[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Ltru] Re: [psg.com #1026] what to do about Guernsey and Jersey



Frank Ellermann <nobody at xyzzy dot claranet dot de> wrote:

>> The presence of both 830 *and* 831 and 832, none of which was
>> permissible under RFC 3066, is not likely to cause any great
>> difficulty in the real world.
>
> If they deprecate 830 too late we also deprecate it => ready.

True.

>> But it SHOULD be noted that if we do allow 831 and 832 (and
>> 833), and if Addison is correct, we CANNOT subsequently
>> register GG and JE (and IM) if those codes are added to ISO
>> 3166/MA, and make them the Preferred-Values for the numeric
>> codes.  I think there is a common assumption that we (or
>> ietf-languages) would do that.
>
> Not my assumption.  I got the idea that 3066bis is a one-way
> street to UN numbers in slow motion.

I did not get that idea, unless by "slow motion" you mean hundreds and
hundreds of years.

<soapbox>
The great pace of national change, fueled by the national
self-determination of colonial holdings during the 20th century and the
rise of nationalism during the 19th and 18th, has essentially concluded.
Most nations on Earth -- probably 75% or more -- have attained more or
less the same status they will continue to have for many, many years to
come.  Even the great changes of the 1990s -- the breakup of the Soviet
Union and Yugoslavia, the reunification of Germany and Yemen -- affected
fewer than 10% of the currently recognized nations.  Minor changes in
governmental structure will not materially affect the relevancy of ISO
3166; changes in name or other major changes are not likely to occur at
the pace necessary to render ISO 3166 worthless for our purposes.  I
expect to be around in 50 years to see how well these predictions have
held up.  :-)
</soapbox>

> You (all here) didn't
> like my plan to do this fast (at date-B) and for all regions.

I did not.  We have to stay backward compatible with RFC 3066, and that
means sticking with alpha-2 region subtags.  I will not argue this
further.

> As long as GG, IM, and JE still _can_ be listed a.s.a.p. it's
> no real problem if their Preferred-Value sticks to 831...833.

I don't believe they can be added if 831 through 833 already exist.
Addison and Mark can confirm or deny this.

>> they cannot be superseded by ISO codes that are added later.
>
> Is there a "normal" procedure that the UN numbers are "always"
> changed first, and that 3166-1 follows suite later ?  We could
> add some text to the draft allowing for this delay, avoiding
> premature registrations of unnecessary / redudant UN numbers.

ISO 3166/MA states that they do not add a code for a country unless it
appears in the UN list.  This allows them to avoid the controversy of
"defining" what is a country, while giving them authority to assign
codes to the things the UN has defined as countries.  This is as it
should be.  So yes, there is a procedure that UN coding comes before
ISO.

What ISO does not promise is that UN coding *will automatically* lead to
ISO coding.  So far it seems to have happened 100% of the time, even for
Åland, which is why I suggested waiting until ISO made a decision before
adding subtags for these entities.  But many people have pointed out
that we can't delay the draft indefinitely, waiting for ISO to make a
decision.  (They're not known for adhering to published schedules in
doing so.)

Later:

> In the hypothetical case TP -> TL, and TL changing its name
> again, we get TP -> TL and TL -> XX.  Is that really necessary,
> why not just update TP -> XX in this case ?  A very minor nit.

I would greatly prefer this, but it seems not to have met with much
approval.

> | Codes assigned by ISO 639, ISO 15924, and ISO 3166 that do
> | not conflict with existing subtags of the associated type and
> | whose meaning is not the same as an existing subtag of the
> | same type are entered into the IANA registry as new records
> | and their value is canonical for the meaning assigned to
> | them.
>
> That's apparently an error, let's say "ISO 3166, and UN M.49"
> instead of (only) "and ISO 3166".

No, that's on purpose.  ISO alpha-2 codes are considered the standard
form for region subtags, while UN M.49 numeric codes are considered a
fallback in case the ISO approach causes problems (as with CS) or for
the macrogeographical regions that are outside the scope of ISO 3166.
UN codes and ISO 3166 codes do not supersede one another.

> And where's the rule about a new code that does _not_ conflict
> with an existing code, but whose meaning _is_ the same as an
> existing subtag ?  The TL -> XX in my example ?  We should
> simply delete this part resulting in:
>
>   Codes assigned by ISO 639, ISO 15924, ISO 3166, and UN M.49
>   that do not conflict with existing subtags of the associated
>   type are entered into the IANA registry as new records and
>   their value is canonical for the meaning assigned to them.

If an ISO code does not conflict with an existing subtag, and has the
same meaning as an existing subtag, aren't they the same code?  I must
be missing something here.  Give me a example of what you mean.

> Adding a statement about the "same meaning" case:
>
>   If existing codes have the same meaning a Preferred-Value is
>   added or updated to show the new code.  UN M.49 codes are
>   only added if no ISO 3166 code with the same meaning exists.
>
> In this form GG, IM, and JE would replace 831..833.

I think I understand what you mean.  It would have to be up to the WG to
decide, at this very late stage, whether to change the policy in this
way.

> This isn't
> what we think how it should work, something with my proposal is
> still wrong.  But something in draft -04 is also still wrong.

Understandably, Randy will want something more concrete than this.

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



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




Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.