We should follow procedures and go ahead with the ballot after we finish
discussion. This was a formal request and should be treated as such.
I remember this same question came up awhile back (about Hawaiian), where
someone asked why we didn't have a code for it. I wrote them back and said
there was no alpha-2 but there was an alpha-3 code. They were happy with
that. Perhaps that is irrelevant because it was a different person
The principle we established about not defining an alpha-2 code when there
was already an alpha-3 was so that the code used would be predictable
(i.e. because of the wording in RFC 3066 about which one to use). Yes, in
some cases there are synonyms and it is better to consider these synonyms
than alternatives. But I would think it better to reduce the number of
cases where the system has the burden of having to recognize synonyms. I
also thought that another reason behind the principle is to encourage
people to allowing alpha-3 codes in their systems because they are much
more extensible (obviously) than alpha-2 codes. Using the reasoning of the
requestor for Hawaiian, it doesn't encourage developers to begin
recognizing alpha-3 codes and greatly limits the number of languages that
can be coded.
On Wed, 24 Apr 2002, [iso-8859-1] Håvard Hjulstad wrote:
> I accept the argument below. However, take a look at the following: Walloon
> recently got an alpha-2 and an alpha-3 identifier. That was ok, because they
> didn't have an alpha-3 identifier to begin with. Hawaiian on the other hand
> is in the "unfortunate" situation of already having an alpha-3 identifier,
> so they cannot get an alpha-2 identifier. This is NOT logical.
> The reason for "freezing" was that there is this principle of "alpha-2 if
> exists - alpha-3 otherwise" in parts of industry (WHO IS IMPORTANT! I am not
> questioning that!). But we have already accepted, and we have presented as a
> fact, that "synonym" language identifiers have to be accepted, and that
> there will be more of them as further work with (new parts of) 639 moves
> along. The principle needs (over time) to be changed from "alpha-2 if
> exists - alpha-3 otherwise" to "synonyms allowed".
> We have other principles too. All of them need to be taken into
> consideration in each case.
> Next time a similar case comes up, I will initiate the discussion in a
> similar manner. Then Michael is probably going to scream. An we will weigh
> the scream (and the underlying argument, which I accept) against other
> arguments and principles. And then we will vote.
> Because there is one detail in Michael's reasoning that I do NOT accept: He
> wants to sanctify one principle beyond the democratic process. He doesn't
> even want a ballot. I am sure that the balloting members are able to apply
> all relevant principles and cast their votes accordingly.
> Do we have any discussion on the merit of adding an alpha-2 identifier for
> Hawaiian? Think of it as follows: Would we have included an alpha-2
> identifier if they didn't have an alpha-3 identifier? If not, we don't have
> a problem at all. If the answer would have been yes, we do need to deal with
> it, weighing all arguments in a proper manner.
> Best regards,
> Håvard Hjulstad mailto:[log in to unmask]
> Solfallsveien 31
> NO-1430 Ås, Norway
> tel: +47-64944233 & +47-64963684
> mob: +47-90145563
> -----Original Message-----
> From: ISO 639 Joint Advisory Committee [mailto:[log in to unmask]]On Behalf
> Of Rebecca S. Guenther
> Sent: 24. april 2002 19:08
> To: [log in to unmask]
> Subject: Re: New ISO 639-1 identifier ? - Hawaiian
> I agree with Michael's reasoning here. A principle may not be a law, but
> what's the point of establishing principles if you don't attempt to honor
> them. I see no reason not to stand by this principle here.
> On Wed, 24 Apr 2002, Michael Everson wrote:
> > At 16:02 +0200 2002-04-24, Håvard Hjulstad wrote:
> > >I also know the principle ...! I also know that it is a principle, and
> not a
> > >law.
> > As IETF Language Tag Reviewer, I formally object to this. It is an
> > agreement made by the JAC while RFC 3066 was being revised. The
> > agreement was to ensure internet stability for language tagging.
> > The JAC must NOT go back on its promise. This is a serious matter of
> > >QUESTION: Is anyone in favour of adding an alpha-2 identifier for
> > >If I don't get any feedback in favour of this within the next week, I
> > >send out a ballot with a recommendation to vote "no".
> > Certainly not. It violates the rules of the JAC. That answer should
> > be given to the Hawai'ians. No ballot should be sent out.
> > --
> > Michael Everson *** Everson Typography *** http://www.evertype.com