[Ltru] Re: Finishing off #1026 (Was: Re: status? last call?)

["Doug Ewell" <dewell at adelphia.net>]

Randy Presuhn <randy underscore presuhn at mindspring dot com> wrote:

>> The draft is not consistent with itself.  Section 2.2.4 (3E), quoted
>> above, says we can register such subtags.  Section 3.4 (11), to which
>> I referred, says we may not.
> ...
>
> I think you mean section 3.3 (11).

Yep.

> The text there in the
> http://www.ietf.org/internet-drafts/draft-ietf-ltru-registry-08.txt
> draft says:
> ...
>         This rule exists so that UN M.49 codes remain available as the
>         value of last resort in cases where ISO 3166 reassigns a
>         deprecated value in the registry.
>
> The difference in our understanding comes from the last sentence.  I
> think Doug is reading it as the *sole* condition under which rule
> 3.3(11) may be employed.

That is correct.

> I, and I think Frank, read it as the *motivation* for the
> rule, rather than a limiting condition.  The two readings play out
> differently if the ISO 3166 registration authority fails to assign a
> code when petitioned.  In Doug's reading, there's no registration,
> and the party needing the code is out of luck.  In the alternative
> reading, after some undefined (I'd leave it up to the collective
> intelligence of the language tag mailing list) period, the UN M.49
> code could be registered.  The question for the WG is which behaviour
> is better / less bad?

It depends on how likely we think this scenario is (using Guernsey as an
example; substitute Jersey or IOM as appropriate):

We add region subtag 831, and then at some point in the future, ISO
3166/MA adds code element GG, resulting in widespread popular
association of the symbol GG with Guernsey, and criticism of the LTRU
project for having assigned 831 instead.

This is what I thought one of Frank's biggest concerns was.  (The other
was that the MA, having reserved GG for Guernsey, would assign it to
some other entity X first, then withdraw it and assign it to Guernsey,
which could no longer have region subtag GG because it was already
permanently assigned in the registry to entity X.  This chain of events
seems too unlikely to worry about.)

The alternative is that we *don't* add region subtag 831, but either (a)
wait for the unrealistic scenario above, resulting in 831 after all, or
(b) wait for the MA to assign GG, resulting in GG.  In this scenario, we
might wait forever, and in the meantime it would be impossible to
indicate Guernsey in a language tag, except with a private-use subtag
(or a non-conformant extension).

I wasn't worried about this impossibility before, since the use of
English, Manx, etc. in these locations seems not to differ from other
locations.  However, I did a bit of checking, and can't figure out a way
to tag Dgèrnésiais (the Norman language of Guernsey) or Jèrriais (the
Norman language of Jersey).  There doesn't seem to be an ISO 639-1 or -2
code element (and therefore a language subtag) for Norman languages in
general, and definitely not for these two specifically.  ISO/DIS 639-3
has only a single code element "xno" for "Anglo-Norman," which may or
may not fit the description of either Dgèrnésiais or Jèrriais; but even
if it is, and even if RFC 3066ter includes "xno" as a language subtag,
there is no way to use it to differentiate between "Anglo-Norman used in
Guernsey" (i.e. Dgèrnésiais) and "Anglo-Norman used in Jersey" (i.e.
Jèrriais) unless region subtags exist for Guernsey and Jersey.

(Of course, ietf-languages could avoid this quagmire by petitioning ISO
639/RA to assign code elements for Dg. and J., or failing that,
registering them directly.)

So if we want to allow 831 and friends to be individually registrable,
cool, but let's not come back later and say, "Wait a minute!  We needed
831 as a safety net!  What were we thinking?"

> I propose adding one sentence to 3.3(11) to make it more consistent
> with 2.2.4 (3E):  "If the petition for a code assignment ISO 3166 is
> refused or not acted on in a timely manner, the Language Subtag
> Reviewer MAY procede with the registration using the UN M.49 code."
> Or words to that effect.

If we do this, the text in Section 2 (6) and Section 4 of
draft-initial-01 will remain unchanged in draft-02.  (It had been
proposed to change it to indicate that the subtags could not be
registered.)  So the decision on this question affects both
draft-registry and draft-initial.

--
Doug Ewell
Fullerton, California
http://users.adelphia.net/~dewell/



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