
https://www1.ietf.org/mailman/listinfo/ltru